THE MICROSOFT JOURNAL FOR DEVELOPERS

JANUARY 2015 VOL 30 NO 1

magazine

14 Top Features
of Visual Basic 14.................16

0115msdn_CoverTip.indd 1

12/9/14 3:26 PM

0115msdn_CoverTip.indd 2

12/9/14 4:44 PM

THE MICROSOFT JOURNAL FOR DEVELOPERS

JANUARY 2015 VOL 30 NO 1

magazine

14 Top Features
of Visual Basic 14.................16

Protect Your Online Services with Big Data

and Machine Learning

VISUAL STUDIO 2015

& MICROSOFT AZURE

Alisson Sol, Don Ankney and Eugene Bobukh . . . . . . . . . . . . . . . . . . . . . . 34

VISUAL BASIC 14

Occasionally Connected Data

in Cross-Platform Mobile Apps

WEB-BASED TEST CASE

MANAGEMENT

Kevin Ashley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

BYOD Challenge: Connect Mobile Devices

to the Enterprise

Lucian Wischik, page 16

Manoj Bableshwar, page 22

HDINSIGHT AND HADOOP

Omid Afnan, page 28

Greg Oliver and Bruno Terkaly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Understanding TypeScript

Peter Vogel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

Adding Application Data to the

Windows Phone 8.1 Map Control

Keith Pijanowski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Cross-Platform Cloud Automation

with JavaScript

Steven Edouard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

COLUMNS
CUTTING EDGE

Mobilize an Existing Web Site

Dino Esposito, page 6

DATA POINTS

Looking Ahead to Entity

Framework 7
Julie Lerman, page 10

TEST RUN

Logistic Regression Classification

with Multi-Swarm Optimization
James McCaffrey, page 74

DONT GET ME STARTED

An Upstart Again

David Platt, page 80

Untitled-7 2

12/4/14 2:22 PM

Untitled-7 3

12/4/14 2:23 PM

Instantly Search
Terabytes of Text
25+ fielded and full-text search types
dtSearchs own document filters
support Office, PDF, HTML, XML, ZIP,
emails (with nested attachments), and
many other file types
Supports databases as well as static
and dynamic websites
Highlights hits in all of the above

APIs (including 64-bit) for NET, Java,

C++, SQL, etc.

JANUARY 2015 VOLUME 30 NUMBER 1

magazine

KEITH BOYD Director

MOHAMMAD AL-SABT Editorial Director/mmeditor@microsoft.com
KENT SHARKEY Site Manager
MICHAEL DESMOND Editor in Chief/mmeditor@microsoft.com
LAFE LOW Features Editor
SHARON TERDEMAN Features Editor
DAVID RAMEL Technical Editor
WENDY HERNANDEZ Group Managing Editor
SCOTT SHULTZ Vice President, Art and Brand Design
JOSHUA GOULD Art Director
SENIOR CONTRIBUTING EDITOR Dr. James McCaffrey
CONTRIBUTING EDITORS Rachel Appel, Dino Esposito, Kenny Kerr,
Julie Lerman, Ted Neward, David S. Platt, Bruno Terkaly, Ricardo Villalobos

Henry Allain President, Redmond Media Group

Becky Nagel Vice President, Digital Strategy
Michele Imgrund Vice President, Lead Services Division
Tracy Cook Director, Client Services & Webinar Production
Irene Fincher Director, Audience Development & Lead Generation Marketing
ADVERTISING SALES: 818-674-3416/dlabianca@1105media.com

lightning fast Redmond Magazine

Dan LaBianca Chief Revenue Officer

Chris Kourtoglou Regional Sales Manager
Danna Vedder Regional Sales Manager/Microsoft Account Manager
David Seymour Director, Print & Online Production
Anna Lyn Bayaua Production Coordinator/msdnadproduction@1105media.com

covers all data sources eWeek

results in less than a second InfoWorld

Rajeev Kapur Chief Executive Officer

Henry Allain Chief Operating Officer
Richard Vitale Senior Vice President & Chief Financial Officer
Michael J. Valenti Executive Vice President
Erik A. Lindgren Vice President, Information Technology & Application Development

hundreds more reviews and developer

case studies at www.dtsearch.com
dtSearch products: Web with Spider

Desktop with Spider

Engine for Win & NET-SDK

Network with Spider

Engine for Linux-SDK

Publish (portable media) Engine for Android-SDK beta

Document Filters included with all products, and
also available for separate licensing

Jeffrey S. Klein Chairman of the Board

MSDN Magazine (ISSN 1528-4859) is published 13 times a year, monthly with a special issue in
December by 1105 Media, Inc., 9201 Oakdale Avenue, Ste. 101, Chatsworth, CA 91311. Periodicals
postage paid at Chatsworth, CA 91311-9998, and at additional mailing offices. Annual subscription
rates payable in US funds are: U.S. $35.00, International $60.00. Annual digital subscription rates
payable in U.S. funds are: U.S. $25.00, International $25.00. Single copies/back issues: U.S. $10,
all others $12. Send orders with payment to: MSDN Magazine, P.O. Box 3167, Carol Stream, IL 60132,
email MSDNmag@1105service.com or call (847) 763-9560. POSTMASTER: Send address changes to
MSDN Magazine, P.O. Box 2166, Skokie, IL 60076. Canada Publications Mail Agreement No: 40612608.
Return Undeliverable Canadian Addresses to Circulation Dept. or XPO Returns: P.O. Box 201, Richmond
Hill, ON L4B 4R5, Canada.
Printed in the U.S.A. Reproductions in whole or part prohibited except by written permission. Mail
requests to Permissions Editor, c/o MSDN Magazine, 4 Venture, Suite 150, Irvine, CA 92618.
Legal Disclaimer: The information in this magazine has not undergone any formal testing by 1105 Media,
Inc. and is distributed without any warranty expressed or implied. Implementation or use of any information
contained herein is the readers sole responsibility. While the information has been reviewed for accuracy,
there is no guarantee that the same or similar results may be achieved in all environments. Technical
inaccuracies may result from printing errors and/or new developments in the industry.
Corporate Address: 1105 Media,Inc.,9201 Oakdale Ave.,Ste 101,Chatsworth,CA 91311,www.1105media.com
Media Kits: Direct your Media Kit requests to Matt Morollo, VP Publishing, 508-532-1418 (phone),
508-875-6622 (fax), mmorollo@1105media.com
Reprints: For single article reprints (in minimum quantities of 250-500), e-prints, plaques and posters contact:
PARS International, Phone: 212-221-9595, E-mail: 1105reprints@parsintl.com, www.magreprints.com/
QuickQuote.asp

Ask about fully-functional evaluations

List Rental: This publications subscriber list, as well as other lists from 1105 Media, Inc., is available
for rental. For more information, please contact our list manager, Jane Long, Merit Direct. Phone: 913685-1301; E-mail: jlong@meritdirect.com; Web: www.meritdirect.com/1105
All customer service inquiries should be sent to MSDNmag@1105service.com or call 847-763-9560.

The Smart Choice for Text Retrieval since 1991

www.dtSearch.com 1-800-IT-FINDS
Printed in the USA

Untitled-2 1

11/3/14 11:03 AM

EDITORS NOTE

MICHAEL DESMOND

Back to Basic
When Microsoft convened its Connect(); event in New York City
in November, the company released so many vital development
technologiesincluding open sourcing the .NET Core Runtime
and librariesthat it prompted us to publish an extra issue of
MSDN Magazine covering some of the new updates. Included
in that coverage were looks at updated cross-platform tooling in
Visual Studio, a dive into ASP.NET 5 and the Azure SDK 2.5, and,
of course, a run through the newest versions of C# and C++.
What we werent able to include in that special issue, but do
feature in our pages this month, is a trifecta of articles focused on
developments at Connect();. These include Lucian Wischiks 14
Top Improvements to Visual Basic 14, Omid Afnans Hadoop
Made Easier for Microsoft Developers, and Manoj Bableshwars
Web-Based Test Case Management with TFS.

Ultimately, the most important

improvement in this latest
version of Visual Basic may be the
ironclad commitment Microsoft
has shown for it.
I was particularly interested in Wischiks article, given that
Microsoft essentially rebuilt Visual Basic from the ground up for
the version 14 release. Wischik, a member of the Visual Basic/C#
Language Design Team at Microsoft, told me that his team spent
more than four years rewriting the programming language in
Visual Basic (it was written in C++ before). He described the effort
as both an opportunity and a challenge. Powerful new functionality, such as native refactoring and support for analyzers, are

enabled by Roslyn compiler technology that would be impossible to access without the rewrite. And Wischik says the wholesale
update offered an opportunity to rearchitect the language from
scratch based on what the group had learned over the years.
But the risks involved were not insignificant, he says. Youre
investing a lot of developer years just to regain the functionality
you already had, even before you start to add any new end-user
value. Theres also the risk that you might fail to reproduce all
of the quirks or bugs of the old code base that people might
inadvertently depend upon.
Complicating matters is the surprising complexity of Visual Basica
function of the extensive array of features that the language, with its
cultural roots in Classic Visual Basic, had accrued over the years. As
Wischik observes, the team faced the challenging task of both replicating and modernizing the developer experience in Visual Basic.
Looking forward, Wischik expects more IDE improvements to
arrive after Visual Studio 2015 ships. Weve been working on a Read
Eval Print Loop (REPL), a sort of souped-up immediate window
for quick experimenting and prototyping. We also want to write
more analyzers and other plug-ins, he says.
Wischik urges Visual Basic developers to migrate to Visual Studio
2015, because it provides a host of benefitsincluding faster compile
times and support for analyzerswhile allowing round tripping of
code back to Visual Studio 2013. From there, developers can make
the decision to move to Visual Basic 14, which boasts new features
like string interpolation, nameof and the ?. operator.
Ultimately, the most important improvement in this latest version of Visual Basic may be the ironclad commitment Microsoft
has shown for it.
Its the strongest vote of confidence a business can make,
Wischik says. We believe our product has such a strong future
in the decades to come, that we can invest all our developer years
now to prepare for that future.

Visit us at msdn.microsoft.com/magazine. Questions, comments or suggestions for MSDN Magazine? Send them to the editor: mmeditor@microsoft.com.
2015 Microsoft Corporation. All rights reserved.
Complying with all applicable copyright laws is the responsibility of the user. Without limiting the rights under copyright, you are not permitted to reproduce, store, or introduce into a retrieval system MSDN Magazine or any part of MSDN
Magazine. If you have purchased or have otherwise properly acquired a copy of MSDN Magazine in paper format, you are permitted to physically transfer this paper copy in unmodified form. Otherwise, you are not permitted to transmit
copies of MSDN Magazine (or any part of MSDN Magazine) in any form or by any means without the express written permission of Microsoft Corporation.
A listing of Microsoft Corporation trademarks can be found at microsoft.com/library/toolbar/3.0/trademarks/en-us.mspx. Other trademarks or trade names mentioned herein are the property of their respective owners.
MSDN Magazine is published by 1105 Media, Inc. 1105 Media, Inc. is an independent company not affiliated with Microsoft Corporation. Microsoft Corporation is solely responsible for the editorial contents of this magazine. The
recommendations and technical guidelines in MSDN Magazine are based on specific environments and configurations. These recommendations or guidelines may not apply to dissimilar configurations. Microsoft Corporation does not make
any representation or warranty, express or implied, with respect to any code or other information herein and disclaims any liability whatsoever for any use of such code or other information. MSDN Magazine, MSDN, and Microsoft logos are
used by 1105 Media, Inc. under license from owner.

4 msdn magazine

Some things are

just better together.
Make your Windows Azure environment
more appetizing with New Relic.

Get more visibility into your entire Windows Azure environment.

Monitor all your Virtual Machines, Mobile Services, and Web Sites
- all from one powerful dashboard.

newrelic.com/azure
2014 New Relic, Inc. All rights reserved.

Untitled-2 1

3/10/14 12:39 PM

CUTTING EDGE

DINO ESPOSITO

Mobilize an Existing Web Site

Its a common scenario. You arrive at a hotel, airport or some other
public space. You connect to the local wireless network and youre
redirected to a login page to enter your credentials. The login page
on your mobile device is tiny. All you can do is pinch-and-zoom
the screen, take the input form to a reasonable size and then type
in your credentials.
This is an annoying experience, yet it happens more often than
not. And its more than simply an annoyance. The future of Web
sites that dont render well on all devices is bleak. Not offering
mobile users a good experience can cut into business profits by
driving users away from the site and toward a competitors site.

The simplest approach

that doesnt require a full-site
rewrite and wont break any
existing features is to create a
separate site specific for your
mobile audience.
The magic wand too many developers wave to cast Web spells is
responsive Web design (RWD). RWD is definitely an option, but,
like anything else, it has its pros and cons. The problem with RWD
isfor the most partit requires a new project.
The basic idea of RWD is everything is fluid. Any elements will
never be wider than the screen. So by definition, RWD adapts to
an infinite number of screen widths. The point of RWD is to create a new Web site inspired by a new idea of content and targeting
a fixed and smaller number of break points. Pushing RWD or a
library like Twitter Bootstrap onto an existing Web site designed
for the desktop is hardly a viable option.
Furthermore, selling the RWD approach is easy. Theres one site,
one brand and one code base serving multiple UIs. Implementing
RWD is less easy, as it changes the Web development experience.
You need a clear and strong vision of how the site will present
information. RWD may be easy to sell, but it doesnt support
improvisation. Content adaptation isnt magic. It comes from
attentive and diligent planning.
6 msdn magazine

RWD also changes the consolidated rules of project management and commonly ends up costing more than developing a plain
non-responsive site. Gantt charts tracking progress make little sense
as any new step requires team members to measure effects across
the full spectrum of supported views. So in the end, when having
one responsive Web site isnt an option, what can you do to meet
your customers expectations and make their life easier?

Add a Connected Mobile Site

When your Web site helps your company stay in business and plays
a central role in attracting, maintaining and serving customers,
then providing a mobile-specific experience is vital to keep those
customers coming back. The simplest approach that doesnt require
a full-site rewrite and wont break any existing features is to create
a separate site specific for your mobile audience. This approach is
referred to as building an m-site.
An m-site, much like a native application, is most likely to be
consumed by users on the movepeople who are standing up, in
a hurry, busy or waiting in line. Users willing to connect to a site
under these conditions and using a tiny device truly expect some
benefit from the site. Therefore, the site must be direct, concise,
accurate, and fast to load and respond.
Taking the m-site approach can be inexpensive for two reasons. Its
a brand-new project, but its expected to inherit core services from
the existing site. More often than not, it inherits large shares of the
business logic, which doesnt mean it duplicates code base. Also, an
m-site usually implements the 80/20 rule. It typically supports no
more than 20 percent of the features available in the full site. The
selection of the 20 percent features, though, isnt a walk in the park.
When planning an m-site, you should reason with the mindset
of a site user on the go. Which pieces of information does a user
on the go really expect to consume on your site? Which functions
Desktop Site
Laptop

Serve
Requested
Page

Mobile

Landing Page
for Mobile Users

Mobile Site
Laptop

Landing Page
for Desktop Users
(If Any)

Mobile

Serve
Requested
Page

Figure 1 The Desktop/Mobile View Switcher Algorithm

Untitled-1 1

12/2/14 4:42 PM

does the user expect to perform? What can you do to smooth and
speed up interaction as much as possible? This latter point involves
using HTML5 features to the fullest.
On a site you know will be used on mobile devices, its arduous to
not set the type attribute of an input element to a number or date when
numbers or dates are expected. This would greatly simplify typing. By
the same token, there may be new use cases introduced on top of similar
business logic to make life easier and more enjoyable to mobile users.

Accessing Your M-Site

Because an m-site is a discrete Web site, which URL should visitors
use? Is a subdomain like m.yourserver.com the best option? Would
a .mobi top-level domain name or /mobile subdirectory under the
desktop work better? An m. subdomain is easier to manage than
Figure 2 Implementation of a Mobile Router HTTP Module
public class MobileRouter : IHttpModule
{
private const String FullSiteModeCookie = "FullSiteMode";
public void Dispose()
{
}
public void Init(HttpApplication context)
{
context.BeginRequest += OnBeginRequest;
}
private static void OnBeginRequest(Object sender, EventArgs e)
{
var app = sender as HttpApplication;
if (app == null)
throw new ArgumentNullException("sender");
var isMobileDevice = app.Context.Request.IsMobileDevice();
// Mobile on desktop site, but FULL-SITE flag on the query string
if (isMobileDevice && HasFullSiteFlag(app))
{
app.Response.AppendCookie(new HttpCookie(FullSiteModeCookie));
return;
}
// Mobile on desktop site, but FULL-SITE cookie
if (isMobileDevice && HasFullSiteCookie(app))
return;
// Mobile on desktop site => landing page
if (isMobileDevice)
ToMobileLandingPage(app);
}
private static Boolean HasFullSiteFlag(HttpApplication app)
{
var fullSiteFlag = app.Context.Request.QueryString["m"];
if (!String.IsNullOrEmpty(fullSiteFlag))
return String.Equals(fullSiteFlag, "f");
return false;
}
private static Boolean HasFullSiteCookie(HttpApplication app)
{
var cookie = app.Context.Request.Cookies[FullSiteModeCookie];
return cookie != null;
}
private static void ToMobileLandingPage(HttpApplication app)
{
var landingPage =
ConfigurationManager.AppSettings["MobileLandingPage"];
if (!String.IsNullOrEmpty(landingPage))
app.Context.Response.Redirect(landingPage);
}
}

8 msdn magazine

a .mobi domain simply because its a subdomain and not a new

domain name. A directory is probably more error-prone as it means
the mobile site will be hosted under the same application as the
desktop. This can make the whole solution more brittle.
The simplest approach is to use an m. subdomain with some
redirection logic that automatically sends mobile users to the
m.yourserver.com site. This solution is good for users because
they can use the direct m. subdomain and the canonical URL. For
developers, though, it has the drawback of requiring device detection logic on both the desktop and the mobile site.

Implement Site Redirection

When two sites are in place and work under the same URL, you should
process the host name and decide what to do for every request. If the
host name belongs to the desktop site and the requesting browser is
detected to be a desktop browser, then everything works as expected.
Otherwise, the user should see a landing page, and be informed
theyre trying to access a desktop site with a mobile device. Then
theyll have a chance to save their preference for future similar situations. That preference is stored to a cookie and checked next. If
the request refers to a URL in the mobile site and the user seems to
have a desktop browser, consider showing a different landing page
or just let the request go as usual. Finally, if a request is placed from
a mobile device to the mobile site, it will be served as expected. You
can have your site look into the device capabilities and determine the
most appropriate view. Figure 1 presents a diagram of the algorithm.
In ASP.NET, you can implement the routing algorithm as an
HTTP module and register it on both the desktop and the mobile
site. The module captures the BeginRequest event and uses plain
redirection orif possiblerewriting the URL to change the target
page as appropriate, as shown in Figure 2.
Once installed in the desktop site, this HTTP module captures
every request and checks the requesting browser. If the browser is
running within a mobile device, the module redirects to the specified landing page. The landing page will be a mobile-optimized
page offering a couple of links to the desktop site and the mobile site.
Where do you place the landing page to let mobile users decide what
to do? Generally, it doesnt matter. However, if you put it on the mobile
site, you can deploy your mobile site with all the required routing logic
without touching the desktop site code base. Changes to the desktop
site are limited to configuring the HTTP module in the web.config file.

What Does Mobile Really Mean?

The world of the mobile Web isnt simply split into two camps of desktop computers and all the rest. Mobile encompasses smartphones
and tablets at the very least. An m-site that targets smartphones
may not work as effectively for tablets and vice versa. Serving
smartphone-optimized pages to tablets goes against the common
users expectation of getting a desktop-like experience.
Should you have an additional t-site for tablets and a site for
legacy devices and perhaps one for wearable devices? All in all,
smartphones deserve an ad hoc experience and a dedicated site.
Tablets may blissfully render the same content as desktop computers with minor adjustments, such as a different set of CSS files
with more padding and larger fonts and some media query support
Cutting Edge

for when the orientation switches to portrait or landscape. These

adjustments require changes to the existing desktop site code base.
However, its not too much of an intrusive change.
The problem remains of how to detect smartphones and whats a
definition of a smartphone that works for your business scenario.

Device Detection Tools

Theres no common definition for a smartphone. A smartphone
isnt a physical attribute like an OS version number. When browsers
request a page, they send the user agent string, meaning some
server-side code must process the user agent string and determine
whether the device is a tablet, a smartphone or something else.
Device detection requires ad hoc tools, which have varying
levels of reliability and accuracy. For example, you can use some
Modernizr plug-ins or common routines to check user agents for
known substrings. This may work reliably enough if all you want
to do is separate desktops from everything else.
In this context, detecting anything like an Android OS tells the
browser its not a desktop. However, detecting Android doesnt
help determine if its a smartphone or tablet. The WURFL framework (wurfl.sourceforge.net) is a popular choice with tools that work
on-premises, in the cloud and even the Web server level (no code,
just configuration) on IIS and other environments.
If you opt for building an m-site to serve ad hoc content to
mobile users, you need to be clear which devices fall under the
jurisdiction of the m-site (choosing among smartphones, tablets
and more) and how you detect devices so users have a single URL
to reach the site regardless of the device.

MSDN
Magazine
Online

Wrapping Up
Users expect to seamlessly use Web pages on their smartphones
and tablets just as they do on their desktops. Processing power or
cache size is limited on mobile devices, but that cant be an excuse
to serve pages that require zooming and take seconds to load. Separating desktop from anything else isnt an insightful choice, either.
Some level of device detection is required to provide a good service.
When it comes to device detection and form factors, complexity
and development costs will grow.
Many companies think RWD is the classic Columbus Egg
an obviously simple and brilliant solution to a thorny problem.
Responsive Web sites are more complex and expensive to create
than commonly reckoned. More important, an RWD approach
works for new or significantly reworked sites. You cant just easily
add RWD to mobilize an existing site that was devised for the desktop. Nevertheless, addressing the needs of your mobile audience is
a business necessity. And the sooner you address it, the better. Q
DINO ESPOSITO is the coauthor of Microsoft .NET: Architecting Applications
for the Enterprise (Microsoft Press, 2014) and Programming ASP.NET
MVC 5 (Microsoft Press, 2014). A technical evangelist for the Microsoft
.NET Framework and Android platforms at JetBrains, Esposito frequently
speaks at industry events worldwide and shares his vision of software at
software2cents.wordpress.com and on Twitter at twitter.com/despos.

THANKS to the following technical expert for reviewing this article:

Jon Arne Saeteras
msdnmagazine.com

Its like MSDN Magazine

only better. In addition to all the
great articles from the print
edition, you get:

Code Downloads
The MSDN Magazine Blog
Digital Magazine Downloads
Searchable Content
All of this and more at
msdn.microsoft.com/magazine

magazine

DATA POINTS

JULIE LERMAN

Looking Ahead to Entity Framework 7

Development of the next version of Entity Framework is well
underway. I got my first glimpse of what the EF team was working
on at TechEd North America 2014, when Program Manager Rowan
Miller talked about the goals for Entity Framework 7 (EF7) and
demonstrated some very early bits.
That was five months ago as Im writing this column, and although
EF7 is still an early alpha, it has already come a long way. In this column I want to make you aware of what EF7 will bring to developers,
the motivations behind decisions being made about EF7, and what
this version means to existing apps that use EF6 or earlier. Ill also
give you a peek at what some of the code will look like.

Open Source, but Now on GitHub

The first thing to know about EF7 is that, like EF6, its open source.
But rather than being developed on CodePlex, EF7 is on GitHub,
along with the rest of the upcoming version of ASP.NET. The URL
for EF7 development is github.com/aspnet/EntityFramework. As with EF6,
youll be able to see the details of EF7 as it evolves. You can explore
the source, as well as its progress through branches and commits,
follow the discussions, raise issues, fork the source, and submit
pull requests for the team to examine and potentially commit to
the code base.

EF6 Is Not Going Away Any Time Soon

Dont worryyou wont be forced to move to EF7. Think back to
ADO.NET DataSets and DataReaders. Much like ASP.NET Web
Forms, which are still supported and even benefiting from occasional
tweaks, ADO.NET is still part of the Microsoft .NET Framework,
even though EF has been the primary data access technology for
.NET for many years. Not much has happened to enhance those
technologies, but theyre still there, and still supporting a lot of legacy
code (mine included). One of the big advantages EF6 has over those
other technologies is that its open source, so even though the team
at Microsoft wont be making big investments in EF6, the community will still be able to. And the EF team is committed to EF6. They
will continue to make tweaks, closely inspect pull requests and
update EF6. Even though theyve been going full-bore on EF7 for
a good part of 2014, EF6 was updated. Version 6.1.0 was released in
February 2014; 6.1.1 in June 2014; and as Im writing this article, 6.1.2
This is based on an alpha version of Entity Framework 7. All information is
subject to change.
10 msdn magazine

is in beta, to be released soon. I was originally concerned, but am

no longer worried, about being able to keep older apps going. The
only ones I worry about are the earliest apps that used EF with the
.NET Framework 3.5, ObjectContext and more. But if you havent
updated those apps to leverage all of the great improvements to
EF over the years, you may not be worrying too much about EF7,
anyway. You can also find all of the earlier packages for EF on
NuGet, all the way back to EF 4.1.10311.

EF7: The Short List

Heres the high-level view of whats exciting in EF7:
Support for non-relational data stores and even in-memory
data for testing.
Support for machines and devices that dont use the full
.NET Framework. This means you can use EF7 in Windows
Phone and Windows Store apps, as well as on Linux and
Macintosh machines that are running Mono.
Support for many features that developers have requested
but couldnt be achieved with the existing code base.
Continued support for applications that use the full .NET
Framework such as Windows Presentation Foundation
and other client applications.
EF7 will be distributed in the same manner as ASP.NET 5
and can be used with ASP.NET 5 apps.

Familiar Coding Surface, New Code Base

Each version of EF has evolved the framework, adding new capabilities
and fine-tuning performance and APIs. As Ive written about before
in this column, as well as in an overview article in the December
2013 issue, Entity Framework 6: The Ninja Edition (bit.ly/1cwjyCD),
the latest version brought EF to a new level, giving the framework
many of the features users have been asking for along the way, such
as asynchronous database execution, tapping into the query pipeline, customizing Code First conventions and so much more. I drill
much further into these features in my Pluralsight course, Entity
Framework 6, Ninja Edition: Whats New in EF6 (bit.ly/PS-EF6).
There were even more features developers sought for EF that
Microsoft was eager to implement, but the 10-plus-year-old code
base EF is built onwith a continued dependency on ObjectContext
and less flexible coding patternsprevented the team from getting
to this next level of capabilities. A difficult decisionwhich surely
many of you have faced with your own legacy softwarewas made
to rebuild Entity Framework from scratch.

Untitled-1 1

10/30/14 4:09 PM

EF7 isnt creating a new framework for data access. Instead, its
building a new, more sustainable base on which to support not
only the features and workflow youve depended on for years with
EF, but also one that will enable so much more. The team struggled with whether this should be the next EF or a new data access
technology. At one point, I even questioned whether it would be
EF Light. But the core functionality of EF is still there and after
a lot of consideration, I agree it makes sense to think of this as the
next version of Entity Framework. You can read more about this
in the teams blog post, EF7 v1 or v7? (bit.ly/1EFEdRH).

Shedding Some Legacy Weight,

Keeping the Good Parts
Yet theres also news about EF7 thats worrisome for some developers. While the most common EF classes, patterns and workflows will
remain intact, some of the lesser-used members will be left behind.
But please dont panic; Ill talk more about this in a bit.
Enabling developers to continue using familiar patterns and
even be able to port good amounts of existing code to EF7 was
a critical goal. Youll still use DbContext, DbSet, LINQ queries,
SaveChanges and many of the means of interaction that have been
part of EF for a long time.
Heres a DbContext class I defined in EF7:
public class BreweryContext : DbContext {
public DbSet<Brewery> Breweries { get; set; }
public DbSet<Beer> Beers { get; set; }
}

And heres a simple update in EF7 thats the same as in EF6. Im

using a synchronous save, but all of the async methods are there,
as well:
public void StoreBeers(List<Beer> beers) {
using (var context = new BreweryContext()) {
context.Beers.AddRange(beers);
context.SaveChanges();
}
}

And a simple query:

using (var context = new BreweryContext()) {
return context.Breweries.Where(b=>b.Location.Contains("Vermont"));
}

Im using the version of EF7 thats in the packages with version

beta2-11616. EF7 isnt really beta at this time, but the beta2 is
related to a NuGet package-naming decision. By the time this
article is published, EF7 will have evolved further, so please consider this a look, not a promise.
I still have a DbContext and define DbSets, just as Ive always
done. OnModelCreating is still there, although Im not using it here.
EF4.1 introduced the DbContext API, which was much more
focused on typical EF usage. Underneath, it still relied on the
original ObjectContext that provides database interaction,
manages transactions and tracks the state of objects. Since then,
DbContext has become the default class to use and youd dip down
to the lower-level APIs if you wanted to do a rare interaction with
the ObjectContext. EF7 will shed the obese ObjectContext; only
the DbContext will remain. But some of those tasks youve relied
on the ObjectContext for will still be accessible.
Some of the very complicated mappings that are difficult to support
and not commonly used will go away with EF7. The aforementioned
12 msdn magazine

blog post says, For example, you could have an inheritance hierarchy that combined TPH, TPT and TPC mappings, as well as
Entity Splitting, all in the same hierarchy. If youve ever attempted
to work directly with the MetadataWorkspace API and run away
screaming, you know its an intricate and complex beast, useful for
being able to support this kind of flexibility. But that complexity has
prevented the team from being able to support other scenarios for
which users have asked. By simplifying the mapping possibilities,
the MetadataWorkspace API also became simpler and much more
flexible. You can easily get to metadata about your model schema
from the DbContext API in EF7, which gives you a low-level
capability to perform advanced techniques without having to have
the low-level ObjectContext at your disposal.

Dropping EDMX, but Database First Will Continue

Entity Framework currently has two ways to describe a model.
One uses an EDMX in the designer; the other involves the classes,
a DbContext and mappings that are used by the Code First APIs.
If youre using the EDMX and designer, at run time EF creates an
in-memory model from the XML behind the EDMX. If you choose
the Code First path, EF creates the same in-memory model by reading
the classes, DbContext and mappings you provided. From that
point on, EF works the same, regardless of how you describe your
model. Note that with the EDMX/Designer workflow, you also get
POCO classes and a DbContext to work with in your code. But
because the EDMX is there, they arent used to create that in-memory
model. This is important to understand as you read the next sentences: EF7 will not support the designer-based EDMX model. It
will not have the ability to read the EDMX XML at run time to create
the in-memory model. It will use only the Code First workflow.

I still have a DbContext

and define DbSets, just as Ive
always done.
When the team blogged about this, it caused panic among developers. Partly this was due to the fact that many still dont realize
you can reverse-engineer a database to POCO classes, DbContext
and mappings. In other words, you can start with a database to
get a Code First model. This has been possible since the EF Power Tools Beta was first released in early 2011. Its supported by the
EF6.1 designer and it will definitely be supported for EF7. Ive said
many times that the Code First moniker is a little confusing and
misleading. It was originally called Code Only, but the name was
changed to Code First to make a nice match with Database First
and Model First.
So you dont need the designer or an EDMX to start with an
existing database.
But what if you have existing EDMX models and dont want to
lose the ability to use a designer? There are third-party designers
that support Entity Framework, such as the LLBLGen Pro Designer,
Data Points

which already supports EF Code First (bit.ly/11OLlN2), and the Devart

Entity Developer (bit.ly/1yHWbB2). Look for those tools and possibly
others to potentially provide designer support for EF7.
There is yet another path to keep in mind: sticking with EF6!

Smaller Footprint, More Devices and OSes

Additionally, Microsoft strove to streamline the distribution of
the EF APIs. The NuGet package folder for EF6.1.1 is about 22MB.
This includes a 5.5MB assembly for the .NET Framework 4.5 and
another in case youre using .NET Framework 4. With EF7, there
are a number of smaller DLLs. Youll combine only the DLLs necessary to support your workflow. For example, if youre targeting
SQL Server, youd use a core EntityFramework.dll, a DLL for SQL
Server and another with APIs common to relational data stores.
If you want to use migrations, thats a separate assembly you can
skip. Otherwise, you may want to create and execute migrations
from the Package Manager Console. Theres an API for commands.
Using the NuGet package manager, the proper packages will be
identified and downloaded via their dependencies, so you wont
have to worry too much about the details.

When Entity Framework

was first introduced, Microsoft
had a vision of it being used for a
variety of data stores, though
the first pass focused on
relational databases.
What this does is minimize the EF7 footprint on the end users
computer or device, which is especially important on devices.
ASP.NET is going this route, as well. Both of these technologies
are dropping their reliance on the full .NET Framework. Instead,
theyll distribute only the DLLs necessary for accomplishing the
tasks of a given application. This means the already-streamlined
version of .NET used by Windows Phone and Windows Store apps
will be able to use EF7.
It also means that OSes like OS X and Linux that use Mono
rather than the full .NET Framework will also be able to support
client-side Entity Framework.

Beyond Relational
When Entity Framework was first introduced, Microsoft had a
vision of it being used for a variety of data stores, though the first pass
focused on relational databases. Non-relational databases existed at
that time, but were not widely used, unlike the NoSQL databases
especially document databasesthat are so popular today.
While EF is an Object Relational Mapper (ORM), developers who
use it want to be able to use the same constructs to interact with
non-relational databases. EF7 will provide a high level of support
msdnmagazine.com

for this, but keep in mind what high level really means. There are
vast differences between relational databases and non-relational
databases and EF will not make any attempt to mask those differences. But for basic querying and updates, youll be able to use the
patterns with which youre already familiar.
Figure 1 shows code from a sample app that targets Microsoft
Azure Table Storage, which is a non-relational document database. The sample comes from EF Program Manager Rowan Miller
at github.com/rowanmiller/Demo-EF7. Note that the sample runs against
the 11514 version of the EF7 alpha nightly builds.
The OnConfiguring method is new. Its a way to affect how
EF configures the DbContext at run time, somewhat like you
can do today with the DbConfiguration class. Notice the
builder.UseAzureTableStorage extension method, which exists
because Ive also installed the EntityFramework.AzureTableStorage
package into my project.
EF7 uses this pattern for its various providers. Heres an OnConfiguring method in a DbContext class within a project that
targets SQLite:
protected override void OnConfiguring(DbContextOptions builder) {
string dir = ApplicationData.Current.LocalFolder.Path;
string connection = "Filename=" + Path.Combine(dir, "VermontBrewery.db");
builder.UseSQLite(connection);
}

This project has the EntityFramework.SQLite package installed,

so now I have the UseSQLite extension method instead.
Back in the WarrantyContext class in Figure 1, you can see the
familiar OnModelCreating override for DbContext and in there
Im doing some special mapping. Again, I have methods provided
by the EntityFramework.AzureTableStorage NuGet package. I
get to pick and choose the packages I want based on the features
I need. Azure Table Storage relies on a key-value pair for unique
identity and to support table partitioning. In order to retrieve or
store data, its critical to know what values are to be used for the
PartitionKey and the RowKey, so the API provides a method
PartitionAndRowKeythat allows you to map the properties to
the appropriate keys. The concept is no different from how youve
been able to use the fluent API or Data Annotations to specify the
property that maps to a relational databases primary key.
Thanks to this mapping, I can write a familiar LINQ query to
retrieve some data:
var warranty = _context.Warranties
.Where(w =>
w.BikeModelNo == modelNo
&& w.BikeSerialNo == serialNo)
.SingleOrDefault();

So youre seeing a typical LINQ query, but its being executed

against the Azure Table Storage data store, just as you can do with
a relational database today.
This same demo also updates warranty objects; creates and inserts
new ones using DbSet.Add; and uses DbContext.SaveChanges to
persist everything back to the data store, just as its done today with
EF6and has been done throughout the history of EF.
Also interesting to consider is how Entity Framework has always
supported a set of canonical features for mapping to relational
databases, but left it up to the database providers to specify how those
would translate to the database they target. EF7 will have a high-level
set of canonical features that can be understood by relational and
January 2015 13

non-relational data stores. Theres also a lower-level set of features that

focus on relational databases, and theyre encapsulated in the EntityFramework.Relational assembly. All of the relational database providers
will depend on those and, just like today, their specific handling of
database interaction will be housed in their own provider APIs, like
the EntityFramework.SQLite I used earlier. Youll find extension
methods in the providers that spin off of an AsRelational method,
which is in the Relational API. Its an extension method of DbContext.
Theres even an in-memory data store provider, which is for unit
testing when you want to avoid a database interaction that might
be involved in the logic youre testing. Typically in these scenarios you use fakes or mocking frameworks to impersonate the
database interaction.
If you were setting up a test to perform a query or update the database, youd have some code to instantiate the database, such as:
using (var context = new BreweryContext()) {
// Perform some action against the context
}

You can easily switch to an in-memory store by first installing

the entityframework.InMemory package to your test project, defining a DbContextOption for InMemoryStore and then specifying that the context should use that option. Again, this is possible
thanks to extension methods provided by this API:
var options = new DbContextOptions().UseInMemoryStore();
using (var context = new BreweryContext(options)){
// Perform some action against the context
}

More Features, More Capabilities,

Much More Flexible
You can already see the benefits of the new code base in the flexibility the extension methods provide, and in the ability to affect
the Entity Framework pipeline with the OnConfiguring overload.
There are extensibility points throughout the new code base, not
only for changing EF7, but also for making it simpler for you to
plug in your own logic to EF7.
The new core code base gives the EF team a chance to solve
some age-old problems. For example, the version Im using already
has support for batch updating, which is the default for relational
databases. Ive played with code that allows me to use my own
methods inline in LINQ queries without receiving the dreaded
Entity Framework cannot translate this method into SQL. Instead,
Figure 1 A DbContext Defined for
Working with Azure Table Storage
public class WarrantyContext : DbContext
{
public DbSet<WarrantyInfo> Warranties { get; set; }
protected override void OnConfiguring(DbContextOptions options) {
var connection =
ConfigurationManager.ConnectionStrings["WarrantyConnection"]
.ConnectionString;
options.UseAzureTableStorage(connection);
}
protected override void OnModelCreating(ModelBuilder builder) {
builder.Entity<WarrantyInfo>()
.ForAzureTableStorage()
.PartitionAndRowKey(w => w.BikeModelNo, w => w.BikeSerialNo);
}
}

14 msdn magazine

EF and the providers are able to parse out which part of the query
will become SQL and which will get run locally on the client. Im
sure there will be protection from and guidance for avoiding some
potential performance issues for that particular feature.
The team was able to add the long-requested Unique Foreign
Keys capability for models. Theyre also looking closely at providing
support for table-valued functions and cleaner ways to handle disconnected data, which is something Ive focused on for many years
with Entity Framework. Its a common problem with disconnected
applicationsnot just when Entity Framework is involvedand
its not easy to create algorithms that will work consistently in
every scenario. So a new approach is needed, for sure.
Theres a lot more to get excited about with EF7. I highly recommend
a close look at the posts on the ADO.NET Team Blog at blogs.msdn.com/
adonet. In addition to the post I linked to earlier, Rowan Miller wrote
in-depth about the decision to drop designer support in EF7; see
EF7 - What Does Code First Only Really Mean at bit.ly/1sLM3Ur.
Keep an eye on that blog, as well as the GitHub project. The wiki on
GitHub (bit.ly/1viwqXu) has links to how to access the nightly builds;
how to download, compile and debug the source code; some walkthroughs and the design meeting notes. The team is eager for your
feedback and is excited to receive pull requests.

A Decision Not Taken Lightly

Its important to me to write about EF7 to help allay some fears
about such a big change and that some of the existing EF features
that might be integral to your applications will not make it into
EF7. These fears are not unfounded and the team is not taking them
lightly, nor am I. But understanding that EF6 will not go away and
will continue to evolve with contributions from the community is
critical. If you want to take advantage of the forward movement,
youll have some tough choices to make. Upgrading big applications will not be easy and you should weigh the options carefully.
Perhaps you can break up your application, rewriting only some
portions to benefit from EF7.
Again, as Im writing this column, EF7 is still in its early stages, and
Im not sure how far along it will be by the time you read this. But the
current available source and NuGet packages are there to explore,
experiment with and provide feedback. Bear in mind that the team
may not always keep all of the provider APIs (such as Redis, SQLite
and others) up-to-date as they evolve the core API. According to the
post at bit.ly/1ykagF0, EF7 - Priorities, Focus and Initial Release, the first
release of EF7 will focus on compatibility with ASP.NET 5. Subsequent
releases will add more features. Still, even though EF7 is not yet stable
enough to begin building applications with, theres definitely enough
there to let you start planning ahead.
Q
J ULIE L ERMAN is a Microsoft MVP, .NET mentor and consultant who lives
in the hills of Vermont. You can find her presenting on data access and
other.NET topics at user groups and conferences around the world. She blogs at
thedatafarm.com/blog and is the author of Programming Entity Framework
(2010), as well as a Code First edition (2011) and a DbContext edition (2012), all
from OReilly Media. Follow her on Twitter at twitter.com/julielerman and see
her Pluralsight courses at juliel.me/PS-Videos.

THANKS to the following Microsoft technical expert for reviewing this article:
Rowan Miller
Data Points

Untitled-1 1

10/30/14 4:10 PM

VISUAL BASIC

14 Top Improvements in
Visual Basic 14
Lucian Wischik
Visual Basic 14 is the newest version of Visual Basic,
and it will ship as part of Visual Studio 2015. This version has been
rewritten from scratch in about 1.3 million lines of VB codeearlier
versions were actually written in C++. The team took advantage of
the rewrite to rethink every part of VB from the ground up. I asked
the team to pick out their top 14 improvements. They chose their
favorites from across the boardin the coding experience, in the
project system fundamentals and in the language itself.
Better Coding Experience

1. Refactorings Just the fact that we finally get refactoring built

directly into the product is huge..NET MVP Jim Wooley
It used to be you had to buy add-on products just to get essential
refactorings for Visual Basic, like extracting a method or inlining
a temporary variable. There were a few refactorings for C#, and
Microsoft had partnered with Developer Express to make its
Refactor! add-in available to Visual Basic users since Visual Studio
2005. Now refactorings are built into Visual Studio 2015! To use
them, click on an identifier, or highlight a sub-expression. Then hit
Ctrl+Dot, or right-click and choose Quick Actions. This brings up
a light bulb context menu of relevant actions, as shown in Figure 1.
Note that the refactorings are context-aware. For instance, if you
extract the right-hand side of Dim circleArea = Math.PI * radius
* radius out into a method, Visual Basic suggests the name GetCircleArea for that method. And it puts you into inline-rename
mode if you want to change the name further. Whats smart about
this inline-rename is that it can detect and warn you about name
This article refers to prerelease versions of Visual Basic 14 and
Visual Studio 2015. All information is subject to change.

This article discusses:

A better coding experience
Enhanced project system fundamentals
Language improvements

Technologies discussed:
Visual Basic 14, Visual Studio 2015
16 msdn magazine

clashes if you pick a name thats already in use, it can avoid clashes
where possible, and it works across your entire solution, even
changing names in C# projects, as well.
2. Analyzers The feature is spectacular. I have so many ideas
for how to use this feature ... all the little things I see in peoples
code.Overheard from a Windows PowerShell MVP
Analyzers are a way to put those light bulbs, code actions and error
squiggles into your own hands. You can use them to enforce coding
guidelines throughout your team. If youre flown in to debug a problem, you can plug in an analyzer to quickly find common code defects
in the entire solution. And many of the libraries you use can become
code-aware with their own built-in analyzers. For instance, suppose
you havent yet used the Microsoft Azure Storage library or havent
read articles on best practices. Because the library now comes with
an analyzer that detects common pitfalls in use of its APIs, you can
be immediately confident that youre using it properly. Its like having
an expert code reviewer stand over your shoulder as you type.
You can add analyzers to your project under the new References
| Analyzers node, (or via NuGet). Once there they become part
of your projects compilationthey run live as you type to show
live error squiggles. They run when you build your project in
Visual Studio or on the command line, and will even run on build
servers. Analyzers get a chance to crack open the internals of the
compiler, to look at the syntax trees of the projects source code
and its types and members. Developers are pleasantly surprised
to discover how easy it is to code their expert domain knowledge
into analyzers, thanks to these syntax trees and types+members.
My favorite analyzer is one that detects where my team has been
using Async Sub methods that should have been Async Function
As Task and issues a warning. This is a rough corner of asynchronous programming that not enough people are aware of, and
it leads to difficult-to-catch concurrency bugs, so its great that
my team can now catch the error at compile time. To get started
writing your own analyzers, go to roslyn.codeplex.com.
3. No Need to Cursor off the Line We do the right thing
now.Dustin Campbell, Visual Basic Team Member
As a VB user, youve long been used to typing some code, then
doing a quick down-then-up cursor to see if any error squiggles

appear. Or youd write code to fix an error squiggle, but then have
to do the down-then-up for the squiggle to disappear.
Now you dont have to go through all of that. Just leave the cursor
where it is, and error squiggles will appear or disappear themselves.
4. References in XML Doc Comments For people passionate
about docs, this is an enormous step in the right direction..NET
MVP Sam Harwell
Are you passionate about XML doc comments? Heres a small example:
'''
'''
'''
'''
'''
Sub
End

<summary>
Similar to <see cref="List(Of Integer).Count"/>
</summary>
<param name="e">Thing to count</param>
<remarks></remarks>
Count(e As IEnumerable)
Sub

In previous versions of VB, when you typed out cref and paramname in your comments, you got completion-list help, but beyond
that you were on your own. The compiler did some minimal validation to check that the names existed, but they were typeset in grey
and not easy to find, look up or refactor.
Now in Visual Basic 14 the cref and param-name arguments are
colorized properly. You can hover over them to see tooltips. When
you do a rename-symbol refactoring (Ctrl+R, Ctrl+R), Visual Basic
renames all references to a symbol, including those in cref and paramname. You can right-click on one of them and Go to Definition, or
Find All References. If you want to refer to a method that has several
overloads, you can now unambiguously refer to the single overload
you want. These changes all make it easier to type references in your
XML doc-commentsand get them right.

Project System Fundamentals

5. References Node in Solution Explorer Fact is, we all have to

tweak references daily.Lucian Wischik, Visual Basic Team Member

Figure 2 shows how a typical Visual Basic 14 project looks in
Solution Explorer.
Whats new is the References node. This used to be hidden and
you had to click Show All Files to see itbut that also showed lots
of irrelevant files.
This previous behavior might have made sense 10 years ago when
youd start with a Windows Forms project and it would generally have
the right set of references. But its a reality of modern development
nowadays that the References node is used frequentlyespecially
to manage NuGet references. Its a small but handy convenience to
be able to find it easily in Solution Explorer.
6. Shared Projects This is really just nice tooling on top of
linked files that makes them easier to work with.Windows
Developer MVP Morten Nielsen

Figure 1 Visual Basic 14 Now Has Built-in Refactoring

msdnmagazine.com

Suppose you want to share code between two or more projects.

Its a common-enough situation, for example when maintaining
both Windows Presentation Foundation (WPF) and Windows
Phone versions of an app. The goal is always the same: maximize
code reuse, so, for example, a bug fix you make for one project will
automatically benefit the other project.
In the past, you could choose between two techniques: use linked
files to share common source code, or rearchitect your shared code into
a Portable Class Library to share a common binary. Now Visual Basic
14 allows a third, powerful technique: Shared Projects.
Why would you use Shared Projects? The task of sharing code
is deep and challenging with no good one-size-fitsall solution.
Portable Class Libraries are a good, clean solution, but they force
you to architect your code so the common code never calls into
the WPF or Phone projects; it only calls into system APIs that are
present on both WPF and Phone. Shared Projects are easier to use
because they dont require this rearchitecting.
To set up a shared project, right-click on your solution and
select Add | New Project | VB | Shared Project. Next, right-click
on each projects Reference node in turn and choose Add | Shared
Projects. A Shared Project is just a collection of the source files,
XAML files, images and other assets that will be included in each
project that references it.
For each of your projects you can also set up My Project | Compile |
Advanced Compile Options | Custom Constants with custom
constantsWPF and PHONE, for example. Then, in your shared
code, you can call into project-specific APIs like this:
#If WPF Then
' nothing needed
#ElseIf PHONE Then
ShowBatteryStatus()
#End If

7. 50 Percent Faster Compile Times 50 percent is no joke.

.NET MVP Sam Harwell
The Visual Basic compiler used to be written in C++. For Visual Basic 14, the team has rewritten it completely in VB, which has
made it considerably faster! Here are two comparisons:
A large solution build (1.3 million lines of code) went from
68 seconds to 41 seconds.
A cold solution load (a Windows Store app) went from 6.7
seconds to 4.6 seconds
Thats a considerable time saving. Its great when your mind can
stay on track rather than meandering in the delay between finish
coding and press F5 to hit a breakpoint.
The 50 percent performance increase might come as a surprise
to people who think that C++ is faster than VB. The truth is

Figure 2 The References

Node Is Now Shown in
Solution Explorer
January 2015 17

Figure 3 Lambdas and LINQ Expressions in the Watch Window

that algorithms, data structures and concurrency are where you

get real speed benefits. The performance boost from rewriting
in VB comes from many places: rethinking the data structures;
being able to express algorithms more cleanly and refactor more
safely; using async and the threadpool; using the Visual Studio
profiling tools to discover CPU and memory-allocation hotspots;
and using analyzers to detect simple .NET performance gotchas
like unnecessary boxing.
8. Lambdas and LINQ Expressions in a Watch Window Awesome!Visual Studio Uservoice User Marco Senn-Haag
LINQ and lambdas are a great way to summarize data. One of
the places this is needed most is at debug time, in the Watch and
Immediate windows. In the past, any use of LINQ or lambdas here
generated an error:
Evaluation of lambda expressions is not valid in the debugger.

Now, as you can see in Figure 3, it just works! For instance, if

youre at a breakpoint where you have a collection called customers,
you can get a quick read on the collection by writing this in the
watch window:
From c In customers Where c.OrderStatus = "Unfulfilled" Select c.LastName

names so you can more easily read the

error message. And it shows the error
code, as well, which is handy for sorting
by error code. Even better, the error
code is a hyperlink to an Internet search,
often more useful than a link to the
MSDN documentation page. Also, you
can filter each column in the error list
like you do in Excel.
Sometimes when making big changes
its easy to get your solution into a state where a lot of downstream code is broken. In the past, VB would show only the
first 101 errors. This made it difficult to get an idea of just
how widespread the break was, or to get an overview of what
kinds of changes youd have to make. Now, Visual Basic 14
shows all errors without limit. (This can be configured under
Tools | Options | Text Editor | Basic | Advanced | Show diagnostics
for closed files.)

Language Improvements

10. Null-Propagating Operators The constant need for null

checking bites everyone both in productivity and in bugs..NET
MVP Deborah Kurata
Suppose you have a Customer class with an Address field that
might legitimately be null, perhaps because your app doesnt
require addresses to be typed in. Previously, any code that needed
to do something with the addresslike display it to the user
would need to be safeguarded with null checking to take this into
account. Such null checks quickly get tedious. With Visual Basic
14 you can elegantly handle the possibility of a null like this, using
the new ?. operator:

Did you know you can use the Immediate Window without
even launching your program? For instance, if youve just written
Console.WriteLine("{0} ({1})",
customer.Name,
a module with a function GetName, you can open the Immediate
customer.Address?.Country)
window (Debug | Windows | Immediate) and type ? GetName() and it will
evaluate.
Visual Studio 2015 will also have better
support for Edit and Continue such as in
Async and Iterator methods, as well as in
more common situations like inside of
and around LINQ queries and lambdas,
even allowing you to add a new query or
lambda expression to an existing method.
Although this didnt make it into the
Visual Studio 2015 Preview, youll be
able to do all of this in the final release.
9. Better Error List The before and
after really tell a great story.Visual Basic
Team Member Anthony D. Green
The Error List in Visual Basic 14 has
numerous practical improvements, ones
that answer long-standing user requests
(see Figure 4). Before, the Error List used
to show fully qualified type names; now Figure 4 The Error List in Visual Studio 2015 (bottom) is more Readable and Versatile
it shows only minimally qualified type Than in Visual Studio 2013 (top)
18 msdn magazine

Visual Basic

Untitled-1 1

12/8/14 11:11 AM

The ?. operator is simply shorthand for the common but cumbersome

pattern of assigning to a temporary variable and then checking for null:
Dim _temp = customer.Address
Console.WriteLine("{0} ({1})",
customer.Name,
If(_temp Is Nothing, Nothing, _temp.Country))

You can use ?. in a sequence and mix it with the regular dot operator, a?.b.c?.d, for example. It reads left-to-right. Any null value used by
?. will just stop the sequence short and give the answer Nothing, and
any null value used by . will throw a NullReferenceException as usual.
The ?. operator is a null-conditional version of the . operator. There
are null-conditional versions of most other operators, as well: indexing, array?(i); delegate invocation, delegate?(args); dictionary lookup,
dict?!key; and XML axis properties, xml?.@attr, xml?.<key>, xml?...<key>.
You can use ?. in other handy ways, too:
If customer?.Age > 50 Then ...
' If branch taken only if customer is non-null AND is older than 50
Dim name = If(customer?.Name, "blank")
' Pick a default name if customer is null
Dim first = customers?.FirstOrDefault()
' Only invoke this method if customers is non-null

11. Multiline String Literals Is it sad that Im so excited for

multiline string literals?Forum User Scaramouche
This is how you used to write multiline strings, peppered with vbCrLfs:
Dim json = "{" & vbCrLf &
" 'Name': 'Bad Boys'," & vbCrLf &
" 'ReleaseDate': '1995-4-7T00:00:00'," & vbCrLf &
" 'Genres': ['Action','Comedy']" & vbCrLf &
"}"

Understandably, a common request has been to allow string

literals that span multiple lines. Now, in Visual Basic 14, you can:
Dim json = "{
'Name': 'Bad Boys',
'ReleaseDate': '1995-4-7T00:00:00',
'Genres': ['Action','Comedy']
}"

A related noteworthy featurealso commonly requestedis that

you can now put comments within multiline statements. Previously,
they werent allowed inside LINQ expressions like the following:
Dim q = From x In y ' This is a from clause
Where x < z ' And this is the where
Select x
' This select is redundant

12. String Interpolation String interpolation makes for much

simpler code, and makes the intent pop out.Channel9 User Judah
String interpolation is an easier way of writing strings with
expressions in them, like so:
Dim s = $"hello {p.Name} you are {p.Height:0.00}m tall"

This is simply shorthand for the following:

Dim s = String.Format("hello {0} you are {1:0.00}m tall", p.Name, p.Height)

String interpolation is often easier to write than an explicit call to

String.Format because it saves you having to juggle the positional
placeholders {0} and {1}. And, of course, theres full colorization and
IntelliSense for the expressions inside the holes. String interpolation
works particularly well with programmatic strings, as in these examples:
Dim fn = $"C:\Documents\{folder}\{file}.{ext}"
Dim url = $"http://{site}/{path}/{file}?search={query}"

As is normal for String.Format, this formats the string using

the current culture. If youre constructing a programmatic string
that includes a floating point number, such as when youre passing
latitude and longitude to a Web service, youll most likely want
InvariantCulture instead. This will be supported in Visual Studio
2015, but at press time the design wasnt yet settled.
20 msdn magazine

Note that string interpolation isnt yet in Visual Studio 2015

Preview, but will be present in Visual Studio 2015 before its final.
13. NameOf This will be such a help for so many scenarios I
can think of.Roslyn Forum Member ewwloyd
The NameOf operator is a better way of embedding a string literal
in your code, when that string literal refers to a name in your source
code. Heres one example:
Sub f(s As String)
If s Is Nothing Then Throw New ArgumentNullException(NameOf(s))
End Sub

The NameOf operator isnt evaluated at run time: Its a compile-time

constant, in this case the constant string s. The reason to use NameOf(s)
is that it safeguards you against typos. For instance, if you rename
the method parameter, then the NameOf argument will be renamed
automatically. That wouldnt have happened with just a string literal.
Heres another place where NameOf works well:
Private _age As Integer
Property Age As Integer
Get
Return _age
End Get
Set
_age = Value
RaiseEvent PropertyChanged(
Me, New PropertyChangedEventArgs(NameOf(Age)))
End Set
End Property

Note that NameOf isnt yet in Visual Studio 2015 Preview, but
will be present in Visual Studio 2015 before its final.
14. Open Source Were trying to engage the community. There
are a lot of smart people out there. Well look at pull requests from
the community just like we do our own ideas.C#/Visual Basic
Architect Anders Hejlsberg
The final improvement is not in Visual Basic itself, but in the
process of working with VB.
The source code of the VB compiler is now open source. So is
the design process of the language itself. Each new feature proposal
is made in the open, with full public scrutiny. The members of the
Microsoft Visual Basic Language Design Team are essentially now
stewards of the language. The team looks at proposals, considers
them deeply, sees if there are unexpected gotchas or corner-cases,
and determines if they meet the bar for inclusion in the language.
The Language Design Meeting minutes are published openly. Its
truly an exciting time to be on the Visual Basic Language Design
Team, and an exciting time to be a user of VB.

Wrapping Up
There are a lot of improvements in Visual Basic 14. This article has
covered almost half of them. The general theme has been to make
the existing VB work better in easy-to-use ways, without introducing difficult new concepts. For more information, check out
roslyn.codeplex.com and blogs.msdn.com/vbteam.
Q
LUCIAN WISCHIK is on the Visual Basic/C# Language Design Team at Microsoft,
with particular responsibility for VB. Before joining Microsoft he worked in
academia on concurrency theory and async. Hes a keen sailor and long-distance
swimmer. Reach him at lwischik@microsoft.com.

THANKS to the following Microsoft technical experts for reviewing this article:
Dustin Campbell and Anthony D. Green
Visual Basic

Untitled-1 1

10/13/11 11:25 AM

VISUAL STUDIO 2015

Web-Based Test Case

Management with TFS
Manoj Bableshwar
Application Lifecycle Management with Team Foundation for manual testing. To access the Test Hub, you can navigate to it
Server (TFS) is all about leveraging an integrated toolset to manage
your software projects, from planning and development through
testing and deployment. As a core part of Team Foundation Server,
the Test Hub enables you to create and run manual tests through an
easy-to-use Web-based interface that can be accessed via all major
browsers on any platform. In this article, Ill delve into the phases
of manual testingplanning and creating tests, reviewing them
with stakeholders, running the tests, and tracking the test progress
of the team. Ill touch on different value propositions, such as the
flexibility to customize workflows; end-to-end traceability; criteriabased test selection; change tracking and audit; sharing test steps
and test data; stakeholder review; and most important, ease of use,
especially for testers whove been using Excel-based frameworks
This article discusses:
Planning, creating, running and tracking manual tests
on any platform
End-to-end traceability of test activity across user stories, test
cases and bugs
Reviewing test scenarios with stakeholders, measuring feature
quality and gauging if a feature is ready to ship

Technologies discussed:
Visual Studio Online, Team Foundation Server
22 msdn magazine

by clicking on the Test tab in on-premises TFS, just the way you
access the Work tab to manage backlogs or the Build tab to monitor builds. Alternatively, you can sign up for a free Visual Studio
Online (VSO) account at visualstudio.com and activate the 90-day
account trial to try out Test Hub.

Plan Test Activity for the Sprint

Sprints or iterations are units of planning for teams that practice
Agile or Scrum methodologies. It makes sense to plan test efforts
for a sprint, just as its done for user stories. To get started with test
planning, create a test plan by providing a name and associating it
with a team and a sprint. The test plan can have an owner and test
cycle dates for out-of-band test activity such as a beta release signoff or a user acceptance test cycle. Test plans in TFS are work items,
so you get all the benefits of work items, such as change-tracking
with work item history; permissions based on area paths; rich-text
summary fields; file attachments and more. However, the most
important benefit of work items is customization. Work item customization makes it possible to align the workflows and fields of
artifacts used for tracking activities with the business processes
used by the organization. This concept can be extended to better
reflect the test activities practiced as part of your software development model, by customizing test plan work items. Moreover,
the process of customizing test plan work items is similar to that
of other work items, such as bugs or user stories. For example, the

default states of a test plan can be

changed from Active and Inactive
to, say, Authoring, Testing, or
Archived. Additional user fields
such as reviewers, approvers, signoff owner, and so forth, needed for
accountability or audit requirements, can be added to the test
plan. As you integrate your processes into the test plan, you may
want to restrict access to it, so that
only certain people, such as team
leads or test managers, have access
to create and modify test plans. The
Manage Test Plans permission can
be used to moderate access to test
plans at a user or team level.
Once youve set up a test plan,
youll be eager to create and run
tests. But before that, its import- Figure 1 The Excel-Like Grid Can Be Used to Create Multiple Tests
ant to think about the best way to
organize those tests to enable reuse and end-to-end traceability of story and the test case, thus providing end-to-end visibility of a user
test efforts. Test suites are artifacts that are contained in a test plan story, its test scenarios and open bugs. This helps you measure the
and enable grouping of test cases into logical units. Test suites are of quality and ship-readiness of a feature.
three types: requirement-based test suites (RBS), query-based test
suites (QBS) and static test suites. Static test suites work like folders Criteria-Based Testing with Query-Based Suites
to organize RBS and QBS. If you want to group test cases yourself, Regression-test coverage is as important as test coverage for new
you can manually pick and add test cases to static test suites.
features. Teams typically set up regression-test coverage based on
criteriaall priority 1 tests, all end-to-end scenario tests, all automated tests and so forth. Test Hub supports criteria-based testing
with QBS; these suites are created by defining a query on test cases.
Test cases that match the query criteria are automatically populated
in the QBS, without any need to manually refresh the QBS. QBS
can also be used in other scenarios, such as tracking test cases for
bugs that are being fixed in the current sprint.

Once youve set up

a test plan, youll be eager to
create and run tests.

Like test plans, test suites are work items, so all the customization
benefits mentioned earlier apply to test suites. Some examples of
custom fields for a test suite are summary fields describing instructions to set up the test application and fields to describe the nature
of tests such as functional or integration, test complexity, and so
on. Just as with test plans, you can moderate access to test suites at a
user or team level with the Manage Test Suites permission. Changes
to test cases contained in the suite, owner, state or other fields can
be tracked in the test suite work item history.

End-to-end Traceability with

Requirement-Based Suites
Requirement-based suites correspond to user stories (or product
backlog items for scrum and requirements for CMMI-based projects)
that the team is working on in the current sprint. The goal of creating an RBS by picking a user story is to enable traceability. Test cases
created in an RBS are automatically linked to a user story, making it
easy to find the scenarios covered to test the user story. Bugs, if any,
that are filed while running these test cases are also linked to the user
msdnmagazine.com

Creating Test Cases with an Excel-Like Grid

Test cases are the basic units of testing, each containing test steps
that describe a set of actions to be performed, and expected results
that describe what has to be validated at each test step. Each test
step can have an optional attachment, for example, a screenshot
that illustrates the output. Like test plans and test suites, test cases
are work items, so all benefits of work item customization apply
to test cases, as well.
There are two ways to create test cases. The first option is to use
the test case work item form, which lets you create one test case at
a time. The second option, and the one that really lets you breeze
through creating test cases, is the Excel-like grid shown in Figure 1.
The grid resonates very well with manual testers, who, typically,
wouldve written and tested their test cases in Excel. With the grid,
testers can create multiple test cases at a time, fluently typing test
titles, steps, and expected results while navigating the grid with
tabs, arrows, and the Enter key. Its a simple experience to insert,
delete, cut, copy and paste rows. Whats more, the grid can display
all test case fields, such as state, tags, automation status and so
January 2015 23

on, plus these fields can be bulk-marked for multiple test cases.
If you have an intermittent Internet connection or are just more
comfortable writing test cases in Excel, youre welcome to do that.
Just copy and paste all the test cases youve written in Excel into
the grid and save them to populate them into the system. In fact,
if your team is just adopting the TFS Test Hub for testing, the grid
can help you import your test cases from Excel. Check out the
Test Case Migrator Plus utility at tcmimport.codeplex.com for advanced
import requirements from Excel.

Share Test Steps and Test Data

Some test scenarios need specific test data as input to be meaningfully tested. Also, it makes sense to repeat tests with different
variants of test data, for example, valid and invalid input sets or
different combinations of items in a shopping basket. Parameters
can be used to associate a test case with test data. With mature test
teams that cover large and complex test scenarios, its quite possible
that many test cases use similar test data to drive testing. Shared
parameters can help you consolidate and centrally manage such
test data. You can also import test data from Excel and use it to
drive tests through shared parameters.
Just as with the test data, its possible the test steps are common
across multiple test cases, for example the steps to log into an
application or navigate to a form. Such common test steps can
be consolidated into shared steps. The advantage of using shared
steps is that a change, such as an updated application URL or an
additional authentication step while logging in, can be updated in
the shared step. Changes to shared parameters or shared steps will
reflect across all referenced test cases instantly.

Figure 2 Web-Based Test Runner

24 msdn magazine

Review Tests with Stakeholders

Before running tests, its a good idea to share the tests with stakeholders, such as product managers or business analysts, to solicit
their comments. In cross-division or cross-organization development and test teams, such as outsourced test projects, a formal
signoff may be required before proceeding with test execution. To
share tests with stakeholders for review, you can export a test plan or
a bunch of test suites by e-mail or print them to PDF or hard copy.
The output of the e-mail dialog can be edited before sending it to
stakeholders. You can also copy and paste to Word documents when
stakeholders are required to respond with inline review comments.

Running Tests with the Web-Based Test Runner

To prepare the team to run tests, the test lead can assign tests to team
members. The owner of a test case and the tester of a test case can
be different people; the test lead has the flexibility to shuffle testers
or even take the help of vendors to have tests executed. The most
valuable feature of the Web-based Test Runner, which is used to run
manual tests, is its cross-platform support. Because the Test Runner
is browser-based, you can run it on any platform that supports any
major browserInternet Explorer, Chrome, Firefox and Safari.
The Test Runner presents the test steps and expected results in
a narrow window, making it easy to read and execute the steps on
the application being tested (see Figure 2). Image attachments
that were created while writing the test case are visible and can be
zoomed into. If your test case is driven by test data, then each row
of parameter values included in the test case will correspond to
one iteration of the test.
A test can have different outcomesPassed, Failed, Blocked and
Not Applicable. The Blocked state
can be used when tests are waiting
on an external dependency, such
as a bug fix, and Not Applicable is
useful when a test doesnt apply
to the current featurea service
release, for example. As you walk
through validating the test steps,
you mark them passed or failed.
For failed steps, you can jot down
comments for issues you observed
while testing. You can report the
failure to developers by creating
a bug, right in the context of the
Test Runner session. The bug is
auto-populated with all the steps
performed before you encountered the issue. The bug can also be
updated with additional comments
and screenshots before filing it.
The bug is linked to the test case
that was run while filing it and the
requirement being tested, thus
enabling end-to-end traceability.
On the other hand, if you find
that the discrepancy between the
Visual Studio 2015

Untitled-7 1

12/8/14 4:05 PM

expected results and the application is because the application

was recently updated, you can fix
the test case inline, while its running. If youre in a really long test
session running many tests and
need to take a break, you can pause
the tests and resume them later. If
you find that a test is failing for
you, and wish to find out when it
last passed or which team member
got to execute it successfully, looking at the recent results of the test
case will answer those questions.
While the Test Runner helps you
walk through each test step of a test
case in detail, the bulk-mark feature
helps you pass or fail multiple tests at
once. If youre validating high-level
test scenarios highlighted by the test
case title, but not actually walking
through detailed test steps, you can
quickly mark each tests outcome,
without launching the Test Runner.
The bulk-mark feature is particularly helpful when a large number
of tests have been executed offline
and their status has to be reflected
back in the system.

Figure 3 Tracking Test Results

Track Test Progress with Charts

Is my feature ship-ready? Is my team on track to complete testing
this sprint? Are all the test cases I planned for this sprint ready to
run? These are some of the questions in which test leads, test managers and stakeholders are interested. The Test Hub lets you create
a rich set of charts to help answer such questions (see Figure 3).
Charts come in two sets: test case charts that can be used to track
the progress of test authoring activities, and test result charts that
can be used to track test execution activities. And these charts
can be different kinds of visualizationspie, column, stacked bar,
pivot table and so forth. Test case fields, such as owners, state,
priority and the like can be used as pivots for test case charts.
Test result charts come with the test suite, outcome, tester, run by,
priority and more as pivots. For example, to find the test status of
user stories, you can create a stacked bar chart with test suite and
outcome as pivots for all the requirements-based suites being tested
in the current sprint. These charts can either be created for a
bunch of test suites or for a test plan to roll up information for the
entire test plan. You can also share the insights with stakeholders by
pinning the charts to the homepage. Finally, all the charts display
real-time metrics, without any lag or processing delay.

Wrapping Up
The Test Hub isnt just for manual testers. Its a tool that product
owners and business analysts can use to gauge how their features
26 msdn magazine

measure up against the acceptance criteria. The grid can be used

to keep track of acceptance criteria for requirements, and can later
be used for sign-off. To summarize, the Test Hub offers:
Customization of workflows with test plan, test suite and
test case work items.
End-to-end traceability from requirements to test cases
and bugs with requirement-based test suites.
Criteria-based test selection with query-based test suites.
Excel-like interface with the grid for easy test case creation.
Reusable test steps and test data with shared steps and
shared parameters.
Sharable test plans, test suites and test cases for reviewing
with stakeholders.
Browser-based test execution on any platform.
Real-time charts for tracking test activity.
Test Hub provides an easy yet comprehensive way to test the
user stories you plan to release in a sprint. Test Hub is available
on-premises with TFS, as well as in the cloud with Visual Studio
Online. Get started with a free 90-day trial right away at visualstudio.com.
To see Test Hub in action, watch the demo at aka.ms/WebTCMDemo. Q
MANOJ BABLESHWAR is a program manager with the Visual Studio Online team at
Microsoft. His team ships Manual Testing Tools to Visual Studio Online.

THANKS to the following Microsoft technical expert for reviewing this article:
Ravi Shanker
Visual Studio 2015

Untitled-6 1

5/28/14 4:02 PM

MICROSOFT AZURE

Hadoop Made Easier for

Microsoft Developers
Omid Afnan
This year Microsoft announced the availability of HDInsight,
its Big Data cloud offering. With this service youre able to create
Hadoop, HBase and now Storm clusters within Microsoft Azure. By
using the Azure Management Portal, you can follow some simple
steps to create a Hadoop cluster, associate Azure storage with it and
choose the number of cores you need. With the click of a button,
within minutes you have a cluster provisioned with the necessary computation resources and a 100 percent Apache Hadoop
distribution up and running in the cloud. Thats it: No hardware
procurement, no sorting through Hadoop stacks, no need to learn
how to install Hadoop. Youre ready to run Big Data queries!
You may be saying: Thats great now what? There are a number
of common cases people are trying to address with Big Data. You
or your organization may have some specific cases in mind, like
Internet of Things (IoT) telemetry crunching, social sentiment
analysis or user preference analysis. A case thats interesting to
many of our customers is that of analyzing Web services or application logs. In this article Im going to show you how this scenario
This article discusses:
Creating Hive queries in HDInsight
Developer tooling for Big Data scenarios
Analyzing Web server logs

Technologies discussed:
Microsoft Azure, Visual Studio, HDInsight, Hadoop
28 msdn magazine

can be boot-strapped using HDInsight and how you can use the
newly released Microsoft Azure HDInsight Tools for Visual Studio
to simplify your development experience.

The Log Analysis Scenario

Most apps and services produce a large trail of data intended for
tracking usage, understanding failures or managing the execution
environment. Data streams such as error logs, software quality
management (SQM) output, click-stream instrumentation and
Web server logs can easily accumulate at the rate of gigabytes or
terabytes a week. With the exception of debugging application failures or service outages, these logs are often left untouched due to
their sheer size and the semi-structured nature of their contents.
However, many organizations are beginning to tackle this
problem in order to discover quality issues more quickly, understand customer usage better and, ultimately, drive up customer
engagement and satisfaction. A common base case is the analysis
of Web server logs to derive two kinds of information:
Frequency of particular errors happening on the Web site
Traffic patterns into or out of the Web site
Parsing and aggregating this log data is an activity that allows a
high degree of parallelism. The ideal would be to have subsets of
records farmed out to individual servers that can parse, transform
and summarize those records. Then youd merge these partial
results to create the final aggregated data. This is, of course, exactly
what the MapReduce engine in Hadoop does and the Hive Query
Language (HiveQL) allows you to express your target aggregations

project type enables the easy creation of the

related files needed for specifying a working
application in a given languagein this case
HiveQL. Youll also notice a Hive sample project that contains the code example presented in
this article. Using the Hive project lets you easily
group, save, modify and share related queries.

Getting the Data

Now, in Visual Studio, open Server Explorer
to browse Azure services and objects in your
subscription. Use this to navigate to the Azure
| HDInsight node where you can find your
recently created Hadoop cluster (called WeblogAnalytics in this example). Expand this node
to find two children: Hadoop databases and
the default blob storage you associated with the
account at creation time. The blob storage now
has a default container with the same name as
your cluster, and Hadoop applications can read
and write to this container and its subfolders
Figure 1 Creating a Hadoop Cluster in the Microsoft Azure Management Portal
using the URIs provided.
You can now copy or load your data into this container, and
in a SQL-like query syntax. As your data grows larger with increased
usage, this approach scales effectively to keep overall processing then simply reference the data by supplying the correct URI. If
your Web site is located outside of Azure, use AzCopy to move
time under control.
Programming against Hadoop MapReduce is made significantly the desired logs to this storage container, as needed. If you have
easier with high-level languages like HiveQL. The HDInsight tools existing data in Azure, you can access that data without further
in Visual Studio bring the management of this kind of code into a copying, which is much simpler if youre using Azure Websites to
standard framework familiar to all developers. The tool also exposes host your application. In this case, at cluster creation time, choose
information related to the Hadoop cluster and the data artifacts in the storage account containing your Web site data as the associated
ways that are easy to understand. In the rest of this article Ill walk storage account for your cluster. You have even more flexibility
through a simple log analysis case and highlight how to use the to add multiple storage containers using the HDInsight custom
create option. In these cases you can easily access the required files
language and the tool to get to a working solution quickly.
with a reference such as:

Getting Set Up

Im going to start by creating an HDInsight

cluster for running my queries. The cluster is
needed only when running queries and can be
deleted when not in use, or it can be left running
(in which case you continue to incur charges). Ill
put my persistent data on Azure storage, which
is created separately and can be used with many
different clusters over time. Figure 1 shows where
this is done in the Azure Management Portal.
After the cluster is created it will show up in the
Management Portal, where you can view or delete
it as needed. Note that theres also a Query Console,
which you can open from the Management Portal
to submit simple queries directly from the browser.
At this point you should have already installed
the Azure SDK 2.5, which includes the HDInsight
Tools for Visual Studio. When you navigate to
the File | New | Project screen, youll find a new
category for HDInsight projects. Like other
Visual Studio projects, the Hive Application
msdnmagazine.com

wasb://container@storage.blob.core.windows.net/example/data/filename.ext

Figure 2 Creating a Hive Table Definition

January 2015 29

Figure 3 Inserting Calculated Values

DROP TABLE IF EXISTS ClientErrors;
--create table ClientErrors for storing errors users experienced and
their frequencies
CREATE EXTERNAL TABLE ClientErrors(sc_status int, cs_referer string,
cs_page string, cnt int)
ROW FORMAT DELIMITED FIELDS TERMINATED BY ',';
--populate table ClientErrors with data from table weblogs
INSERT OVERWRITE TABLE ClientErrors
SELECT sc_status,
cs_referer,
concat(cs_uristem,'?', regexp_replace(cs_uriquery,
'X-ARR-LOG-ID=[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]
{12}','')) cs_page,
count(distinct c_ip) as cnt
FROM weblogs
WHERE sc_status >=400 and sc_status < 500
GROUP BY sc_status,
cs_referer,
concat(cs_uristem,'?', regexp_replace(cs_uriquery,
'X-ARR-LOG-ID=[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}',''))
ORDER BY cnt;

For this article Im going to use some log data thats available
as part of the sample code in HDInsight. You can navigate to the
HDISamples folder in the default container where youll find a set
of folders containing samples, including the file:
HDISamples/webloganalyticsstorage/webloganalytics/HdiSamples/
WebsiteLogSampleData/SampleLog/909f2b.log

Once the raw data is available, you can turn it into a Hive table
that can be referenced in other Hive queries as a rowset from which
you can select rows. To do this you create it as an EXTERNAL table,
which is a meta-data operation in Hadoop and doesnt change the
data on disk. The only step you have to take is to define the column
types and name them so you can refer to the data in the familiar
SQL-style fashion. This is an important aspect of the Big Data
modelthat you can do a late binding between the format of your
data and the data itself. Theres no need to create schemas up front,
or worry about cleansing your data at load time. In fact, theres no
checking that the rows in your data adhere to any schema you define.
Instead, its your job to build your queries such that different versions
of data or missing data is handled as your business logic requires.
Again, you can use the Visual Studio IDE to simplify this task.
Figure 2 shows the kind of table creation tool you may have seen
in SQL Server tools. You can simply define table columns by
adding entries to the Table Columns area, providing a name, data
type and additional comments. Beyond a name, nothing else is
needed to create a basic, empty table definition. However, in this
case, I want to apply this definition to an existing file, so I use the
External table property to specify the URI mentioned earlier.
Figure 4 Creating the Table and Inserting Data
DROP TABLE IF EXISTS RefersPerDay;
--create table RefersPerDay for storing references from external Web sites
CREATE EXTERNAL TABLE IF NOT EXISTS RefersPerDay(year int, month int, day int,
cs_referer string, cnt int)
ROW FORMAT DELIMITED FIELDS TERMINATED BY ',';
--populate table RefersPerDay with data from the weblogs table
INSERT OVERWRITE TABLE RefersPerDay
SELECT year(s_date), month(s_date), day(s_date),
cs_referer, count(distinct c_ip) as cnt
FROM weblogs
WHERE sc_status >=200 and sc_status <300
GROUP BY s_date, cs_referer
ORDER BY cnt desc;

30 msdn magazine

Metadata operations are achieved by running DDL statements

in the Hadoop cluster. The create table dialog shows the DDL
generated, which in this case works out to:
CREATE EXTERNAL TABLE IF NOT EXISTS weblogs (s_date date, s_time string,
s_sitename string, cs_method string, cs_uristem string,
cs_uriquery string, s_port int, cs_username string, c_ip string,
cs_useragent string,
cs_cookie string, cs_referer string, cs_host string, sc_status int,
sc_substatus int,
sc_win32status int, sc_bytes int, cs_bytes int, s_timetaken int)
ROW FORMAT DELIMITED FIELDS TERMINATED BY ' '
STORED AS TEXTFILE LOCATION 'wasb://webloganalytics@
webloganalyticsstorage.blob.core.windows.net/HdiSamples/
WebsiteLogSampleData/SampleLog/'
TBLPROPERTIES ('skip.header.line.count'='2');

The DDL includes a couple of extra items. In this example,

the sample file is using spaces to separate column values in each
line, so I specify the field delimiter to be a space using the ROW
FORMAT clause. Hive also allows you to set some specific metadata
values related to a table. In the last line of the DDL Im indicating
that the text file contains two rows of header information that can
be skipped. This DDL can be cut and pasted into another HiveQL
script, or you can push the submit button to kick off the table
creation. This results in a task being sent to the Hadoop cluster that
will create the table. In Visual Studio youll see a job status window,
which you can refresh to track the tasks completion status. Once
the task is done, you can use Server Explorer to see the new table,
browse its definition or view the first 100 rows (from the context
menu of the database).

Cooking Data into Information

One of the key things youll want to know is how well your Web
site is working. A starting point might be to look at how often
users are getting client errors on their HTTP requests. You can take
a look at the different kinds of errors received by different source
referrers, group them by error type and list them by most impacted
(highest error count). Youll want to keep these results, so lets put
them in a new Hive table, which you can do using a CREATE
EXTERNAL TABLE statement. This time youll populate the table
by inserting calculated values using a select statement. Figure 3
shows what the code looks like.
Youll notice Im doing some cleanup in the data. For the cs-uriquery field, I simply want to see any query string that was passed in
with the HTTP request, but dont care about the X-ARR-LOG-ID
information stored in the logs. So, I simply replace that portion
of the string with a blank. Of course, as this is a declarative query
statement and not a procedure, its not possible to specify reuse of
the transformation I use for the cs-uri-query field. However, I can
count on the query execution runtime to optimize this situation.
Id further like to look at the top referrers for my Web site. To do
this I can simply select the cs_referer field from the IIS logs and
count the number of referrals by day and referrer. Once again, I
want to keep this data for further use, so I create another Hive
table called RefersPerDay and store the query results there. Figure 4
shows how to do the table creation and data insertion.
All of the code in Figure 3 and Figure 4 can be put into a single
Hive QL file. You can now create a new Hive application project,
enter this code and run the queries. The Hive project allows you
to save your query for later use. These examples are quite simple,
Microsoft Azure

Data Quality Tools

for Developers
A better way to build
in data verification
Since 1985, Melissa Data has provided
the tools developers need to enhance
databases with clean, correct, and
current contact data. Our powerful, yet
affordable APIs and Cloud services
provide maximum flexibility and ease
of integration across industry-standard
technologies, including .NET, Java, C, and
C++. Build in a solid framework
for data quality and protect your
investments in data warehousing,
business intelligence, and CRM.
Verify international addresses
for over 240 countries

Address
Verification

Phone
Verification

Email
Verification

Geocoding

Matching/
Dedupe

Enhance contact data with

phone numbers and geocodes
Find, match, and eliminate
duplicate records

Change of
Address

Sample source code for rapid

application development
Free trials with 120-day ROI
guarantee

Melissa Data.
Architecting data quality success.

| United States | United Kingdom | Germany | India |

Untitled-4 1

9/8/14 4:18 PM

Figure 5 Submitting a Query

but theyre based on the actual format of IIS logs. Therefore, they
should work against your actual server logs and serve as a basis for
creating more complex analyses.
Of course, running the queries means submitting them to your
Hadoop cluster in Azure. To do this, you have to choose the cluster
where you want to submit the job. The HDInsight toolbar allows
you to select your cluster from a list of clusters available across your

subscriptions. As you can see in

Figure 5, I selected the WeblogAnalytics cluster. Choosing Submit
will send the query to that cluster.
Here Im showing an advanced submit, where you can add a friendly
name for the resulting job that will
appear in the cluster. This makes it
easier to track running queries. You
can also add further arguments for
the Hive processor.
As soon as the query is submitted, youll see a job summary
window like the one in Figure 6,
indicating the current state of your
query as its initialized, starts to
run, and eventually completes or
fails. Your Hive queries are turned
into MapReduce tasks that can be
fanned out across a large number of
nodes for execution. This is where
the inherent power of the Hadoop
processing model kicks in to allow
massive scaling. In the job summary view, you can see the job go
through different percentages of the map and reduce phase until
it completes all queries that were submitted together.
When the job finishes, you can review the associated query and
look at job output and exceptions or errors through the links under
Job Details. Of course, the query put summarized data into new
tables and you can go back to Server Explorer to look at those tables
and their content. You can also run new queriesones that use the
data in these new tables for further summarizing or aggregation.

Wrapping Up
With HDInsight you can quickly and easily access large amounts of
data youve generated or transferred to the cloud. You can spin up
Hadoop clusters to process the data, accessing the data from anywhere in Azure. Hive allows you to specify the structure of your data
and to evolve with the changes that occur in your data format over
time. HDInsight Tools for Visual Studio further allows you to explore
your Hadoop clusters and data in intuitive ways. More important,
it allows you to create Hive query projects and manage them like
any other code, helping you move from simple, working queries to
complex information processing over time. Aggregate data created
using HDInsight is best visualized through tools such as Power BI,
which offer direct connection to HDInsight sources. Other capabilities, like machine learning, can also be used in conjunction with
these tools to extend the analytic possibilities. Watch this space for
more on solving Big Data problems using Azure services.
Q
OMID AFNAN is a principal program manager in the Azure Big Data team working
on implementations of distributed computation systems and related developer
tool-chains. He lives and works in China. Reach him at omafnan@microsoft.com.

Figure 6 The Hive Job Summary Window

32 msdn magazine

THANKS to the following technical experts for reviewing this article:

Duc Le and Tony Murphy
Microsoft Azure

/update/2015/01
www.componentsource.com

LEADTOOLS Document Imaging SDKs V19 from $2,995.00 SRP

BEST SELLER

Add powerful document imaging functionality to desktop, tablet, mobile & web applications.
Universal document viewer & conversion framework for PDF, Office, CAD, TIFF & more
OCR, MICR, OMR, ICR and Forms Recognition supporting structured & unstructured forms
PDF & PDF/A Create, Load, Save, View, Edit
Barcode Detect, Read, Write for UPC, EAN, Code 128, Data Matrix, QR Code, PDF417
Zero-footprint HTML5/JavaScript UI Controls & Web Services

ComponentOne Studio Enterprise 2014 v3 from $1,315.60

BEST SELLER

.NET Tools for the Professional Developer: Windows, HTML5/Web, and XAML.
Hundreds of UI controls for all .NET platforms including grids, charts, reports and schedulers
A line of HTML5 and JavaScript products for enterprise application development
Built in themes and an array of designers for creating custom themes and styling
40+ Windows 8.1 & Windows Phone 8.1 controls and Universal Windows app support
All Microsoft platforms supported, Visual Studio 2013, ASP.NET, WinForms, WPF & more

Help & Manual Professional from $583.10

BEST SELLER

Easily create documentation for Windows, the Web and iPad.

Powerful features in an easy accessible and intuitive user interface
As easy to use as a word processor, but with all the power of a true WYSIWYG XML editor
Single source, multi-channel publishing with conditional and customized output features
Output to HTML, WebHelp, CHM, PDF, ePUB, RTF, e-book or print
Styles and Templates give you full design control

Aspose.Total for .NET from $2,449.02

BEST SELLER

Every Aspose .NET component in one package.

Programmatically manage popular file formats including Word, Excel, PowerPoint and PDF
Work with charts, diagrams, images, project plans, emails, barcodes, OCR and OneNote files
alongside many more document management features in .NET applications
Common uses also include mail merging, adding barcodes to documents, building dynamic
reports on the fly and extracting text from most document types

1996-2014 ComponentSource. All Rights Reserved. All prices correct at the time of press. Online prices may vary from those shown due to daily fluctuations & online discounts.

US Headquarters
ComponentSource
650 Claremore Prof Way
Suite 100
Woodstock
GA 30188-5188
USA

Untitled-1 1

European Headquarters
ComponentSource
30 Greyfriars Road
Reading
Berkshire
RG1 1PE
United Kingdom

Asia / Pacific Headquarters

ComponentSource
3F Kojimachi Square Bldg
3-3 Kojimachi Chiyoda-ku
Tokyo
Japan
102-0083

We accept purchase orders.

Contact us to apply for a credit account.

Sales Hotline - US & Canada:

(888) 850-9911
www.componentsource.com

11/17/14 11:56 AM

B I G D ATA

Protect Your Online

Services with Big Data
and Machine Learning
Alisson Sol, Don Ankney and Eugene Bobukh
There are currently several methods for protecting online
services, from the Security Development Lifecycle to operational
processes for rapid incident response. One of the primary assets
of online services is usually overlookedthe Big Data created by
request logs and operational event monitoring. This article will
explore usage data processing and machine learning (ML) techniques to improve security, based on the experiences of protecting
online assets in the Microsoft Applications & Services Group (ASG),
including services such as Bing, Bing Ads and MSN.
Most online services create several streams of logged data. While
theres no standard taxonomy for the kinds of measurements you
can store about a service, when youre exploring that data seeking
security issues, you can broadly categorize it as usage data or
operational data. Usage data includes any logged value regarding
This article discusses:
Protecting online services from attacks
Analyzing usage data to determine nature of attacks
Running test attacks against an attack surface

Technologies discussed:
Microsoft Bing, Microsoft Azure, Windows PowerShell

Code download available at:

msdn.microsoft.com/magazine/msdnmag0115
34 msdn magazine

use of the service by its target audience. A common example is a

log entry for requests made to a Web site:
#Fields: date time s-ip cs-method cs-uri-stem cs-uri-query s-port
cs-username c-ip cs(User-Agent) cs(Referer) sc-status sc-substatus scwin32-status time-taken
2014-10-27 20:46:57 127.0.0.1 GET /search q=elect
ion+results&form=PRUSEN&mkt=en-us 80 - 127.0.0.1
Mozilla/5.0+(Windows+NT+6.4;+WOW64;+Trident/7.0;+Touch;+rv:11.0)+like+Gecko
- 200 0 0 5265

This type of log entry contains data not only about the requested
resource, but also the client browser, return code and time taken to
complete the request. More sophisticated services may enrich usage
data with derived information such as geolocation or applicationspecific information like user identification (for logged-in users).
There would be no usage data without actual users, except perhaps
for testing and monitoring agents.
Operational data refers to server and service operational
measurements. This includes CPU utilization or temperature, disk
space, network transfer rate, application exceptions, memory faults,
and similar factors logged as soon as a server is turned on and a
service started. In modern datacenters, log information typically
includes not only computing devices, but also aspects such as
air-conditioning measurements, presence of personnel and visitors
in zones containing sensitive data, doors being opened and closed,
and similar information required by operational security standards.
The code samples in this article will focus on processing usage
data. However, you could apply most of the principles outlined and
demonstrated here to identify vulnerabilities using operational

Azure HDInsight Processing

Binaries
ClusterBinariesContainer

Input

Output

InputContainer

ClusterOutputContainer

ClusterName
Status
ClusterStatusContainer

InputStorageAccount

ClusterStorageAccount

Figure 1 The Azure HDInsight Processing Environment

data. You can also improve your chances of identifying security

incidents by correlating usage data with operational data.

Attacking Endpoints
The pace of changes for a large online service makes it hard to protect
using only typical Security Development Lifecycle practices, such
as code reviews and static analysis tools. Thousands of changes are
committed every month. And there are often at least a few hundred
experiments in flight at any given point. These experiments present
new features to select users to gather feedback before widespread release.
Besides following good development practices and having penetration
test teams constantly trying to pinpoint vulnerabilities, its important to
automate as much of the vulnerability discovery as possible.
Near the end of 2014, Microsoft Bing services generated hundreds
of terabytes of usage data on a daily basis, logging up to hundreds of
billions of requests. Its safe to assume some of these requests were
actually attacks, trying to identify or exploit vulnerabilities. A typical
query to Bing is made sending to the service a URL request:
http://www.bing.com/search?q=election+results&form=PRUSEN&mkt=en-us

In this example, the user is searching for election results. There

are two other parameters in the URL that identify the Web form
that originated the request and the market setting (in this case,
Figure 2 Windows PowerShell Script XML Configuration File
<?xml version="1.0" encoding="utf-8"?>
<Configuration>
<SubscriptionName>Your-SubscriptionName</SubscriptionName>
<ClusterName>Your-ClusterName</ClusterName>
<ClusterStorageAccountName>Your-ClusterStorageAccountName
</ClusterStorageAccountName>
<ClusterBinariesContainer>Your-ClusterBinariesContainer
</ClusterBinariesContainer>
<MapperBinary>UsageDataMapper.exe</MapperBinary>
<ReducerBinary>UsageDataReducer.exe</ReducerBinary>
<ClusterOutputContainer>Your-ClusterOutputContainer</ClusterOutputContainer>
<ClusterStatusContainer>Your-ClusterStatusContainer</ClusterStatusContainer>
<InputStorageAccountName>Your-InputStorageAccountName
</InputStorageAccountName>
<InputStorageAccountKey>Your-InputStorageAccountKey</InputStorageAccountKey>
<InputContainer>Your-InputContainer</InputContainer>
<DeployBinaries>true</DeployBinaries>
<DeployFlavor>Release</DeployFlavor>
<JobTimeOut>3600</JobTimeOut>
</Configuration>

msdnmagazine.com

indicating the user language is English and in the United States).

You may understand this as a call to the search application within
the Bing domain, with parameters q, form and mkt. All such
requests are expressed in a canonical format, like so:
http://www.bing.com/search?q=[]&form=[]&mkt=[]

There are other applications within the Bing domain answering

similar requests. A request asking for election results in the video
format would be:
http://www.bing.com/videos?q=election+results&form=PRUSEN&mkt=en-us

As online services grow, new applications and features are added

dynamicallysome for convenience and others for compatibility.
Different formats are often allowed for the same request. Bing
videos would also accept this request:
http://videos.bing.com/?q=election+results&form=PRUSEN&mkt=en-us

Assuming you could derive from the usage logs a list with all
canonical requests to a service, you could then probe for vulnerabilities trying to inject known malicious payloads into the parameter
values. For example, an intruder could use the following request
to verify if the Bing video application is vulnerable to cross-site
scripting (XSS) in the query parameter:
http://www.bing.com/videos?q=<script>alert("XSS")
</script>&form=PRUSEN&mkt=en-us

An intruder scanning for vulnerabilities will also test the

responses for malicious payloads injected into other parameters, in
all possible combinations. Once a vulnerability is found, an attack
can be launched. Attack URLs are usually included in spam messages, in hopes a small percentage of users will carelessly click on
the links. Some of those users may even suspect URLs containing
JavaScript keywords. However, encoding the requests makes it more
difficult to promptly identify attacks:
http://www.bing.com/videos?&form=PRUSEN&mkt=en-us&q=%3CscRipt%3Ealert(%2
2XSS%22)%3C%2FScriPT%3E

You can write an application that accepts the list of canonical

requests to a service as input, injecting malicious payloads for
each kind of possible vulnerability and detecting from the service
answer if the attack succeeded. You can find the code for individual
detectors for several kinds of vulnerabilities (XSS, SQL injection
and open redirects) online. In this article, well focus on finding
the attack surface for online services from the usage data logs.

Processing Environment
Weblogs are usually distributed across several machines, making
sequential log file reads extremely efficient (even better if the files
are partitioned across different storage devices in a distributed file
system). That makes processing Weblogs a great application for the
MapReduce framework.
For this example, well place Weblogs in Microsoft Azure Blobs
under the same container called InputContainer. As a processing
platform, well use Azure HDInsight Streaming MapReduce jobs.
Theres good information already available online on how to set up
and configure HDInsight clusters. The code explained in this article
will generate binaries you should place in a container accessible to
the HDInsight cluster, referred to as ClusterBinariesContainer. As
code executes and processes input, it will create output in another
container called the ClusterOutputContainer, along with status
information saved to the ClusterStatusContainer. A visualization of
the Azure HDInsight processing configuration is shown in Figure 1.
January 2015 35

Figure 3 Function to Extract Key and Values from Log Line

private static string ExtractKeyAndValuesFromLogLine(string inputLogLine)
{
StringBuilder keyAndValues = new StringBuilder();

Figure 4 Function to Get Just Parameter Names from Query

private static string ExtractParameterNamesFromQuery(string query)
{
StringBuilder sb = new StringBuilder();

string[] inputColumns = inputLogLine.Split(DataFormat.MapperInputColumnSeparator);

string uriReference = inputColumns[DataFormat.MapperInputUriReferenceColumn];
string uriQuery = inputColumns[DataFormat.MapperInputUriQueryColumn];
string parameterNames = ExtractParameterNamesFromQuery(uriQuery);

// Go through each parameter adding to output string

string[] nameValuePairs = query.Split(DataFormat.ParametersSeparator);
Array.Sort(nameValuePairs, StringComparer.InvariantCultureIgnoreCase);
List<string> uniqueParameterNames = new List<string>();
foreach (string nameValuePair in nameValuePairs)
{
int indexOfSeparatorParameterNameFromValue =
nameValuePair.IndexOf(DataFormat.ParameterNameFromValueSeparator);
string paramName = nameValuePair;

// Key = uriReference + separator + parameterNames

keyAndValues.Append(uriReference);
keyAndValues.Append(DataFormat.ReferenceFromQuerySeparator);
keyAndValues.Append(parameterNames);
keyAndValues.Append(DataFormat.MapperOutputColumnSeparator);
keyAndValues.Append(DataFormat.OneOccurrence);

paramName = nameValuePair.Substring(0,
indexOfSeparatorParameterNameFromValue);
uniqueParameterNames.Add(paramName);
sb.Append(paramName);
sb.Append(DataFormat.ParameterNameFromValueSeparator);
sb.Append(DataFormat.OneOccurrence);
sb.Append(DataFormat.ParametersSeparator);

return keyAndValues.ToString();
}

You need to replace the placeholder names in Figure 1 with values

for your specific configuration. You can set these in a configuration
file. The Windows PowerShell script that will create and execute
the HDInsight job will read the XML configuration file shown in
Figure 2. After configuring the file, youll most likely execute the
script for usage data analysis from within an Azure PowerShell
prompt, having properly configured your Azure account to have
the authorization to access the storage and compute services (see
Get-AzureAccount and Add-AzureAccount cmdlets help).

}
return sb.ToString();
}

Figure 5 Reducer Main Loop

public static void Main(string[] args)
{
string currentKey, previousKey = null;
int count = 0;
Console.SetIn(new StreamReader(args[0]));
string inputLine;
while ((inputLine = Console.ReadLine()) != null)
{
string[] keyValuePair =
inputLine.Split(DataFormat.ReducerInputColumnSeparator);
currentKey = keyValuePair[0];
if (currentKey != previousKey)
{
Console.WriteLine(DataFormat.ReducerOutputLineFormat, previousKey, count);
count = 1;
previousKey = currentKey;
}
else count++;
}

Map to Canonical Requests

Getting the attack surface for the online services using the
MapReduce processing environment consists of creating a mapper
application to extract the URLs from the Weblogs and transform
them into canonical form. That value becomes the key for the
reducer, which will then eliminate duplicates. Thats the same principle used in the sample word count application available for HDInsight.
Removing any comments and validation code, the following code
demonstrates the main entry point for the mapper application:
public static void Main(string[] args)
{
Console.SetIn(new StreamReader(args[0]));
string inputLogLine;
while ((inputLogLine = Console.ReadLine()) != null)
{
string outputKeyAndValues =
ExtractKeyAndValuesFromLogLine(inputLogLine);
Console.WriteLine(outputKeyAndValues);
}
}

This code goes through every input line and extracts the unique
key, as well as any complementary values relevant to the problem
being solved. For example, if you were seeking the most common
user queries, the key would be the value passed for the query
parameter. Raw log lines appear as follows:
#Fields: date time s-ip cs-method cs-uri-stem cs-uri-query s-port
cs-username c-ip cs(User-Agent) cs(Referer) sc-status sc-substatus scwin32-status time-taken
2014-10-27 20:46:57 127.0.0.1 GET /search q=elect
ion+results&form=PRUSEN&mkt=en-us 80 - 127.0.0.1
Mozilla/5.0+(Windows+NT+6.4;+WOW64;+Trident/7.0;+Touch;+rv:11.0)+like+Gecko
- 200 0 0 5265

Columns cs-uri-stem and cs-uri-query have the relevant information you need to have parsed to get the canonical form of the
36 msdn magazine

Console.WriteLine(DataFormat.ReducerOutputLineFormat, previousKey, count);

}

request (the sample code doesnt include multiple hosting processing). The function to extract key and values from each log line
is outlined in Figure 3.
The only missing logic relates to extracting the parameter names
from the query column. Code to perform the task is shown in
Figure 4. The input for that functionthe previously provided
sample linewould be a string like this:
q=election+results&form=PRUSEN&mkt=en-us

The canonical form used in the sample code will remove the
parameter values, sort the parameter names and transform this
into a still valid query string:
form=1&mkt=1&q=1&

Sorting the parameter names helps avoid duplication, because

Web requests dont depend on parameter order. The placeholder
used for the parameter value is 1, instead of [] because its shorter.
It may also be used for other things like counting the number of
times parameters appear in all combinations of request parameters, as shown in Figure 4.
Big Data

Reduce the Attack Surface

The mapping code sequentially reads the Weblog lines, then outputs
a key and value for each. MapReduce has a combine phase, which
assembles all records with the same key for processing by the reducer
code. If an input log had several lines doing search queries, by now
those produce identical output:
search?form=1&mkt=1&q=1&
search?form=1&mkt=1&q=1&
search?form=1&mkt=1&q=1&

1
1
1

Figure 5 has the outline for the reducer code. It reads input lines
and splits those into key and values. It keeps
a counter until the key changes, and then
outputs the result.
You could easily modify the code and script
provided in this article for other purposes.
Change function ExtractKeyAndValuesFromLogLine to have the parameter values
as the keys and youd have a useful distribution of value frequency. In its current form,
the output will be a list with the attack surface, showing the normalized application
path and frequency of requests:
search?form=1&mkt=1&q=1&
video?form=1&mkt=1&q=1&

may still be activated by humans or automated processes. While

seeking an attacker, you may uncover the need to have a compatibility mode for old URLs, and automatically redirect those to new
versions of the application. Thats a good business result.
Developing the attacking application is a journey you should take
with care. Assuming your detectors are all perfect, youll now impose
a load onto your service that needs to be properly throttled. You need
to avoid creating a denial-of-service attack or affect performance for
real users. Its also essential to avoid numerous false positives. If that

3
10

Understand Service Traffic

The attack surface will already be valuable to
help you understand whats happening with
your service, even if you dont perform proactive
penetration testing to expose vulnerabilities.
Although request frequency changes per hour,
per day of the week, and with other seasonal
factors, the normal behavior over time changes
with feature releases or even coordinated
attacks. For example, Bing receives hundreds of billions requests per day. The attack
surface list typically has hundreds of thousands of values per day. Not all such paths are
even expected. Figure 6 summarizes whats
found on the attack surface on a typical day.
The first row indicates the top 10 canonical
paths amount to 89.8 percent of the usual
request traffic. The next top 10 paths add up
to 6.3 percent of the request count (or 96.1
percent for the top 20). Those are really the
top applications for the service. Everything
else amounts to less than 4 percent. Such a
request pattern is very different for a site with
syndicated content, like MSN.com.
Its particularly interesting to note that
about two-thirds of the requests are to
unique paths. Some of those requests may
come from attackers trying to probe for application parameters that might trigger certain
functionality. Yet the very nature of online
services generates a lot of such traffic. Links
to your service stored a few years back
msdnmagazine.com

January 2015 37

happens, incident reports from the attacking

service will soon be ignored.

Figure 6 Typical Distribution of

Request Paths for Bing in 2013

module as part of a solution and start receiving

evaluation requests. The output is then subject to a scoring process, which will indicate
Canonical Paths
Percentage
Learn from Service Data
whether the ML module correctly identified
Top 10
89.8
ML lets you automate several processes that
the requests (true positives and true negatives),
Top 20
96.1
would be difficult to do by directly coding
missed suspicious requests (false negatives) or
Paths with <= 1,000 requests 99.9
instructions or rules. For example, it would
generated a false alert (false positive).
Paths with <= 100 requests
99.6
be hard to conceive the code for a computer
If the initial experiments produced a good
Paths with <= 10 requests
97.6
vision application that would detect a person
enough ML module based on synthetic data,
is in front of a camera. After labeling thouthat module should be fairly accurate with few
Paths with = 1 request
67.5
sands of images with depth information,
incorrectly evaluated actual user requests. You
however, the Kinect team at Microsoft was able to train an ML can then properly label those that were wrongly evaluated and add
module to do this with sufficient accuracy. The labeled depth them to the ground truth. A few more experiments and training
images indicating not only human presence, but also body posi- should now generate a new ML module with restored accuracy. As
tion, enabled the learning process.
you carefully iterate this process, the initial synthetic data becomes
An attack service that generates requests of known categories (XSS, a smaller part of the ground truth used in the training process, and
SQL injection and so on) automates an important part of the process iterations of the ML module get better at accurately evaluating user
to use ML methods for evaluating online traffic. It generates a large requests. For additional validation, you can use the ML module
body of synthetic ground truth. Check the usage logs to easily iden- for offline applications examining usage logs and identifying
tify all such attack requests known to be made at a certain time by malicious requests. After sufficient development, you can deploy
the attack service. Theyre now mixed with user requests, for which the ML module online to evaluate requests in real time and prevent
there isnt yet a known classification (normal or malicious request). attacks from ever hitting back-end applications.
To create a stratification of the data that truly matches the
traffic hitting the service, you have to understand the nature of Wrapping Up
that traffic. Generating an experimental sample with 0.1 percent of While you should continue to follow solid development processes
the usage data for a service receiving 100 billion requests a day still (including the Security Development Lifecycle), you should also
results in 100 million requests. Only synthetic data can help create assume your online service may be under attack at any point.
such an initial ground truth.
Usage logs can provide you with insightful information about how
Assuming you have high-quality ground truth and adequate such attacks are occurring. Knowing your attack surface will help
tools, the iterative cycle for the learning process for an ML solu- you proactively attack your service to identify and close vulneration to evaluate user requests is outlined in Figure 7. Starting with bilities before theyre exploited. Building that attack service then
synthetic data in the ground truth, you can make experiments and creates synthetic ground truth, enabling the use of ML techniques
kick off a training process in the ML module to classify requests to train an ML module to evaluate service requests. Building the
(into categories such as normal, XSS, SQL injection and so on) or attack service is not a trivial task, but the immediate and long-term
make a regression (indicating the confidence a request belongs business results more than justify the investment.
Q
to one or more of the categories). You could then deploy this ML
Learning Cycle

Ground Truth:
Training Data
(Log Labeled Data)

Scoring
True Positive
True Negative
False Positive
False Negative

Action

Boosting GT
Boosting GT

Solution

DON ANKNEY is a senior security researcher in the Microsoft Information Platform

Group where he works on applying Microsoft machine learning investments to
the services security space. He was a founding member of Black Lodge Research,
an educational non-profit focused on security, and regularly teaches secure
development techniques at regional meet-ups, workshops and un-conferences.

ML Module
Prepare
Features

Feature
Experiments
ML Training

Classification
or Regression

Figure 7 Learning Cycle to Create a Machine

Learning-Based Solution
38 msdn magazine

ALISSON SOL is a principal architect for Microsoft. He has many years of software
development experience, with a focus on image processing, computer vision, ERP
and business intelligence, Big Data, machine learning and distributed systems.
Prior to starting at Microsoft in 2000, he cofounded three software companies,
published several technical papers and recorded several patent applications. Read
his blog at AlissonSol.com/blog.

Scoring
Process

E UGENE B OBUKH is a senior security/data program manager at Microsoft.

Focusing on applying scientific approach to security problems, he conducted security
testing and performed security reviews for more than 200 Microsoft releases since
2000, including .NET, Silverlight, and Internet Explorer. Some of Bobukhs work is
described at blogs.msdn.com/b/eugene_bobukh.

THANKS to the following Microsoft technical experts for reviewing this article:
Barry Markey and Viresh Ramdatmisier
Big Data

Untitled-1 1

11/17/14 2:56 PM

vslive.com

2015 Dates Announce

Redmon
Code Home
August 10 14
Microsoft HQ
Redmond, WA

JUST
ADDED!

Sa FranciVc
Code By The Bay
June 15 18
The Fairmont
San Francisco, CA

Visual Studio Live!

is hittin the open road on the

ultimate code trip! For 22
years, we have helped tens of
thousands of developers
navigate the .NET Highway,
featuring code-lled days,
networking nights and the
best in independent training:

SAN FRANCISCO

Multi-track Events
 Focused, Cutting-edge .NET
Education
 Covering the Hottest Topics
 Relevant and Useable
Content
 Expert Speakers, Including
Many Microsoft Instructors
 Interactive Sessions
 Discuss Your Challenges
Find Actionable Solutions


Las Vegas

Code on the Strip

March 16 20
Ballys Hotel & Casino
Las Vegas, NV

MAKE PLANS TO JOIN US IN 2015!

SUPPORTED BY

PRODUCED BY
magazine

Untitled-2 2

12/8/14 2:37 PM

NAVIGATE THE
.NET HIGHWAY

New York
WITH

NOW

The Code That Never Sleeps

September 28 October 1
NY Marriott at Brooklyn
Bridge Brooklyn, NY

Locations
t Cho
Fro!

Orland

Code in the Sun

PART OF LIVE! 360
Tech Events with Perspective

AuVti
Dont Mess with Code
June 1 - 4
Hyatt Regency
Austin, TX

SharePoint Live!
SQL Server Live!
Modern Apps Live!
TechMentor
November 16 20
Loews Royal Pacic Resort
Orlando, FL

vslive.com
Untitled-2 3

12/8/14 2:38 PM

MICROSOFT AZURE

Occasionally Connected
Data in Cross-Platform
Mobile Apps
Kevin Ashley
Most mobile apps need to stay offline for some periods of process in the OS. In addition, there are different types of data
time, and mobile users expect their apps to work gracefully in both
connected and disconnected states. Hundreds of millions of mobile
device users might not be aware of needs that the apps have with
regard to online and offline states; they simply expect the apps to
work under any circumstances. In this article Ill show you ways
to enable your mobile app to work in both states and synchronize
data gracefully with the cloud in Windows, iOS and Android using
cross-platform Xamarin tooling and Microsoft Azure Mobile Services.
As a mobile app developer myself, I encountered the need to
synchronize offline data early on. For my Winter Sports ski app
(winter-sports.co) and Active Fitness activity tracking app (activefitness.co),
its somewhat expected that you cant get a decent connection on
the slopes or on the run. So those apps need to be able to synchronize data collected offline, without significant impact on battery
life and reliability. In other words, the apps need to work efficiently
in any conditions.
Theres more than meets the eye when considering persistent
storage. To start, there are multiple approaches to synchronizing,
meaning it can be done from within an app or as a background
This article discusses:
Using cross-platform functionality
Offline synchronizing of structured data
Manually synchronizing serialized data
Synchronizing unstructured data with the cloud

Technologies discussed:
Microsoft Azure Mobile Services, Xamarin

Code download available at:

bit.ly/11yZyhN
42 msdn magazine

stores, such as semistructured data from sensors and potentially

relational data stored in SQLite. Its also important to implement
a conflict resolution policy so data loss and degradation are minimized. Finally, there are a wide variety of data formats to manage,
such as binary, JSON, XML and custom.
Mobile apps typically store several types of data. Structured data,
such as JSON or XML, is typically used for local settings, local files
or caching. In addition to storing data in files, you also have a choice
of storage engines, such as SQLite, to store and query data. Mobile
apps may also store blobs, media files and other types of large
binary data. For these types of data, Ill demonstrate techniques that
make data transfer more reliable on occasionally connected devices.
Ill provide an overview of several technologies. For example,
instead of focusing strictly on offline sync for structured data, Ill
show you a broader picture of both structured and unstructured
(blob) data synchronization techniques. Ill also use a cross-platform
approach everywhere throughout these examples.

The Cross-Platform Approach

Because its becoming increasingly popular to connect sensorenabled devices to the cloud, I included device sensor data in my
project to demonstrate various methods to sync it with the cloud. Ill
discuss three scenarios: offline data sync, manual data synchronization,
and synchronizing large media and binary data. The accompanying
code sample is completely cross-platform, with 100 percent
reusability for Android, iOS and Windows. To achieve that, I used
Xamarin.Forms, cross-platform XAML/C# tooling that works well
on iOS, Android and Windows and is getting integrated with Visual
Studio tools (see the Microsoft Channel 9 video, Cross-Platform
Mobile Development Using Visual Studio, at bit.ly/1xyctO2).
In the code sample are two classes that manage cross-platform data
models: SensorDataItem and SensorModel. This approach may be

used by many sports and fitness tracking apps, such as Active Fitness,
or apps that need to synchronize structured data from local storage
with the cloud. I added latitude, longitude, speed and distance to the
SensorDataItem class as an example of data collected by sensors,
such as GPS, to illustrate the idea. Of course, the data structure in the
actual app might be more complicatedand include dependencies
but my example will give you an idea of the concept.

Synchronizing Structured Data with Offline Sync

Offline sync is a powerful new feature in Azure Mobile Services.
You can reference the Azure Mobile Services package in your Visual
Studio project using NuGet. More important, its also supported
in cross-platform apps with the new version of the Azure Mobile
Services SDK. That means you can use this feature in your Windows,
iOS and Android apps that occasionally need to connect to the
cloud and synchronize their states.
Ill start with a few concepts.
Sync Table This is a new object in the Azure Mobile Services
SDK created to distinguish tables that support synchronization from
local tables. Sync tables implement the IMobileServiceSyncTable<T>
interface and include additional sync methods, such as PullAsync,
PushAsync and Purge. If you want to synchronize your offline
sensor data with the cloud, you need to use sync tables instead
of standard tables. In my code sample, I initialize my sensor data
sync table by using the GetSyncTable<T> call. In the Azure Mobile
Services portal, I created a regular table called SensorDataItem
and added the code in Figure 1 to my client initialization (you can
download the full source code from bit.ly/11yZyhN).
The Synchronization Context This is responsible for synchronizing data between local and remote stores. Azure Mobile
Services ships with SQLiteStore, which is based on the popular
SQLite library. The code in Figure 1 does a few things. It checks
whether a synchronization context has already been initialized and,
if not, it creates a new instance of the SQLite store from the local.
db file, defines the table based on the SensorDataItem class and
initializes the store. To handle pending operations, the synchronization context uses a queue, accessible via the PendingOperations
property. The synchronization context provided by Azure Mobile
Services is also smart enough to distinguish update operations
happening in the local store. Synchronization is done automatically
by the system, so you dont have to manually and unnecessarily call
into the cloud to persist data. This is good because it lowers traffic
and increases device battery life.
The Push Operation This lets you explicitly synchronize data
between the local store and the cloud store by pushing local data
to the server. Its important to point out that in the current version
of the Azure Mobile Services SDK, you need to invoke push and
pull explicitly to synchronize the context. The push operation
executes on the whole synchronization context to help you preserve
relationships between tables. For example, if I have relationships
between tables, my first insert will give me an Id of the object, and
subsequent inserts will preserve referential integrity:

to specify a subset of data, or any OData query. Unlike push operations, which happen on an entire context, pull is executed on the table
level. If items are pending in the synchronization queue, those items
are pushed firstbefore pull operations executeto prevent data loss
(yet another benefit from using Azure Mobile Services for data synchronization). In this example, I pull data that has a nonzero speed
(collected by my GPS sensor, for example), stored earlier in the server:
var query = sensorDataTable.Where(s => s.speed > 0);
await sensorDataTable.PullAsync(query);

The Purge Operation This clears data specified from local and
remote tables, causing synchronization. Similar to pull, you can
use LINQ to specify a subset of data, or any OData query. In this
example, I purge data that has zero distance (which might also
come from my GPS sensor) from my tables:
var query = sensorDataTable.Where(s => s.distance == 0);
await sensorDataTable.PurgeAsync(query);

Appropriately Handling Conflicts This is an important part

of your data synchronization strategy when devices come online
and offline. Conflicts will happen, and the Azure Mobile Services
SDK provides ways to deal with conflicts. For conflict resolution to
work, I enabled the Version property column on the SensorDataItem
object. In addition, I created the ConflictHandler class, which
implements the IMobileServiceSyncHandler interface. When you
need to resolve a conflict, you have three options: keep the client
value, keep the server value or abort the push operation.

Figure 1 Leveraging the MobileServiceSQLiteStore

Object for Synchronization
// Initialize the client with your app URL and key
client = new MobileServiceClient(applicationURL, applicationKey);
// Create sync table instance
todoTable = client.GetSyncTable<SensorDataItem>();
// Later in code
public async Task InitStoreAsync()
{
if (!client.SyncContext.IsInitialized)
{
var store = new MobileServiceSQLiteStore(syncStorePath);
store.DefineTable<SensorDataItem>();
await client.SyncContext.InitializeAsync(store,
new MobileServiceSyncHandler ());
}
}

Figure 2 Data Structure for Data Synchronization

public class SensorDataItem
{
public string Id { get; set; }
[Version]
public string Version { get; set; }
[JsonProperty]
public string text { get; set; }
[JsonProperty]
public double latitude { get; set; }
[JsonProperty]
public double longitude { get; set; }
[JsonProperty]
public double distance { get; set; }

await client.SyncContext.PushAsync();

The Pull Operation This lets you explicitly synchronize data by

pulling data from a remote store into the local store. You can use LINQ
msdnmagazine.com

[JsonProperty]
public double speed { get; set; }
}

January 2015 43

In my example, check the ConflictHandler class. When its

initialized, in the constructor I set it with one of three conflict
resolution policies:
public enum ConflictResolutionPolicy
{
KeepLocal,
KeepRemote,
Abort
}

Depending on the method, each time a conflict happens, in the

ExecuteTableOperationAsync method, I automatically apply my
conflict resolution policy. When I initialize my synchronization
context, I pass my ConflictHandler class to the synchronization
context with my default conflict resolution policy:
await client.SyncContext.InitializeAsync(
store,
new ConflictHandler(client, ConflictResolutionPolicy.KeepLocal)
);

To read more on conflict resolution, see the MSDN sample,

Azure Mobile Services - Handling Conflicts with Offline WP8,
at bit.ly/14FmZan and the Azure documentation article, Handling
Conflicts with Offline Data Sync in Mobile Services, at bit.ly/1zA01eo.

Synchronizing Serialized Data Manually

Before Azure Mobile Services started offering offline synchronization, developers had to implement data synchronization manually.
So, if you develop an app that occasionally needs to synchronize
data and you dont use the Azure Mobile Services offline sync
feature, you can manually do it (although I strongly recommend
looking at the offline sync feature). You can use either direct object
serialization (such as a JSON serializer) into files, or data storing
engines such as SQLite. The main difference between the offline
sync mechanism and manual synchronization is that you need
to do most of the work yourself in the latter. One of the methods
Figure 3 Download Operation (Cross-Platform Code)
public static async Task<bool> DownloadFileAsync(
IFolder folder, string url, string fileName)
{
// Connect with HTTP
using (var client = new HttpClient())
// Begin async download
using (var response = await client.GetAsync(url))
{
// If ok?
if (response.StatusCode == System.Net.HttpStatusCode.OK)
{
// Continue download
Stream temp = await response.Content.ReadAsStreamAsync();
// Save to local disk
IFile file = await folder.CreateFileAsync(fileName,
CreationCollisionOption.ReplaceExisting);
using (var fs =
await file.OpenAsync(PCLStorage.FileAccess.ReadAndWrite))
{
// Copy to temp folder
await temp.CopyToAsync(fs);
fs.Close();
return true;
}
}
else
{
Debug.WriteLine("NOT FOUND " + url);
return false;
}
}
}

44 msdn magazine

to detect whether data has been synchronized is by using the Id

property of any object in your data model. For example, see the
SensorDataItem class used in my earlier examplenote the Id and
Version fieldsshown in Figure 2.
When a record is inserted into a remote database, Azure Mobile
Services automatically creates an Id and assigns it to the object, so the
Id will be a non-null value when the record has been inserted, and
null when the record has never been synchronized with the database:
// Manually synchronizing data
if (item.Id == null)
{
await this.sensorDataTable.InsertAsync(item);
}

Synchronizing deletes and updates manually is a far more

challenging and involved process and is beyond the scope of this
article. If youre looking for a comprehensive synchronization
solution, look at the offline sync features of the Azure Mobile
Services SDK. Of course, this example is simple compared with
real-life scenarios, but if youre looking to implement manual
data synchronization, this can give you an idea on where to start.
Certainly, because the Azure Mobile Services SDK offers a tested,
well-thought-out solution for synchronizing data, I recommend
trying the offline sync approach, especially in apps that require a
solid, tested method to keep local and remote data in sync.

Transferring Binary Data,

Photos and Media to the Cloud
In addition to structured data, apps often need to synchronize
unstructured or binary data or files. Think of a mobile photo app
or an app that needs to upload a binary file to the cloud, such as a
photo or video. Because I explore this subject in a cross-platform
context, different platforms have different capabilities. But are they
really that different? Synchronizing blob data can be achieved in
multiple ways, such as with an in-process service or by leveraging
an out-of process, platform-specific background transfer service.
To manage downloads, I also provided a simple TransferQueue
class, based on ConcurrentQueue. Every time I need to submit
a file for upload or download, I add a new Job object to the
queue. This is also a common pattern in the cloud, where you put
unfinished work in a queue and then have some other background
process reading from the queue and completing the work.
In-Process File Transfer Sometimes you need to transfer files
directly from within your app. This is the most obvious way of handling blob transfers, but as mentioned earlier, it has disadvantages.
To protect the UX, the OS puts a cap on an apps use of bandwidth
and resources. However, this assumes the user is using the app to
work. In the case of occasionally disconnected apps, it may not be
the best idea. The upside of transferring files directly from an app
is that it has full control over data transfer. With full control, an
app can take advantage of shared access signature methods to manage uploads and downloads. You can read about the advantages in
the Microsoft Azure Storage Team Blog post, Introducing Table
SAS (Shared Access Signature), Queue SAS and Update to Blob
SAS, at bit.ly/1t1Sb94. Although not all platforms provide this functionality built in, if youre willing to use a REST-based approach,
you can certainly take advantage of SAS keys from Azure Storage
Services. The downside to transferring files directly from an app
Microsoft Azure

Untitled-1 1

11/26/12 3:02 PM

is twofold. First, you must write more code. Second, the app must
be running, potentially draining battery life and limiting the UX.
The best solutions leverage some of the elegance of built-in data
synchronization techniques.
I provided source code for a basic upload and download operation
in a cross-platform Xamarin app in BlobTransfer.cs (see accompanying code download). This code should work on iOS, Android and
Windows. In order to use platform-independent file storage, I used
the PCLStorage NuGet package (Install-Package PCLStorage), which
allows me to abstract file operations on iOS, Android and Windows.
To initiate an in-process transfer, I call my TransferQueue AddInProcessAsync method:
var ok = await queue.AddInProcessAsync(new Job {
Id = 1, Url = imageUrl, LocalFile = String.Format("image{0}.jpg", 1)});

This schedules a typical in-process download operation, defined

in the BlobTransfer object as shown in Figure 3.
Of course, if you wish to upload a file, you can do it in-process
by using the method shown in Figure 4.
Out-of-Process File Transfer Using OS-Specific Transfer
Service Downloading and uploading using built-in file transfer

services has many advantages. Most platforms offer a service that

can transfer large files (both uploading and downloading) autonomously as a background service. You should leverage such services
whenever possible because they run out of processthat is, your
app isnt capped on the actual data transfer, which may be fairly
expensive in terms of resources consumed. Also, your app doesnt
have to stay in memory all the time to transfer files, and the OS
usually provides a conflict resolution (retry) mechanism to restart
uploads and downloads. Other advantages include less code to
write, the app doesnt need to be active (the OS manages its own
queue of uploads and downloads) and the app is more memory/
resource efficient. The challenge, however, is that this method
requires platform-specific implementation: iOS, Windows Phone
and so on have their own background transfer implementations.
Conceptually, a reliable upload for files in a mobile app using
an OS-specific out-of-process service looks similar to the in-app
implementation. But the actual upload/download queue management is outsourced to the OS transfer service. For Windows
Phone Store apps and Windows Store apps, developers can use
Figure 4 Upload Operation (Cross-Platform Code)
public static async Task UploadFileAsync(
IFolder folder, string fileName, string fileUrl)
{
// Connect with HTTP
using (var client = new HttpClient())
{
// Start upload
var file = await folder.GetFileAsync(fileName);
var fileStream = await file.OpenAsync(PCLStorage.FileAccess.Read);
var content = new StreamContent(fileStream);
// Define content type for blob
content.Headers.Add("Content-Type", "application/octet-stream");
content.Headers.Add("x-ms-blob-type", "BlockBlob");
using (var uploadResponse = await client.PutAsync(
new Uri(fileUrl, UriKind.Absolute), content))
{
Debug.WriteLine("CLOUD UPLOADED " + fileName);
return;
}
}
}

46 msdn magazine

BackgroundDownloader and BackgroundUploader objects. For

iOS 7 and higher, NSUrlSession provides CreateDownloadTask
and CreateUploadTask methods to initiate downloads and uploads.
Using my earlier example, I now need to call my out-of-process
method to invoke a call using an OS-specific background transferring service. In fact, because the service is handled by the OS,
Ill schedule 10 downloads to demonstrate the app isnt blocking
and the execution is handled by the OS (in this example, I used the
iOS background transferring service):
for (int i = 0; i < 10; i++)
{
queue.AddOutProcess(new Job { Id = i, Url = imageUrl,
LocalFile = String.Format("image{0}.jpg", i) });
}

For an example of a background transfer service for iOS, check

BackgroundTransferService.cs. In iOS you first need to initialize a
background session with CreateBackgroundSessionConfiguration
(note this only works on iOS 8 or later):
using (var configuration = NSUrlSessionConfiguration.
CreateBackgroundSessionConfiguration(sessionId))
{
session = NSUrlSession.FromConfiguration(configuration);
}

Then you can submit a long upload or download operation and

the OS will handle it independently from your app:
using (var uri = NSUrl.FromString(url))
using (var request = NSUrlRequest.FromUrl(uri))
{
downloadTask = session.CreateDownloadTask(request);
downloadTask.Resume();
}

You also need to think of a queueing mechanism to reliably

upload and download blobs.

Sample Code and Next Steps

All sample code for this article is available on GitHub at bit.ly/11yZyhN.
To use this source code, you can use Visual Studio with Xamarin
or Xamarin Studio, which is available from xamarin.com. The project
uses cross-platform Xamarin.Forms and the Azure Mobile Services
library with offline sync. For the next steps, it would be interesting to
see an out-of-process service added to community libraries, such as
Xamarin Labs, as well as queuing capabilities and conflict resolution
similar to whats currently provided for structured data in the Azure
Mobile Services Offline Sync SDK.
To recap, Microsoft Azure Mobile Services provide a powerful and
efficient way to synchronize offline data. You can use these services in
a cross-platform scenario on Windows, Android and iOS. Microsoft
also provides easy-to-use native SDKs that work on every platform.
You can improve your apps reliability in disconnected scenarios by
integrating these services and adding offline sync to your apps. Q
K EVIN A SHLEY is an architect evangelist at Microsoft. Hes a coauthor of
Professional Windows 8 Programming (Wrox, 2012) and a developer of top apps
and games, most notably Active Fitness (activefitness.co). He often presents on
technology at various events, industry shows and Webcasts. He works with startups and partners, advising on software design, business and technology strategy,
architecture, and development. Follow him at his blog at kevinashley.com and on
Twitter at twitter.com/kashleytwit.

THANKS to the following Microsoft technical experts for reviewing this article:
Greg Oliver and Bruno Terkaly
Microsoft Azure

Untitled-7 1

11/5/14 4:57 PM

AZURE INSIDER

BYOD Challenge:
Connect Mobile Devices
to the Enterprise
Greg Oliver and Bruno Terkaly
There are still many concerns surrounding the Bring
Your Own Device (BYOD) movement in corporate environments
and how Azure Mobile Services can help developers. A core value
of Mobile Services is that it democratizes identity and lowers the
cost and effort to provide secure access to protected corporate
resources. In this article, Well take a closer look at what it takes to
support more complex scenarios when connecting mobile applications to enterprise resources.
Well start by defining the identity stakeholders and look at the
similarities and differences with respect to employees, trading partners and customers. Well also look at software architectures that
provide secure connectivity for browser and mobile device users
accessing on-premises services and sites. Later in the article, well
demonstrate these techniques using an iOS app that connects to a
Mobile Services back end, as well as an on-premises service endpoint.

Not All Access Is Equal

There are several important goals of this effort. The first is to provide
access to on-premises resources with a minimum of custom code or
infrastructure changes, including existing network configurations.
Another goal is to manage these usage scenarios to ensure control
and visibility, while being agile and flexible.
Not all users who need access behind the corporate firewall are
created equal. And not all data behind the corporate firewall needs
the same level of protection. Some employees need greater access
to protected information. Employees like field salespersons need
access to the most up-to-date information on product prices and
This article discusses:
Supporting complex mobile enterprise apps
Multiple approaches to mobile identity management
Managing different categories of contacts

Technologies discussed:
Microsoft Azure, Azure Mobile Services
48 msdn magazine

inventory levels. They may even need access to accounting data on

their customers, such as recent payments on account. They probably dont, however, need access to human resources applications.
Trading partners have their own specialized needs. Trusted business
partners will frequently federate identity across an extranet. Trading
partners typically manage their own identity infrastructure to authenticate users and will use claims to moderate access to a companys
protected resources. Thus, either the partner or the company will use
trust policy to map the incoming claims to claims understood by a
protected resource, such as an internal Web application.
There are two other less common identity stakeholders. Customers
often need access to their account information, something like
PayPal or Netflix. A prospect is effectively a potential customer.
He might have given up credit card information, but hasnt really
bought anything yet. All these types of users share one common
aspect: They all want to bring and use their iOS, Android and
Windows Phone devices to work at the enterprise.

Enterprise Security Management

At least five software stacks play a role in securing mobile devices
in the enterprise. Here, well take a closer look at Hybrid Connections, Azure Service Bus, Azure Active Directory Application Proxy
and Microsoft Azure API Management.
It would be simple to have employees present their credentials to
Active Directory through a Web UI, get an access token and start
working, but the world of identity is more nuanced and complex.
The notion that one size fits all is simply not true. Employees are
typically forced to use a corporate identity provider.
Most enterprises have brought identity in-house, typically
leveraging Active Directory, which mostly exposes an LDAP-based
protocol. In this new mobile, cloud-enabled world, enterprises
need an HTTP-based communication protocol for mobile devices.
Microsoft has created such an identity system in Azure called Azure
Active Directory (Azure AD). One of the great features of Azure
AD is that enterprises can sync identities to the cloud using directory synchronization. Read more about that process at bit.ly/1ztaB9Z.

Corporate Network
Company A

Perimeter Network
Company A

Active Directory
Federation
Services

Federation Proxy

Employee for
Company A

Internet Users
Employee
Company A
Firewall

Firewall

Employee
Company B

Perimeter DNS

Perimeter Network
Company B

Corporate Network
Company B

Federation Proxy

Active Directory
Federation
Services

Firewall

Firewall

Perimeter DNS

Employee for
Company B

Trust Relationship

Figure 1 Traditional Identity Model for Active Directory Federation

There are a couple of other nice features in Azure AD. Azure

AD provides the convenience of single sign-on (SSO). This lets an
employee access protected resources over time without repeatedly
being forced to log in and provide credentials. The access token is
stored locally as a cookie on the employees device. You can limit
access by using token expiration techniques.
Azure AD also supports OAuth 2.0, perhaps the most popular
open standard for authentication. This provides client applications
secure delegated access to server resources on behalf of a resource
owner. Using OAuth 2.0, Azure AD can authenticate a user and
issue either a SAML 2.0 or JWT token. Read more about how
tokens work at bit.ly/1pDc0rg.

Trading Partners
The traditional way companies have managed authentication across
each others IT infrastructures has been through federation trust.
This involves establishing a trust between the federation services of
the two companies (see Figure 1). This supported secure Web-based
transactions with business partners. However, its not as straightforward as it sounds, and can indeed present challenges.
The partner needs to set up a trust relationship between their
Active Directory and the other enterprises Active Directory. This
lets an employee from one company authenticate through their
own Active Directory and then swap authentication tokens with
the other companys Active Directory in order to gain access to an
app in the other network. This approach requires a lot of coordination and manual workflows for IT administrators.
On-Premises Datacenter
Active Directory
Federation Services

Azure Datacenter
Synchronize
with Directory
Synchronization

Azure Active Directory

On-Premises-Hosted LOB

Cloud-Hosted LOB

VMs

VMs

Web Sites

Web Sites

Databases

Employee

Databases

Figure 2 A Modern Approach to Identity Management

msdnmagazine.com

Figure 2 depicts a more modern approach to identity management. Generally, a copy of employee identities are hosted in Azure,
although some identities may reside in Azure alone. Maintenance
is kept at a minimum because the original copy of most identities
is managed on-premises. A background synchronization process
keeps the cloud copy of identities fully synchronized. Users can
connect directly to Azure AD in a datacenter in the cloud and then
get redirected to the appropriate application.
For mobile app developers, youll find a number of libraries on
GitHub you can use to take advantage of Azure AD (bit.ly/1qmeipz).
Android, iOS, the Microsoft .NET Framework, JavaScript,
Node.js and more are all supported. For example, the iOS sample
shows you how to obtain an interactive authorization code to work
with a work account. You can tie this work account to an Active
Directory server running in your datacenter or live completely in
the cloud with Azure AD. On the back end, you could use either a
REST-based Node.js API or .NET Web API.

Customer Access
You dont usually think of customer-facing sites and services having
access to protected company resources. However, this is more
common than you think. For example, Wells Fargo might want
to expose a customers financial data across the Web via a secure
Web site or mobile app, while Netflix might want to provide secure
access to protected resources such as current account balance or
payment history. This becomes especially important for customers
using mobile devices.
In some scenarios, it even makes sense to leverage social identity
providers, such as Microsoft Account, Facebook, Google or Twitter.
This approach was demonstrated in detail in the November Azure
Insider article (msdn.microsoft.com/magazine/dn818496), in which we wrote
a native iOS application that uses Twitter as the identity provider.
The assumption here is that anyone can download your app or
access the Web site. Theres a publicly available service tier. Prospects
typically have an unauthenticated experience, and see only publicly
available information.Even in this case, there may be some need to
access on-premises resources.Your service could provide an unauthenticated tier of service or respond to a set of default credentials.

Secure Connectivity Tools

Identity aside, there are several tools that can help securely connect
users to resources in enterprise networks. One such example is Azure
Service Bus Relay (see Figure 3), which provides an elegant way to
January 2015 49

Azure Service Bus Relay

Web Service Consumer

HTTP Rest

Customer
Training Partner

Firewall
On-Premises
Web Service

Figure 3 The Azure Service Bus Relay

facilitate access to protected resources. You can think of the Azure

Service Bus Relay as a cloud-hosted service acting as a gatekeeper
to broker the connection between a client and corporate resources.
The Service Bus Relay is authenticated and only opens outbound
ports. It supports Topics and Queues, as well, so you could implement a form of distributed messaging. You could even incorporate
Event Hubs if you needed to ingest large amounts of events.
The advantage of this approach is there are no configuration
changes required to the on-premises networks. That includes
provisioning users or changing firewall settings, because both the
client and back-end service use outbound connectivity to port 80
or 443. In other words, no additional ports need to be opened to
incoming traffic and no proxies are required.

Application Proxy Approach

The second approach to making protected resources available
outside the firewall is to use the Azure AD Application Proxy
service, currently in Preview. This provides secure access to an
organizations internal Web apps and services through HTTP and
HTTPS Web protocols.
The big benefit of this approach is users can enjoy SSO and a native
application experience and still use an unmanaged, non-domain-joined
device. In a previous article, we discussed the concept of a workplace
join, which is a lightweight version of a fully domain-joined device
(bit.ly/1tE7dRR). This supports those BYOD scenarios where users want
to retain control of their personal device.
Azure AD Application Proxy provides tight integration with
Azure AD, supporting SSO, Software-as-a-Service and on-premises
applications, multifactor authentication and device registration.
The connector bridges the resource to the application proxy and,
therefore, the mobile application. These connectors essentially route
traffic and support redundancy, scale and multiple sites.
An Application Proxy provides a few advantages. First, you can
implement fine-grained control at an application, device, user and
location level. Existing client apps dont need modification and
the devices themselves dont need to be modified. It also supports
multifactor authentication (MFA), which can provide an extra level
of protection for highly confidential resources. MFA provides several supporting services, like fraud alerts, IP whitelists, and mobile
phone or SMS as a second authentication factor.
50 msdn magazine

The third advantage of an Application Proxy is it provides isolation between cellular networks and on-premises networks. Back-end
services are never directly exposed because the Application Proxy
itself lives in Azure AD and the connector reaches out directly
from the protected resource. Finally, an Application Proxy protects
against denial-of-service attacks, giving you control over throttling,
queuing and session management.

API Management
A third approach to providing access to protected resources is Azure
API Management, which also behaves as a proxy to your on-premises
resources. This lets you expose Web services implemented within
your firewall to the outside world. To make this option work, youll
need to make some modifications to your on-premises Web server,
as well as create and configure an API in the Azure portal. You can
author your API with either .NET Web API or Node.js.
First, well consider authenticating with Azure API Management
basic authentication to provide outside access to an on-premises
IIS-based Web server. The first step to enabling this scenario is to
make some changes to your on-premises IIS Web server and enable
basic authentication. Within IIS, you can choose the authorization
icon and drill down to basic authentication to add a user. Azure
API Management will access on-premises IIS with this user profile.
The next approach you can use with Azure API Management
is shared secret authentication. This means the API Management
proxy will need to store a secret key. It will place that key in the
HTTP header before trying to connect to the on-premises backend Web service. It should be no surprise the back-end Web service
will need to pull the secret key from the HTTP header and process
the authorization request.
You can implement shared secret authentication by going to the
Azure Management Portal, navigating to the API Management
section and selecting Policies from the menu. Here, youll add
a policy, which is nothing more than a secret key sent to your
on-premises Web server. The policy definition file that youll add
is an XML document.

Hybrid Connections
The fourth and arguably simplest approach to granting access to
protected resources is using Hybrid Connections (see Figure 4), a
feature of the Azure BizTalk Services currently in Preview. Hybrid
Connections leverages Service Bus Relay technology to create a
iOS

Azure Datacenter

Android

Hybrid Connection
(Hostname and Port)

Windows Phone
Web

Corporate Network
Database

Hybrid
Connection
Manager

Figure 4 Microsoft Azure BizTalk ServicesHybrid Connections

Azure Insider

simple, secure and effective connection between on-premises process and an Azure Mobile Services service or an Azure Web Sites
site. You can use Hybrid Connections to connect to any on-premises
resource that uses TCP or HTTP for connectivity, including databases and ERP solutions such as SQL, Oracle and SAP.
A requirement for this is the on-premises process must be
accessible via a static TCP port, such as SQL Server on port 1433. You
can use any framework to connect to the resource, including Node.js,
Java or the .NET Framework. Most of the configuration requirements of
Service Bus Relay are abstracted away by Hybrid Connections. There
are no required changes to the source code of the on-premises process.
To use Hybrid Connections, configure a small amount of metadata in the Azure Management Portal. Install an agent on the server
that hosts your process (SQL Server, MySQL, your Web service and
so on). Theres a Service Bus Relay connection under the covers, using
an SAS key for authentication. You can rotate both service-side and
client-side keys independently from within the portal. Because it uses
Service Bus Relay, no incoming endpoints or proxies are required.
If you prefer more simplicity in your set up and developer efforts,
Hybrid Connections might be your solution.Better yet, because
its available in the free tier of Azure BizTalk Services, you can use
it cost-effectively, only paying for egress bandwidth.
Figure 5 Client Code Leverages Protected
Resources via a Hybrid Connection
// Use the SQL Server connection
try
{
SqlConnectionStringBuilder builder = new SqlConnectionStringBuilder();
builder["Data Source"] = "myserver";
builder["Integrated Security"] = false;
builder["Initial Catalog"] = "mydb";
builder["User ID"] = "greg";
builder["Password"] = "S3cr3t";
int count = 0;
string query = "SELECT COUNT(*) FROM mytable";
using (SqlConnection conn = new SqlConnection(builder.ConnectionString))
using (SqlCommand cmd = new SqlCommand(query, conn))
{
conn.Open();
cmd.CommandType = System.Data.CommandType.Text;
count = Convert.ToInt32(cmd.ExecuteScalar());
}
string results = string.Format("Count of records in mytable is {0}", count);
Services.Log.Info(results);
}
catch (Exception ex)
{
Services.Log.Error(ex.Message);
}
// Use the Web service connection
try
{
const string URL = "http://mydevbox:31106/";
HttpClient client = new HttpClient();
client.BaseAddress = new Uri(URL);
client.DefaultRequestHeaders.Accept.Clear();
client.DefaultRequestHeaders.Accept.Add(
new MediaTypeWithQualityHeaderValue("application/json"));
string response = await client.GetStringAsync("api/values/5");
Services.Log.Info("Accessing the web service succeeded.");
}
catch (Exception ex)
{
Services.Log.Error(ex.InnerException.Message);
}

msdnmagazine.com

Now for a Demonstration

As promised, well now demonstrate some of the techniques outlined here, using a native iOS app plus its AMS back end as the
platform.Well add an on-premises Web service and an on-premises
SQL Server using Hybrid Connections. In the interests of conserving space and utilizing already-existing resources, please refer
to the AMS teams tutorial at bit.ly/1mK1LQF. In the directions youll
find branches to additional tutorial resources ( bit.ly/1vFiPuv and
bit.ly/1xLWuuF) to build out the complete solution, which consists of
the following:
A Web API back end installed as an AMS service and
registered with an Azure AD tenant
A native iOS app registered with the same Azure AD tenant
Bi-directional linkage within the tenant between the Web
API and the AMS service
There are several steps that involve creating metadata in the Azure
Management Portal, while others involve using native development
tools (Visual Studio, Xcode, Android and more) to create the
actual apps.When complete, you should have a working system
that requires you to authenticate into your iOS app by entering
Azure AD credentials. Then it will let you do CRUD operations on
To Do data. Add the code in Figure 5 to the mobile service code.
In the connection information for SQL Server and the addressing of the Web service (a simple template Web service), only the
name of the Hybrid Connection is involved.

Wrapping Up
Identity is becoming increasingly important in the world of BYOD.
Businesses are under increasing pressure to support mobile
devices in the workplace. Its made more complex by the fact that
we live in a hybrid world. Key resources are located in the cloud and
on-premises. Authentication may occur in the cloud or on-premises
or both. Protected resources might include business apps, services
running on-premises or cloud-hosted Web services. With so many
variables, you need multiple flexible approaches.
For full details of this implementation, check the Go Live on
Azure blog post at bit.ly/1rVbtCp, and for a Java-based treatment of
the topic, bit.ly/1zW7XpZ.
Q

GREG OLIVER joined the Microsoft Azure ISV organization in 2010. He spends most
of his time helping companies with their migration plans and implementations.
Most recently, he has been working with a popular consumer games company with
their migration from a competing cloud provider. Prior to joining Microsoft, he
was a technology partner in a startup company.

BRUNO TERKALY is a principal software engineer at Microsoft with the objective

of enabling development of industry leading applications and services across
devices. Hes responsible for driving the top cloud and mobile opportunities across
the United States and beyond from a technology-enablement perspective. He helps
partners bring their applications to market by providing architectural guidance
and deep technical engagement during the ISVs evaluation, development and
deployment. He also works closely with the cloud and mobile engineering groups,
providing feedback and influencing the roadmap.

THANKS to the following Microsoft technical expert for reviewing this article:
Santosh Chandwani
January 2015 51

TYPESCRIPT

Understanding TypeScript
Peter Vogel
In many ways, its useful to think of TypeScript on its own
merits. The TypeScript language specification refers to TypeScipt as
a syntactic sugar for JavaScript. Thats true and probably an essential step in reaching to the languages target audienceclient-side
developers currently using JavaScript.
And you do need to understand JavaScript before you can understand TypeScript. In fact, the language specification (you can read
it at bit.ly/1xH1m5B) often describes TypeScript constructs in terms of
the resulting JavaScript code. But its equally useful to think of TypeScript as a language on its own that shares features with JavaScript.
For example, like C#, TypeScript is a data-typed language, which
gives you IntelliSense support and compile-time checking, among
other features. Like C#, TypeScript includes generic and lambda
expressions (or their equivalent).
But TypeScript, of course, is not C#. Understanding whats
unique about TypeScript is as important as understanding what
TypeScript shares with the server-side language youre currently
This article discusses:
Using TypeScript to develop apps
The similarities and differences of TypeScript and JavaScript
How TypeScript works with JavaScript

Technologies discussed:
JavaScript, C#, TypeScript
52 msdn magazine

using. The TypeScript type system is different (and simpler) than

C#. TypeScript leverages its understanding of other object models
in a unique way and executes inheritance differently than C#. And
because TypeScript compiles to JavaScript, TypeScript shares many
of its fundamentals with JavaScript, unlike C#.
The question then remains, Would you rather write your clientside code in this language or in JavaScript?

TypeScript Is Data-Typed
TypeScript doesnt have many built-in data types you can use to
declare variablesjust string, number and Boolean. Those three
types are a subtype of the any type (which you can also use when
declaring variables). You can set or test variables declared with those
four types against the types null or undefined. You can also declare
methods as void, indicating they dont return a value.
This example declares a variable as string:
var name: string;

You can extend this simple type system with enumerated values
and four kinds of object types: interfaces, classes, arrays and functions.
For example, the following code defines an interface (one kind of
object type) with the name ICustomerShort. The interface includes two
members: a property called Id and a method called CalculateDiscount:
interface ICustomerShort
{
Id: number;
CalculateDiscount(): number;
}

As in C#, you can use interfaces when declaring variables and return
types. This example declares the variable cs as type ICustomerShort:
var cs: ICustomerShort;

You can also define object types as classes, which, unlike interfaces,
can contain executable code. This example defines a class called
CustomerShort with one property and one method:
class CustomerShort
{
FullName: string;
UpdateStatus( status: string ): string
{
...manipulate status...
return status;
}
}

Like more recent versions of C#, its not necessary to provide implementation code when defining a property. The simple declaration of
the name and type is sufficient. Classes can implement one or more
interfaces, as shown in Figure 1, which adds my ICustomerShort
interface, with its property, to my CustomerShort class.

Like C#, TypeScript is a

data-typed language.
As Figure 1 shows, the syntax for implementing an interface
is as simple in TypeScript as in C#. To implement the interfaces
members you simply add members with the same name instead
of tying the interface name to the relevant class members. In this
example, I simply added Id and CalculateDiscount to the class to
implement ICustomerShort. TypeScript also lets you use object
type literals. This code sets the variable cst to an object literal containing one property and one method:
var csl = {
Age: 61,
HaveBirthday(): number
{
return this.Age++;
}
};

This example uses an object type to specify the return value of

the UpdateStatus method:
UpdateStatus( status: string ): { status: string; valid: boolean }
{
return {status: "New",
valid: true
};
}

Figure 1 Add an Interface to a Class

class CustomerShort implements ICustomerShort
{
Id: number;
FullName: string;
UpdateStatus(status: string): string
{
...manipulate status...
return status;
}
CalculateDiscount(): number
{
var discAmount: number;
...calculate discAmount...
return discAmount;
}
}

msdnmagazine.com

Figure 2 This Class Implements the Proper Interface

class CustomerShort implements ICustomerShort
{
Id: number;
FullName: string;
CalculateDiscount( discountedAmount:
( discountClass: string, multipleDiscounts: boolean ) => number ): number
{
var discAmount: number;
...calculate discAmount...
return discAmount;
}
}

Besides object types (class, interface, literal and array), you can
also define function types that describe a functions signature. The
following code rewrites CalculateDiscount from my CustomerShort class to accept a single parameter called discountAmount:
interface ICustomerShort
{
Id: number;
CalculateDiscount( discountAmount:
( discountClass: string, multipleDiscount: boolean ) => number): number
}

That parameter is defined using a function type that accepts two

parameters (one of string, one of boolean) and returns a number. If
youre a C# developer, you might find that the syntax looks much
like a lambda expression.
A class that implements this interface would look something
like Figure 2.
Like the recent versions of C#, TypeScript also infers the datatype of a variable from the value to which the variable is initialized.
In this example, TypeScript will assume the variable myCust is
of CustomerShort:
var myCust= new CustomerShort();
myCust.FullName = "Peter Vogel";

Like C#, you can declare variables using an interface and then
set the variable to an object that implements that interface:
var cs: ICustomerShort;
cs = new CustomerShort();
cs.Id = 11;
cs.FullName = "Peter Vogel";

Finally, you can also use type parameters (which look suspiciously
like generics in C#) to let the invoking code specify the data type
to be used. This example lets the code that creates the class set the
datatype of the Id property:
class CustomerTyped<T>
{
Id: T;
}

This code sets the datatype of the Id property to a string before

using it:
var cst: CustomerTyped<string>;
cst = new CustomerTyped<string>();
cst.Id = "A123";

To isolate classes, interfaces and other public members and avoid

name collisions, you can declare these constructs inside modules
much like C# namespaces. Youll have to flag those items you want
to make available to other modules with the export keyword. The
module in Figure 3 exports two interfaces and a class.
To use the exported components, you can prefix the component
name with the module name as in this example:
var cs: TypeScriptSample.CustomerShort;

January 2015 53

Or you can use the TypeScript import keyword to establish a

shortcut to the module:
import tss = TypeScriptSample;
...
var cs:tss.CustomerShort;

TypeScript Is Flexible About Data Typing

All this should look familiar if youre a C# programmer, except
perhaps the reversal of variable declarations (variable name first,
data type second) and object literals. However, virtually all data
typing in TypeScript is optional. The specification describes the
data types as annotations. If you omit data types (and TypeScript
doesnt infer the data type), data types default to the any type.
TypeScript doesnt require strict datatype matching, either. TypeScript uses what the specification calls structural subtyping to
determine compatibility. This is similar to whats often called duck
typing. In TypeScript, two classes are considered identical if they
have members with the same types. For example, heres a CustomerShort class that implements an interface called ICustomerShort:
interface ICustomerShort
{
Id: number;
FullName: string;
}
class CustomerShort implements ICustomerShort
{
Id: number;
FullName: string;
}

Heres a class called CustomerDeviant that looks similar to my

CustomerShort class:
class CustomerDeviant
{
Id: number;
FullName: string;
}

Virtually all data typing in

TypeScript is optional.
Thanks to structural subtyping, I can use CustomerDevient with
variables defined with my CustomerShort class or ICustomerShort
interface. These examples use CustomerDeviant interchangeably
with variables declared as CustomerShort or ICustomerShort:
var cs: CustomerShort;
cs = new CustomerDeviant
cs.Id = 11;
var csi: ICustomerShort;
csi = new CustomerDeviant
csi.FullName = "Peter Vogel";

This flexibility lets you assign TypeScript object literals to variables declared as classes or interfaces, provided theyre structurally
compatible, as they are here:
var cs: CustomerShort;
cs = {Id: 2,
FullName: "Peter Vogel"
}
var csi: ICustomerShort;
csi = {Id: 2,
FullName: "Peter Vogel"
}

54 msdn magazine

Figure 3 Export Two Interfaces and One Class

module TypeScriptSample
{
export interface ICustomerDTO
{
Id: number;
}
export interface ICustomerShort extends ICustomerDTO
{
FullName: string;
}
export class CustomerShort implements ICustomerShort
{
Id: number;
FullName: string;
}

This leads into TypeScript-specific features around apparent

types, supertypes and subtypes leading to the general issue of
assignability, which Ill skip here. Those features would allow
CustomerDeviant, for example, to have members that arent present in CustomerShort without causing my sample code to fail.

TypeScript Has Class

The TypeScript specification refers to the language as implementing the class pattern [using] prototype chains to implement
many variations on object-oriented inheritance mechanisms. In
practice, it means TypeScript isnt only data-typed, but effectively
object-oriented.
In the same way that a C# interface can inherit from a base
interface, a TypeScript interface can extend another interface
even if that other interface is defined in a different module. This
example extends the ICustomerShort interface to create a new
interface called ICustomerLong:
interface ICustomerShort
{
Id: number;
}
interface ICustomerLong extends ICustomerShort
{
FullName: string;
}

The ICustomerLong interface will have two members: FullName

and Id. In the merged interface, the members from the interface
appear first. Therefore, my ICustomerLong interface is equivalent
to this interface:
interface ICustomerLongPseudo
{
FullName: string;
Id: number;
}

A class that implements ICustomerLong would need both properties:

class CustomerLong implements ICustomerLong
{
Id: number;
FullName: string;
}

Classes can extend other classes in the same way one interface can
extend another. The class in Figure 4 extends CustomerShort and
adds a new property to the definition. It uses explicit getters and setters
to define the properties (although not in a particularly useful way).
TypeScript enforces the best practice of accessing internal fields
(like id and fullName) through a reference to the class (this). Classes
TypeScript

can also have constructor functions that include a feature C# has

just adopted: automatic definition of fields. The constructor function in a TypeScript class must be named constructor and its public
parameters are automatically defined as properties and initialized
from the values passed to them. In this example, the constructor
accepts a single parameter called Company of type string:
export class CustomerShort implements ICustomerShort
{
constructor(public Company: string)
{
}

Figure 4 Properties Defined with Getters and Setters

class CustomerShort
{
Id: number;
}
class CustomerLong extends CustomerLong
{
private id: number;
private fullName: string;
get Id(): number
{
return this.id
}
set Id( value: number )
{
this.id = value;
}
get FullName(): string
{
return this.fullName;
}
set FullName( value: string )
{
this.fullName = value;
}

Because the Company parameter is defined as public, the class

also gets a public property called Company initialized from the
value passed to the constructor. Thanks to that feature, the variable
comp will be set to PH&VIS, as in this example:
var css: CustomerShort;
css = new CustomerShort( "PH&VIS" );
var comp = css.Company;

Declaring a constructors parameter as private creates an internal property it can only be accessed from code inside members of
the class through the keyword this. If the parameter isnt declared
as public or private, no property is generated.

TypeScript isnt only data-typed,

but effectively object-oriented.
Your class must have a constructor. As in C#, if you dont provide one, one will be provided for you. If your class extends another
class, any constructor you create must include a call to super.
This calls the constructor on the class its extending. This example
includes a constructor with a super call that provides parameters
to the base class constructor:
class MyBaseClass
{
constructor(public x: number, public y: number ) { }
}
class MyDerivedClass extends MyBaseClass
{
constructor()
{
super(2,1);
}
}

TypeScript Inherits Differently

Again, this will all look familiar to you if youre a C# programmer,
except for some funny keywords (extends). But, again, extending
a class or an interface isnt quite the same thing as the inheritance
mechanisms in C#. The TypeScript specification uses the usual
terms for the class being extended (base class) and the class that
extends it (derived class). However, the specification refers to
a class heritage specification, for example, instead of using the
word inheritance.
To begin with, TypeScript has fewer options than C# when it
comes to defining base classes. You cant declare the class or members
as non-overrideable, abstract or virtual (though interfaces provide
much of the functionality that a virtual base class provides).
Theres no way to prevent some members from not being inherited.
A derived class inherits all members of the base class, including public
msdnmagazine.com

and private members (all public members of the base class are
overrideable while private members are not). To override a public
member, simply define a member in the derived class with the same
signature. While you can use the super keyword to access a public
method from a derived class, you cant access a property in the base
class using super (though you can override the property).
TypeScript lets you augment an interface by simply declaring an
interface with an identical name and new members. This lets you
extend existing JavaScript code without creating a new named type.
The example in Figure 5 defines the ICustomerMerge interface
through two separate interface definitions and then implements
the interface in a class.
Classes can also extend other classes, but not interfaces. In
TypeScript, interfaces can also extend classes, but only in a way that
involves inheritance. When an interface extends a class, the interface includes all class members (public and private), but without
the class implementations. In Figure 6, the ICustomer interface
will have the private member id, public member Id and the public
member MiddleName.
The ICustomer interface has a significant restrictionyou can
only use it with classes that extend the same class the interface
Figure 5 The ICustomerMerge Interface Defined Through
Two Interface Definitions
interface ICustomerMerge
{
MiddleName: string;
}
interface ICustomerMerge
{
Id: number;
}
class CustomerMerge implements ICustomerMerge
{
Id: number;
MiddleName: string;
}

January 2015 55

extended (in this case, thats the Customer class). TypeScript requires
that you include private members in the interface to be inherited from the class that the interface extends, instead of being
reimplemented in the derived class. A new class that uses the
ICustomer interface would need, for example, to provide an
implementation for MiddleName (because its only specified in the
interface). The developer using ICustomer could choose to either
inherit or override public methods from the Customer class, but
wouldnt be able to override the private id member.

Classes can extend other classes

in the same way one interface
can extend another.
This example shows a class (called NewCustomer) that implements the ICustomer interface and extends the Customer class
as required. In this example, NewCustomer inherits the implementation of Id from Customer and provides an implementation
for MiddleName:
class NewCustomer extends Customer implements ICustomer
{
MiddleName: string;
}

This combination of interfaces, classes, implementation and extension provides a controlled way for classes you define to extend classes
defined in other object models (for more details, check out section
7.3 of the language specification, Interfaces Extending Classes).
Coupled with the ability of TypeScript to use information about
other JavaScript libraries, it lets you write TypeScript code that
works with the objects defined in those libraries.

TypeScript Knows About Your Libraries

Besides knowing about the classes and interfaces defined in your
application, you can provide TypeScript with information about
other object libraries. Thats handled through the TypeScript declare
keyword. This creates what the specification calls ambient declarations. You many never have to use the declare keyword yourself
because you can find definition files for most JavaScript libraries
Figure 6 An Extended Class with All Members
class Customer
{
private id: number;
get Id(): number
{
return this.id
}
set Id( value: number )
{
this.id = value;
}
}
interface ICustomer extends Customer
{
MiddleName: string;
}

56 msdn magazine

on the DefinitelyTyped site at definitelytyped.org. Through these definition files, TypeScript can effectively read the documentation
about the libraries with which you need to work.
Reading the documentation, of course, means you get datatyped IntelliSense support and compile-time checking when using
the objects that make up the library. It also lets TypeScript, under
certain circumstances, infer the type of a variable from the context in which its used. Thanks to the lib.d.ts definition file included
with TypeScript, TypeScript assumes the variable anchor is of type
HTMLAnchorElement in the following code:
var anchor = document.createElement( "a" );

The definition file specifies thats the result returned by the

createElement method when the method is passed the string a.
Knowing anchor is an HTMLAnchorElement means TypeScript
knows the anchor variable will support, for example, the addEventListener method.
The TypeScript data type inference also works with parameter
types. For example, the addEventListener method accepts two
parameters. The second is a function in which addEventListener
passes an object of type PointerEvent. TypeScript knows that and
supports accessing the cancelBubble property of the PointerEvent
class within the function:
span.addEventListener("pointerenter", function ( e )
{
e.cancelBubble = true;
}

In the same way that lib.d.ts provides information about the

HTML DOM, the definition files for other JavaScript provide
similar functionality. After adding the backbone.d.ts file to my project, for example, I can declare a class that extends the Backbone
Model class and implements my own interface with code like this:
class CustomerShort extends bb.Model implements ICustomerShort
{
}

If youre interested in details on how to use TypeScript with

Backbone and Knockout, check out my Practical TypeScript columns at bit.ly/1BRh8NJ. In the new year, Ill be looking at the details
of using TypeScript with Angular.
Theres more to TypeScript than you see here. TypeScript version
1.3 is slated to include union datatypes (to support, for example,
functions that return a list of specific types) and tuples. The TypeScript team is working with other teams applying data typing to
JavaScript (Flow and Angular) to ensure TypeScript will work with
as broad a range of JavaScript libraries as possible.
If you need to do something that JavaScript supports and TypeScript wont let you do, you can always integrate your JavaScript
code because TypeScript is a superset of JavaScript. So the question
remainswhich of these languages would you prefer to use to write
your client-side code?
Q
PETER VOGEL is a principal with PH&V Information Services, specializing in Web
development with expertise in SOA, client-side development and UI design. PH&V
clients include the Canadian Imperial Bank of Commerce, Volvo and Microsoft.
He also teaches and writes courses for Learning Tree International and writes the
Practical .NET column for VisualStudioMagazine.com.

THANKS to the following Microsoft technical expert for reviewing this article:
Ryan Cavanaugh
TypeScript

ement1); areaSerie
areaSeries
ies.A
Add(seriesElement2);
A
dd(se
(s riesElement2
t2);
) areaSeries
areaSeries.Ad
Add(seriesElement3);
A d(seriesElement3); // Add series to the plot area plotArea
plotArea.Series.Add(areaSeries);
Series Add(areaSeries); //page
//page.Elemen
Elements
em ts.Add(
Add(
dd( new
ne LayoutGrid()
LayyoutGrid() ); /// Add
La
Add the page elements to the page AddEA
s, 240, 0); AddEAN1
AN 3Sup
Sup5(pa
5(p ge.Eleme
ent
nts, 480, 0); Ad
ddUPCVersionA(page.Elemen
e ts, 0, 135); AddUPCVersionASup2(page.Elements, 240, 135); Ad
dd
dUPC
UPC
CVers
sionA
o Sup5((page.Elemen
nts,
t 480, 135); AddEAN8(page.Elements, 0,
.Elements, 480, 270);; Add
ddU
UPCVersionE(pa
page.Elementts, 0, 405); AddUPCVersionESu
E p2(page.Elements, 240, 405); AddUPCVersionESup5(page
age.Ele
.Ele
lem
ments,
s, 48
480, 405); // Add the page
e tto the document document.Pages.Add(pa
CaptionAndRectang
a le
e(eleme
ements, EAN/JA
/JAN 13 Bar Cod
de, x, y, 204, 99); BarCode barCode = new Ean13(123456789012, x, y + 21); barCode.
ode.X
X+
+= (20
04 - ba
arCo
ode.GettSymbolWidth()
h ) / 2; elements.Add(barCode); } private vo
dRectangle(element
e s,, EAN
EAN/JAN 13 Bar
ar Code,
C
2 digit supplement, x, y, 204, 99); BarCode barCode = new Ean13Sup2(1
2 2345
56789
678 012
1212,
2, x, y + 21
21);
1); barCo
ode.X
X += (204 - barCode.GetSymbolWidth()) / 2; elements.Add((barC
ts, oat x, oat y) { AddCa
A CaptionAndRectan
angle(elements, EAN/JAN 13 Bar Code, 5 digit supplement, x, y, 204, 99); BarCo
a C de barCo
arC
Code
de = new
n
Ean13Su
Ean1
S p5(12
2345
567890
01212345, x, y + 21); barCode.X += (204 - barCode
e.Get
ddUPCVersionA(Grou
Group elemen
em ts, oat x, oat
oa y) { AddCaptionAndRectangle(element
e s, UPC Version A Bar Code, x, y, 204,
2
99);
9) Ba
B
BarrCode
C
barC
bar
Code = ne
ew UpcVe
pcVersionA
A(12345678901, x, y + 21); barCode.X += (204 - ba
arCo
ddUPCVersionAS
Sup2(
up2 Grou
oup
p elements,, oat
oa x, oat y) { AddCaptionAndRectangle(ele
( ments, UPC Version E Bar Code, 2 digit
git sup
supp
up lement
nt, x,
x y, 204, 99
9); Ba
arCod
de barC
Code = new UpcVersionASup2(12345678
7 90112, x,
x, y +
s.Add(barCode); } private
te vo
oid AddUPCV
PCVersi
ers onASup5(Group elements, oat x, oat
o y) { AddCaptionAndRectangle(ele
em
ment
ents, UPC
UPC Ver
Version E Bar Code, 5 dig
git su
upplem
ment, x, y, 204, 99); BarCode barCode = new UpcVe
ersio
ode.GetSymbolW
Width(
dth )) / 2
2; elements.Add
Add(bar
(ba Code); } private
e voi
v d AddEAN
EAN8(Group
p elements, oat x, oat y) { AddC
ddCapti
tionAn
onA
n dRec
Rec
ect
ctangle(elem
ments, EAN
N/JAN
N 8 Bar Cod
de, x, y, 204, 99); BarCode barCode = new
n
Ean8(1234
34
g(); leDialog.Title
le = Op
Open
en File
F Dialog
g; l
eDialog.Filter = Ad
Adobe PDF
F les
es (*.pdf)
f)|*.pdf|All Files (*.*)|*.*; if (leDi
eDialog.
log.Sh
Show
wDiallog()
og == Dialog
gResult.OK) { pdfViewe
ewer.Op
penFile(leDialog.FileName, ); } Save
Sav File
F Diallog sa
av
aveF
File Dialog; ssaveF
veFileDialo
al g.Filter = Ado
Adobe
e PDF les (*.pdf)
f)|*.pdf|All Files (**.*)|*.*; if (saveFileDialog.ShowD
owDialo
ialo
a g()=
g()==Dia
=Dia
Di logResul
esult

.O
OK) { pdfV
fViewe
ewerr.SaveAs(saveFileDia
Dialog.FileName
e); } if (p
( dfVi
dfV ewe
wer.P
Page

WithDialog(); } else
e se { Mess
Me ageBox.S
Show(
w P
Please open a le to printt); } OpenFile
F Dialog leDi
D alog
log = ne
ew Ope
penFileD
pen
Dialog(); le Dialog.Tiitle = Open File Dialo
og; lleDialog.InitialDirec
ecto
ory = @c:\
:\ ;
leD
leDialog
lo .Filter
ter = All
F
) == DialogResul
es t.O
t.OK) { Dy
D namicPDF
FView
ewerrClass test = new
ew Dynam
D
micPDFVie
ewerC
r lass(); PDF
DFPrin
Printter prin
inter
nter
er = test.OpenF
pe ileForP
orPrin
in
ntter

(leDia
(le
alog.FileName); prin
nter.PrintQuie
Quiet();
() } byt
by e[] cont
contents
t =

pServices; GCH
GC and
andle gc
ch = GCHandle
d .All
Al oc
c(contents, GCH
Hand
ndleTyp
Type.Pinned
d); Int
IntPtr cont
contents
entsIntPtr =gc
ch.
ch.
h.A
AddrOfPinned
nn Obje
ect()
ct ;p
pd
df

Viewer.O
.OpenB
pe ufffe
fer(co
r( ntentsIntPtr
t ,

kmark Page Eleme

lement:, x, y); p
pageEle
ement
en s.A
Add(new Book
kmarrk(
( Bookmarked Text
x , x + 5, y + 20,
0 par
pare
ent
e
ntOutline)); pageEl
g emen
nts.A
ts.Add

(new Label(Thiss tex

text is bookma
okmarked
okma
rked.,
, x + 5, y + 20, 2

ageElements,
s, oat
a x, oat
at y) { // Adds
dss a circltt to the pageElleme
ents AddCaptio
onAndR
AndRecta
ectangle(pag
pag
geEle
eEl ment
men s, Circle Page Elem
ment:
ent:,
t:: x,,

y); pageEl
geElements.A
s.Add(n
dd(new Circle(x
dd
(x + 112.5f
2 ,

shLarge)); } priva
vate
te void AddF
Ad or
ormatted
te
edTextArrea(Group
p page
geElemen
nts, oat x, oa
oatt y)
y { // A
Adds a fo
for
o mattted
d text area to
o thepageE
eElle ments strring formatt
m edHt
edHtml = <p><
<p <i>Dynamic</i><b>PD
b>P F</b
b>&tm
tm;
m Genera
era
ator
o v6.0 for
or
or .NE
matting suppo
ort for
or text th
that appearss in the document.
t. You
Y u havve + com
complet
ple
ete co
ontro
rol ovve
err 8 par
aragraph
ph properties:
p
spacing befo
efore
e, spacing
g after, r
rst liine + indenta
ntation, left indentati
tation, righ
r t in
ndent
ntatio
tion
n, aliignment, al
alllowi
<font face=Tim
Times
es>fontt fac
f e, </fon
nt>
><font poin
o tSiz
Size=6>ffont + size, </fon
</fon
nt><fo
<f ntt co
color
olo
llor=FF0000
00 >
>colo
or, </font>
><b>b
<b> old, </b
</b><
<i>italic and </i><
<u>u
underline</u>
>; + and 2 line pro
opert
rties: lea
eading
ng, an
nd le
ead
ding type. Text
extArea = ne
ew Format
Form tedT
dTextA
Area(for
orrmattedH
Htm
tml, x + 5, y + 20,, 215
5, 60,, F
Font
ontFamil
mi y.He
He
elvet
lv
vetica
e ica, 9, false)
e); // Sets the indent prope
roperty fo
formatte
att dTextAre
ea.Styyle.P
Paragrap
ph.In
ndent = 18; AddC
CapttionAndRect
Rectangl
gle(pa
e(pa
e
(pa
pa
ageE
geElements,
ts, F
ageEleme
m ntts, Fo
Form
mattedT
dTextA
tArea Overow
ow Text:, x + 279
9, y); pag
geEle
ement
men s.Ad
Add(fo
Ad
d(fo
d
ormat
orm
rmat
atttedTextA
tArea)); // Create
ea
an overow formatted
ed ttextt a
area forr the
t over
ow textt Fo
ormattedTextArea
a ove
overowFor
Formatt
ma edTe
Text
xtAr
tArea
ea = formatte
e
a(x + 284, y + 20);
20) p
pageE
Elemen
ements.Ad
dd(o
(over
verow
ow
wForm
wFo
rmat
a tedTe
extArea); } priv
private
ate void A
Add
dImag
mage(Group
mag
up pag
geElem
eme
ents, oat x, oat y) { /// Adds an
n image
e to th
he pa
ageElem
men
ents AddCaptionAndRe
dRectangle((pag
pageElem
ments
en
nts Image
nts,
ge
e Pag
es/DPDFLo
ogo.pn
.png), x + 112.5
5f, y + 50f,
50f, 0.24
0.24f
4f); // Image is size
ed an
nd cente
entered
d in
n the rec
ecta
tangle ima
m ge.SetBo
B unds(215, 60); image.VAlign = VAlign.Cen
enterr; ima
age.Align
n = Align.Center; pag
geE
eElements.Ad
.Add
d(imag
ge); } priv
vate void A
pageElemen
nts Ad
AddCap
dC tion
onAndRecta
angle(pag
ngle(pag
geElements,
eEle
L
Labell & PageNumbe
Page
eringL
erin
g bel Page E
gLab
Elem
ementts:, x, y); string labelText
T
= Labels can be rottated; strring number
mbe Text = PageNum
mbe
eringLabels
els cont
contai
ain page num
mb
be
ering
xt, x + 5, y + 12,
12 22
220, 80,
0 Fontt.Time
messRom
Roman,
an, 12, TextAlign..Cen
nter);; lab
label.Ang
Ang
gle = 8; Page
eNumb
Num erin
ri gLa
gLabel pageNumLabel = new PageNumber
b ingLabel
ab (n
numb
berText, x + 5, y + 55, 220, 80, Font.TimesR
esRoman, 12,
12 TextAl
tAlig
ign.
ign
n Ce
m nts.Add(labe
me
abel);
l); } private
e void AddL
dLin
ne(G p pageElemen
ne(Group
nts, oat x, oa
oat y)) { /// Addss a lin
ne to the
he pageE
pag lements AddCaptionAndRectangle(p
( ageEleme
e nts, Line Pa
age Element:, x, y); page
eElemen
nts.A
s.Add(ne
ew Line(x
x+5
5, y +
w Line(x
x + 220,
2
y + 20, x + 5, y + 80, 3,
3, RgbCo
Rg olor.Green)); } pr
priv
iva
ate void
d AddLi
A
ink
nk(Group
up p
pag
pa eElement
ments, oat x, oat y) { // Adds a link to the pageEleme
em ntts Fo
ont font = Fo
ont.TimesRoman;; st
string text = T
This iss a link
k to Dyna
amic
m :, x, y); Label label = ne
ment
new Lab
Label(textt, x + 5, y + 20, 215
5, 80
5,
0, fon
nt, 12, Rgb
R bColor.
or.B
Blu
lue);; label.Und
Under
erline = true; Link link = new Link(x + 5, y + 20, font.
on GetTe
extWidth
h(tex
xt, 12), 12 - font.G
GetD
Descend
der(1
r(12), ne
ew Url
UrlA
lAction(
n(h
http
E men
Ele
nts.Add(li
( nk);; } priva
p vate
e vvoid
d AddPath(
ath Grroup pageElem
mentts,
oat
oat x, oatt y) { // Ad
Adds a path
h to the pageElements ceTe.DynamicPDF.PageElement
en s.Path
h path = new
w ceTe.DynamicPD
PDF.P
F.PageElemen
men
nts.P
s.Pa
ath(
h(x
x + 5, y + 20,
2 R
P s.A
Path
Add(new Line
eSubPat
Pa h(x
x + 215, y + 40))); path.Su
h.S bPatths.A
Add
dd((new Curv
urve
eToSubPat
Pa h(x + 10
08, y + 80, x + 160, y + 80)); path.SubPaths.Add(new
w Curv
veSu
ubPath(x + 5, y + 40, x + 65,
6 y + 80, x + 5, y + 60))); Add
AddC
Ca ionA
Capt
And
Add(p
path); } private
e void AddR
Rec
ctangle(Gr
G oup
p page
eElemen
nts, oat
oat x, oat
at y)
y) ordere
e dLis
dL t = ord
deredList.GetOverFlowList(x + 5, y + 20); AddCaptionAn
AndR
Recta
ang
gle(pag
ge.Elements, Order
r ed List
L Page
ge
e Ele
El ment
nt Ove
Ove
v row
r :, x, y, 2
8; //
8;
// Create
C
an uno
ordere
ed list Unordered
nor e List unorder
e edList
er
stt = ne
ew Uno
norder
rderedL
edList(x
x + 5, y + 20,
20 400, 90, Font.Helvetica, 10); unorderedList.Items.Add(
Add(Fruits); unordered
ere List.Items.Add(
d Vege
Vege
g table
es); U
Unorde
r eredS
re
edS
d ub
bList unord
t();
(); unorderedSubList.Items.
ms.Add(
dd(( Citrus); unord
ordered
eredSu
edSubLiist.Ite
emss.Ad
Add
d( Non
n-Citr
t us)
s ; Ad
AddC
CaptionAndRectangle(page.Elemen
nts, Unordered Lis
st Page
e Elem
me
ent:, x, y,
y 225, 110); Uno
n rd
dere
edSubLis
bLis
st u
un
norde
deredS
redS
d ubLi
ub st2 = uno
erredSub
bList2.Items.Add((Po
Potato); unorderedS
SubLis
ubLiist2.Item
ms.Ad
dd(B
Beans); Unor
no dere
deredSub
dSubLisst subUnorderedSubList = unordered
e SubL
S
ist.Items[0]].Su
ubLists.A
AddUnorder
rde edSubList(); ssub
bUnor
UnorderedSub
dS
Sub
bList.Ite
te
ems.A
m Add
d(Lime); s

;<M<CFG<;=FI
@EKL@K@M<LJ<

List sub
bUnorderedS
SubList2
st = unorderedSubL
S bList.Items
ist.
s[1].Su
ubLis
sts.A
AddUnorde
eredSu
edS bLis
bL t();
t() su
ubUnorderedSubList2.Items.Add(Man
a go); subUnorrdere
edS
SubList2.I
t2 tem
ms.A
Add(Banana);
); Un
Unord
deredS
dS
Su Lis
Sub
List
s sub
ubUnor
n dere
de dSubList
t();
() subU
Unordered
dSubList3.Items.Add(Swe
Swee
w t Po
Potato); Uno
order
d redS
SubList sub
bUnor
Uno dere
er dSub
dSubList
List4
4 = unorrderedSubList2.It
2 ems[1].S
].SubLists.AddUno
orde
ered
dSub
bList(
s ); subUn
ubU orderedSubLi
ubList
b st4.
s Ite
ems.Ad
A d(S
Ad
Strin
ing Be
ean)
an ; subUno
U rde
Add(Kid
dney Bean
ean); x += 279; pag
a e.Eleme
ements.Ad
dd(u
d nordere
edLisst); unorderedList
Lis = unord
nordered
ere List.GetOver
e FlowList(x + 5, y + 20);
) Add
A CaptionAndRec
ctan
ngle(p
page
ag .Eleme
ents, Unor
Unordere
dere
r d List Page
e Elem
ment Ove
ve ow:
ver
o , x, y, 225
oid
d Add
dTextField(Group pageElemen
dTextF
me ts,, o
oat x,, o
oat y)) { Tex
xtField txtt = new TextFi
xtF eld(txt
t fna
fname, x + 20, y + 40, 120, 20); txt.Defaul
u tValue = This is a Scrrollab
ble
e TextF
Field
d; txt.Borde
derCol
rCol
C o
or = Rgb
RgbC
Colo
or.B
r.Black; txt
xt ackgro
xt.Bac
kgroundC
un o
d(txt); T
d(
TextField
ex
txt1 = new TextFi
Field(
ld tx
xtf1na
f1 me,
me x + 175, y + 40, 120, 20); txtt1.Def
De ault
u Valu
alue = TextField; txt1.Password = true; txt1.MaxLength = 9; txtt1.Bo
ord
derColor = RgbCollor.Black;
or.B
B
txt1.B
1.Backg
ckgroun
ou dCol
olor
or = Rgb
R Colo
Colorr.Al
Al
e ies(); pieSeries.DataLabel = da
er
da;
a; plo
p tAre
Area
a.Series
rie .Add(pieS
Series
s); pieSeries.Eleme
lementss.Add(
Add(27,
27, Website A); pieSeries.Elements.Add
d (19, Website B)); pieSerries
es.Element
men s.Add(21
d(21, W
Web
ebsi
s te
e C);
); pieS
p eries.El
Elemen
ements[0
me ts[0
s[0].Co
s[0
].C lor
or = a
ess.Elements[2].Color = aut
utogra
og dien
dientt3;RgbC
RgbColo
or.Alice
eBlue; txt2.Too
.ToolTip = Multilin
ne; page
pageEl
Elements.Add(txt2); AddCaptionAndRectangle(pageEleme
men
nts, Tex
xtFi
Field Form
orm Pag
Page
e Elemen
Elemen
nt:,
:, x, yy,, 504
04, 85);
5) } priva
ate
e void
oid AddC
dComb
omb
C
Comb
oBox(cm
mbnam
bna e, x + 51, y + 40, 150,
15 20); cb.B
BorderColo
or = RgbColor.Blac
Black; cb.Ba
b.Back
ckgroundColor = RgbColor.AliceBlue; cb.Font = Font.Helve
elvetica
a; cb
b.Fon
Fo tSiz
Sizze = 12; cb.I
cb.Items
temss A (Item
tems.Add
e 1);
em
); cb.
cb.Items
tems.Add
.Add(It
Ad (It
( Item
em 2); cb
d table);
di
) cb.Item
ms[
s[
[Edita
able].Select
cted
ed = true;; c
cb.Editable = tru
ue; cb.ToolTip = Edi
Ed tabl
ab e Co
C mbo Box; pageElements.Add(cb); ComboBox cb1 = new
ew Combo
omb
b Box(
B x(cmb
mb1nam
me, x + 303,
303, y + 40, 150, 20);
20); cb1.B
cb1 Borde
rder
derColor = R
= Font.H
Helvetic
ca; cb
cb1.FontS
nt ize = 12
2; cb1.Itemss.A
Add(IItem 1); cb1.Ittems.Add(It
Item
em 2);
) cb1.
cb1.It
Items.Add(Item 3); cb1.Items.Add(Item 4); cb1.Itemss.A
Add
d(No
on-Edi
ditab
table
e);
); c
cb1.Items[
[NonNon-Edit
Editable
able].S
].Se
elected = true; cb1.
1 Edita
nts.Ad
nt
s
d(cb
(cb1); Converter.Co
C nvert(http://www.go
goog
ogle.c
com, Outputt.pdf);Convert
ve er.C
er.Conve
onvert(GetDocPath(DocumentA.rtf), Output.pdf);System.Dia
iagno
ostics
css.Pro
P ocess
ess.S
s Start(Outp
tput.p
ut.pdf)
pdf);; AsyncC
As ncConverter
rt aCo
Co
onverte
nve er = new A
err(aCo
(aC nverrter_Converted); aConverter.ConversionErro
or += new Con
nversionErrorEv
ventH
tHandler(aConverter_ConversionError); aConverter.Convert(@C:\t
C:\ em
mp\Docum
mp
mentA.rtf, @C:\tem
ment
em
mp\Ou
p\Output
tputA.pd
A.pdf);
f);
) aConver
v rter.
ter.Co
onvert(@C
ve t(@C:\temp\DocumentC.rtf, @C:\temp\OutputC
ver
C.pdf)); aCo
onve
erter.Convert(
e h
http://
p://www.yahoo.com, @C:\Temp\yahoo.pdf); ConversionOptions
ion op
option
ns = new Conversio
ns
sionOpt
nOpttions
ions(720
(720,, 720,
72 72, true); ce
eTe.D
Te. ynamicPDF
te \\o
temp
output.pdf, options); ceTe.DynamicPDF.Conve
ersion.Con
nvertter.Convert(C:\\\te
emp\\Document2.docx, C:\\temp\\output.pdf, options); string
g sam
ampl
mp eHtm
mp
H mll = <ht
h ml><
ml><body
body><p>
><p>
p This is a very simpl
m e HTML
ML strring includ
<t le border=\1\>
<tab
1 <tr><td>100</td><td>200</td> + <ttd>3
300<
</td></tr><tr><
<td>
>400</td><td>500</td><td>600</t
< d></tr></table><
></bod
/body><
y></
</h
</
/htm
ht >;Conve
html
o version.Co
n.Co
C nver
nverter.
ter.C
ConvvertHtmlString(sa
ample
mpleHtmll, C:\\
C \temp\
emp\\Sam
\Sam
erName
er
e, Path.Comb
o ine(GetPath(), LetterPortrait.pdff)); prrintJo
ob.D
DocumentName
e = Lett
Letter
e Portrait; if (printJob.P
ob. rinter.Color)
r
pri tJob.PrintOp
prin
P ntO
n
tions.Co
s. lor
o = true;
true; if (prin
(prin
ri tJob
tJo .Printer.Col
C late) printJob.P
b.P
.P
PrintOpti
Opti
p ons.
ons Co
on
ollat = tru
ollate

;peXd`ZG;=:fdgi\_\ej`m\G;=Jfclk`fej]fi%E<K;\m\cfg\ij
Z\K\ Jf]knXi\j ;peXd`ZG;= gif[lZkj gifm`[\ i\Xc$k`d\ G;= ^\e\iXk`fe# dXe`glcXk`fe# Zfem\ij`fe#
gi`ek`e^#m`\n`e^#Xe[dlZ_dfi\%Gifm`[`e^k_\Y\jkf]Yfk_nfic[j#k_\fYa\Zkdf[\cjXi\\oki\d\cp
\o`Yc\ Ylk jk`cc jlggcp k_\ i`Z_ ]\Xkli\j pfl e\\[ Xj X [\m\cfg\i% I\c`XYc\ Xe[ \]Z`\ek# k_\ _`^_$
g\i]fidXeZ\ jf]knXi\ `j \Xjp kf c\Xie Xe[ lj\% @] pfl [f \eZflek\i X hl\jk`fe n`k_ Xep f] fli
Zfdgfe\ekj#j`dgcpZfekXZkZ\K\Jf]knXi\ji\X[`cpXmX`cXYc\#`e[ljkip$c\X[`e^jlggfikk\Xd%

NNN%;PE8D@:G;=%:FD

KIPFLIG;=JFCLK@FEJ=I<<KF;8P
K
nnn%;peXd`ZG;=%Zfd&\mXcfiZXcc/''%-*(%,''-s"(+('%..)%/-)'
n

Untitled-1 1

9/8/11 12:41 PM

WINDOWS PHONE 8.1

Adding Application Data

to the Windows Phone 8.1
Map Control
Keith Pijanowski
In my November article on the mapping capabilities
of Windows Phone 8.1, I introduced the Windows Phone 8.1 Map
control and the Mapping Services API. I showed how to use the
Map control to add mapping functionality to a phone application,
for example, adding images to mark a specified location, calculating
the geo-coordinates of a specified address (geocoding), determining the address of a specified geo-coordinate (reverse geocoding),
and providing driving and walking directions.
This article will show how to get application data into the Map control by adding controls to the Map control using XAML, and binding
application data to those controls. If your application will be used
offline, as well, the underlying data thats used by the Map control
This article discusses:
Adding controls to the Map control using XAML
Binding application data to controls
Scaling controls
Downloading maps for offline use
Updating downloaded maps

Technologies discussed:
Windows Phone 8.1, Visual Studio 2013 Update 3

Code download available at:

msdn.microsoft.com/magazine/msdnmag0115
58 msdn magazine

can also be downloaded for offline use. This allows the map control
to fit in with the offline functionality of your app. Ill also show how
to take the Map control offline and keep the offline data up-to-date.

Adding Controls to the Map Control Using XAML

In my previous article, I used code to add an ellipse control to a
map in order to circle a specific location on the map. Thats a good
approach if you need complete control. However, if you wish to get
creative and annotate a map with many different types of controls,
using code can be clunky. Adding controls to the map using XAML
is more efficient. Also, the XAML approach is much easier when you
need to map a collection of locations within a map using data binding.
First Ill show how to add controls using XAML and how to
bind application data to these controls. Then Ill delve further into
data binding and show how to data bind a collection of locations
to the Map control.
Adding Basic XAML Controls There are two ways to add controls to the Map control in order to add application data to a map:
You can add the controls as children of the Map control or you can
use the MapItemsControl to contain the controls.
Figure 1 shows how to add controls as children of the Map
control. Because Children is the default content property of the
Map control, it doesnt need to be explicitly specified in the XAML
markup. Figure 2 shows an alternate approach to adding controls
to the Map control, which makes use of the MapItemControl to
contain added controls.

If you use the MapItemControl technique shown in Figure 2, be

aware you wont be able to access the controls in your codebehind.
IntelliSense will acknowledge your control names as valid variables, but at run time these variables will always be null and if you
reference them youll get a NullReferenceException. The MapItemControl is typically used to contain data templates that are used
for binding a collection of objects to the Map control. (This will be
discussed later in this article.) Therefore, if youre not binding to a
collection of objects, its better to add your controls as children of
the Map control, as shown in Figure 1.
Data Binding The code in Figure 1 makes use of data binding
to set the Location property of both the image control and the text
block. The Text property of the TextBlock also utilizes data binding.
To quickly review, the Location property is an attached property
of type Geopoint. Its used to specify where on the map the control
will be placed. The NormalizedAnchorPoint attached property
allows the location of the control to be fine-tuned. For example, you
can use the NormalizedAnchorPoint property to center a control
over a location, or place a control on the upper left of a location.
Both of these properties were discussed in detail in my first article
(msdn.microsoft.com/magazine/dn818495).
Figure 1 Adding Controls as Children of the Map Control
<Maps:MapControl
x:Name="myMapControl" Grid.Row="1"
MapServiceToken="{StaticResource MapServiceTokenString}" >
<!-- Progress bar which is used while the page is loading. -->
<ProgressBar Name="pbProgressBar" IsIndeterminate="True" Height="560"
Width="350" />
<TextBlock Name="tbMessage" Text="{Binding Message}"
Maps:MapControl.Location="{Binding Location}"
Maps:MapControl.NormalizedAnchorPoint="1.0,1.0"
FontSize="15" Foreground="Black" FontWeight="SemiBold"
Padding="4,4,4,4"
Visibility="Collapsed"
/>
<Image Name="imgMyLocation" Source="Assets/PinkPushPin.png"
Maps:MapControl.Location="{Binding Location}"
Maps:MapControl.NormalizedAnchorPoint="0.25, 0.9"
Height="30" Width="30"
Visibility="Collapsed" />
</Maps:MapControl>

Figure 2 Containing Controls Within a MapItemControl

<Maps:MapControl
x:Name="myMapControl"
MapServiceToken="{StaticResource MapServiceTokenString}">
<Maps:MapItemsControl>
<TextBlock Name="tbAddress" Text="{Binding Message}"
Maps:MapControl.Location="{Binding Location}"
Maps:MapControl.NormalizedAnchorPoint="1.0,1.0"
Visibility="Collapsed" />
<Image Name="imgMyLocation" Source="Assets/PinkPushPin.png"
Maps:MapControl.Location="{Binding Location}"
Maps:MapControl.NormalizedAnchorPoint="0.25, 0.9"
Height="30" Width="30"
Visibility="Collapsed"/>
</Maps:MapItemsControl>
</Maps:MapControl>

msdnmagazine.com

Data binding allows XAML controls to get their values from

underlying objects. An object created from the following class will
be used to hold the values needed for the XAML controls shown
in Figure 1:
public class LocationEntity
{
public Geopoint Location { get; set; }
public string Message { get; set; }
}

Data binding allows XAML

controls to get their values from
underlying objects.
Figure 3 shows the complete OnNavigatedTo event for a page
containing the controls in Figure 1. (If a page or view within an
application requires a lot of setup, consider placing this code in the
creation of a view model and running it asynchronously.) This code
sets the devices current location into an instance of the LocationEntity
class along with a short message. Notice that the LocationEntity
object is set into the DataContext property of the Map control.

A Few Tips

In the code shown in Figure 3, the OnNavigatedTo event performs time-consuming work. Specifically, this event calls the
GetGeopositionAsync function of a Geolocator object. This function gets the devices current location, which is needed before the
image control and the TextBlock control can be properly positioned on the map. In situations like this, any controls added to the
map as children via XAML are initially displayed in the upper-left
corner of the map until they can be positioned on the map.
This creates a poor UX. To correct this, simply set up any added
controls as Collapsed within your XAML. The added controls can
be made visible once the control locations are calculated and the
object in which the controls are bound is set up. (Refer to Figure
1 and the last few lines of code in Figure 3.)
Consider using the DesiredAccuracyInMeters property of the
Geolocator object. Setting this property to 100 meters or less will
get the most accurate data available. If the device has GPS capabilities, the GPS will then be used to determine the devices current
location. If this value is greater than 100 meters, the Geolocator
object will optimize for power and use non-GPS data, such as Wi-Fi
signals. The 100 meter threshold could change as devices evolve,
so set this property based on the needs of the application and not
the underlying threshold, which triggers different behavior. Finally,
the DesiredAccuracyInMeters property overrides anything set in
the Geolocators DesiredAccuracy property.
Also, consider using the overload of GetGeopositionAsync,
which accepts a maximumAge parameter and a timeout parameter. The maximum age parameter is a TimeSpan that specifies the
maximum acceptable age of cached location data. The timeout
parameter is also a TimeSpan and will cause the GetGeopositionAsync function to throw an exception if determining the current
location takes longer than specified.
January 2015 59

vslive.com/lasvegas

Las Vegas

MARCH 16 20

BALLYS HOTEL & CASINO LAS VEGAS, NV

Code on the Strip

Visual Studio Live!s rst stop on its 2015
Code Trip is Las Vegas, situated ttingly near
Historic Route 66. Developers, software
architects, engineers, and designers will cruise
onto the Strip for ve days of unbiased and
cutting-edge education on the Microsoft
Platform. Navigate the .NET Highway with
industry experts and Microsoft insiders in
90+ sessions and fun networking events all
designed to make you better at your job.

EVENT PARTNERS

SUPPORTED BY

PRODUCED BY
magazine

Untitled-4 2

12/10/14 3:26 PM

NAVIGATE THE
.NET HIGHWAY
TRACKS INCLUDE:
Visual Studio / .NET Framework
JavaScript / HTML5 Client
Modern App Development
ASP.NET
Cross-Platform Mobile
Development

Database and Analytics

Windows Client
Cloud Computing
Windows Phone
SharePoint Cloud Business Apps

RegiVter NOW and

Sav $400!
Use promo code VSLJAN4

CONNECT WITH VISUAL STUDIO LIVE!

twitter.com/vslive @VSLive
facebook.com Search VSLive
linkedin.com Join the
Visual Studio Live group!

Scan the QR code to

register or for more
event details.

TURN THE PAGE FOR MORE EVENT DETAILS

vslive.com/lasvegas

Untitled-4 3

12/10/14 3:10 PM

vslive.com/lasvegas

Las Vegas

MARCH 16 20

BALLYS HOTEL & CASINO LAS VEGAS, NV

AGENDA AT-A-GLANCE
Ballys Hotel & Casino
will play host to
Visual Studio Live!, and
is offering a special
reduced room rate to
conference attendees.

CONNECT WITH
VISUAL STUDIO LIVE!
twitter.com/vslive
@VSLive
facebook.com
Search VSLive
linkedin.com Join the
Visual Studio Live group!

Scan the QR code to

register or for more
event details.

Untitled-4 4

ASP.NET

Cross-Platform Mobile
Development

Cloud Computing

JavaScript /
HTML5 Client

START TIME

END TIME

Visual Studio Live! Pre-Conference Workshops: Monday, March 16, 2015

7:30 AM

9:00 AM

Pre-Conference Workshop Registration - Coffee and Morning Pastries

9:00 AM

6:00 PM

M01 - Workshop: Native Mobile App Development for iOS, Android

and Windows Using C# - Marcel de Vries & Rockford Lhotka

6:45 PM

9:00 PM

Dine-A-Round

START TIME

END TIME

Visual Studio Live! Day 1: Tuesday, March 17, 2015

7:00 AM

8:00 AM

Registration - Coffee and Morning Pastries

8:00 AM

9:00 AM

Keynote: To Be Announced

9:15 AM

10:30 AM

T01 - An Overview of the Xamarin Programming

Platforms - Laurent Bugnion

T02 - JavaScript for the C# (and Java) Developer

- Philip Japikse

10:45 AM

12:00 PM

T07 - Building Cross-Platform Applications with

XAML, Xamarin, Xamarin.Forms and MVVM Light
- Laurent Bugnion

T08 - Write Next Generation, Object Oriented

JavaScript, with TypeScript. - Rachel Appel

12:00 PM

1:30 PM

1:30 PM

2:45 PM

T13 - Microsoft - To Be Announced

T14 - It's the Top 10 of Cool HTML5 Features Every

Developer Should Know Right Now - Gill Cleeren

3:00 PM

4:15 PM

T19 - Creating Responsive Cross-Platform Native/

Web Apps with JavaScript and Bootstrap
- Ben Dewey

T20 - I Just Met You, and This is Crazy, But Here's

My NaN, So Call (me) Maybe? - Rachel Appel

4:15 PM

5:30 PM

Welcome Reception

START TIME

END TIME

Visual Studio Live! Day 2: Wednesday, March 18, 2015

7:00 AM

8:00 AM

Registration - Coffee and Morning Pastries

8:00 AM

9:00 AM

Keynote: To Be Announced - Rockford Lhotka, CTO, Magenic

9:15 AM

10:30 AM

W01 - Cross-Platform Mobile Apps Using AngularJS

and the Ionic Framework - Ben Dewey

W02 - Getting Started with jQuery and How it

Works Together with ASP.NET WebForms and MVC
- Gill Cleeren

10:45 AM

12:00 PM

W07 - Building Cross-Platform Apps in C#

- Rockford Lhotka

W08 - Build Data-Centric HTML5 Single Page

Applications with Breeze - Brian Noyes

12:00 PM

1:30 PM

1:30 PM

2:45 PM

W13 - Building Your First Universal Application for

Windows Phone and Windows Store - Brian Peek

W14 - Busy Developer's Guide to MEANJS

- Ted Neward

3:00 PM

4:15 PM

W19 - Getting Started with Windows Phone 8.1

Development - Nick Landry

W20 - Creating Angular Applications Using Visual

Studio LightSwitch - Michael Washington

4:30 PM

5:45 PM

W25 - Strike Up a Conversation with Cortana on

Windows Phone - Walt Ritscher

W26 - JavaScript Patterns for the C# Developer

- Ben Hoelting

7:00 PM

9:00 PM

Visual Studio Live! Evening Event

START TIME

END TIME

Visual Studio Live! Day 3: Thursday, March 19, 2015

7:00 AM

8:00 AM

8:00 AM

9:15 AM

TH01 - From WPF to Universal Apps: A Guide

Forward - Rockford Lhotka

TH02 - AngularJS for ASP.NET MVC Developers

- Miguel Castro

9:30 AM

10:45 AM

TH07 - UX Design Principle Fundamentals for NonDesigners - Billy Hollis

TH08 - Securing Angular Apps - Brian Noyes

11:00 AM

12:15 PM

TH13 - XAML: Achieving Your Moment of Clarity

- Miguel Castro

TH14 - Unit Testing JavaScript Applications

- Ben Dewey

12:15 PM

1:30 PM

Lunch - Visit Exhibitors

Birds-of-a-Feather Lunch - Visit Exhibitors

Registration - Coffee and Morning Pastries

Lunch

1:30 PM

2:45 PM

TH19 - Tricks and Shortcuts for Amazing UI in XAML

- Billy Hollis

TH20 - Building Single Page Web Applications Using

JavaScript and the MVVM Pattern - Ben Hoelting

3:00 PM

4:15 PM

TH25 - WPF Data Binding in Depth - Brian Noyes

TH26 - Busy Developers Guide to SignalR and

WebSockets - Ted Neward

START TIME

END TIME

7:30 AM

8:00 AM

Post-Conference Workshop Registration - Coffee and Morning Pastries

8:00 AM

5:00 PM

F01 - Workshop: Hybrid Mobile Apps with Cordova, Angular, and Azure - Brian Noyes

Visual Studio Live! Post-Conference Workshops: Friday, March 20, 2015

Speakers and sessions subject to change

12/10/14 3:10 PM

NAVIGATE THE
.NET HIGHWAY

Presented in partnership with

BONUS CONTENT! Modern Apps Live! is

now a part of Visual Studio Live! Las Vegas
at no additional cost!
SharePoint Cloud
Business Apps

Database and
Analytics

Visual Studio /
.NET Framework

Windows Client

Windows Phone

Modern Apps Live!

(Separate entry fee required)

M02 - Workshop: Deep Dive into Visual Studio 2013, TFS

and Visual Studio Online - Brian Randell

M04 - Workshop: Modern App Technology

Overview - Android, iOS, Cloud,
and Mobile Web

M03 - Workshop: UX Design Process Essentials

Steps and Techniques - Billy Hollis

T03 - What's Up with ASP.NET vNext (Who

Moved My Cheese?) - Mark Michaelis

T04 - Cloud or Not, 10 Reasons Why You

Must Know Web Sites - Vishwas Lele

T05 - New IDE and Editor Features in Visual

Studio - Deborah Kurata

T06 - Dening Modern App Development

T09 - Microsoft - To Be Announced

T10 - Azure 10-10: Top 10 Azure

Announcements in T-10 Months
- Vishwas Lele

T11 - Visual Studio Online: An Update Every

Three Weeks - Brian Randell

T12 - Modern App Architecture

T15 - ASP.NET MVC For The WebForms

Developer - Philip Japikse

T16 - Busy Developer's Guide to NoSQL Ted Neward

T17 - Defensive Coding Techniques in C#

- Deborah Kurata

T18 - ALM with Visual Studio Online (TFS)

and Git

T21 - Slice Development Time With ASP.NET

MVC, Visual Studio, and Razor - Philip Japikse

T22 - Busy Developers Guide to Six

Developer Cloud Platforms - Ted Neward

T23 - Automated Cross Browser Testing of

Your Web Applications with Visual Studio
CodedUI - Marcel de Vries

T24 - Reusing Business and Data Access

Logic Across Platforms

W03 - ASP.NET vNext in All Its Glory

- Adam Tuliper

W04 - T-SQL for Application Developers Attendees Choose - Kevin Goff

W05 - Using Microsoft Application Insights

to Implement a Build, Measure, Learn Loop
- Marcel de Vries

W06 - Coding for Quality and Maintainability

W09 - Understanding the Power of ASP.NET

MVC vNext - Ben Hoelting

W10 - SQL Server Reporting Services for

Application Developers - Attendees Choose
- Kevin Goff

W11 - Build It and Ship It with TFS and

Release Management - Brian Randell

W12 - UX Design for Modern Apps

W15 - Improving Web Application

Performance - Ben Hoelting

W16 - A Tour of Azure for Developers

- Adam Tuliper

W17 - Programming C# 6.0 - Mark Michaelis

W18 - Applied UX: iOS, Android, Windows

W21 - To Be Announced

W22 - 10 Ways to Guarantee Your Azure

Project Will Fail! - Michael Collier

W23 - Writing Asynchronous Code using

.NET 4.5 and C# 5.0 - Brian Peek

W24 - Leveraging Azure Services

W27 - WCF & Web API: Can We All Just

Get Along?!? - Miguel Castro

W28 - Intro to Azure API Management

- Michael Collier

W29 - The Busy Developers Guide to

Virtualization with Hyper-V - Brian Randell

W30 - Unit Testing Modern Apps

TH03 - Fiddler and Your Website

- Robert Boedigheimer

TH04 - SQL Server in Azure Virtual Machines

- Eric D. Boyd

TH05 - Managing the .NET Compiler

- Jason Bock

TH06 - Analyzing Results with Power BI

TH09 - Web Essentials

- Robert Boedigheimer

TH10 - Microsoft Azure SQL Database - SQL

Server in the Cloud - Leonard Lobel

TH11 - Real World Scrum with Team

Foundation Server 2013 & Visual Studio
Online - Benjamin Day

TH12 - Building a Modern Android App

with Xamarin

TH15 - Node.js for .NET Developers

- Jason Bock

TH16 - Database Development with SQL

Server Data Tools - Leonard Lobel

TH17 - Git for the Microsoft Developer

- Eric D. Boyd

TH18 - Building an iOS and Android App

with Xamarin.Forms

TH21 - Building Rich UI with ASP.NET MVC

and Bootstrap - Walt Ritscher

TH22 - Creating Visual Studio Cloud Business

Applications for SharePoint 2013 Using
AngularJS - Michael Washington

TH23 - To Be Announced

TH24 - Building a Windows App

TH27 - ASP.NET - Rudiments of Routing

- Walt Ritscher

TH28 - Cloud Business Apps and SharePoint

Workows - Michael Washington

TH29 - Zero to Hero: Untested to Tested with

Visual Studio Fakes - Benjamin Day

TH30 - Building a Responsive

Single Page App

(Separate entry fee required)

F02 - Workshop: SQL Server 2014 for Developers - Leonard Lobel

F03 - Workshop: Modern App Development

In-Depth: iOS, Android, Windows, and Web

vslive.com/lasvegas

Untitled-4 5

12/10/14 3:38 PM

Figure 3 The OnNavigatedTo Event Creating Objects Needed

for Data Binding
protected async override void OnNavigatedTo(NavigationEventArgs e)
{
// Call the navigation helper base function.
this.navigationHelper.OnNavigatedTo(e);
// Get the user's current location so that it can be
// used as the center point of the map control.
Geolocator geolocator = new Geolocator();
// Set the desired accuracy.
geolocator.DesiredAccuracyInMeters = 200;
// The maximum acceptable age of cached location data.
TimeSpan maxAge = new TimeSpan(0, 2, 0);
// Timeout used for the call to GetGeopositionAsync.
TimeSpan timeout = new TimeSpan(0, 0, 5);

a progress bar shown across the Map control tells the user that its
the Map control doing work.

Data Binding a Collection

Binding a collection to the Map control is similar to binding a
collection to a ListBox, ListView, GridView or any other control
that displays collections. To illustrate this with the Map control,
consider the following class:
public class Team
{
public string TeamName { get; set; }
public Uri ImageSource { get; set; }
public string Division { get; set; }
public string StadiumName { get; set; }
public Geopoint StadiumLocation { get; set; }
public Point NAP { get; set; }
}

Geoposition geoposition = null;

try
{
geoposition = await geolocator.GetGeopositionAsync(maxAge, timeout);
}
catch (Exception)
{
// Add exception logic.
}
// Set up the LocationEntity object.
LocationEntity locationEntity = new LocationEntity();
locationEntity.Location = geoposition.Coordinate.Point;
locationEntity.Message = "You are here";
// Specify the location that will be at the center of the map.
myMapControl.Center = locationEntity.Location;
// Set the map controls data context so data binding will work.
myMapControl.DataContext = locationEntity;
// Zoom level.
myMapControl.ZoomLevel = 15;
// The map control has done most of its work. Make the controls visible.
imgMyLocation.Visibility = Windows.UI.Xaml.Visibility.Visible;
tbMessage.Visibility = Windows.UI.Xaml.Visibility.Visible;
// Collapse the progress bar.
pbProgressBar.Visibility = Windows.UI.Xaml.Visibility.Collapsed;
}

The previous two tips should substantially improve performance.

However, in areas with poor Internet connectivity and on devices
with low memory and slower CPUs, the Map control and the
Mapping Services APIs may still need time, resulting in the user
experiencing a delay. In such cases, the Map control will display
a map of the globe at the current zoom level until a location is set
into the Map controls center point property. This is also a poor UX
because the user should be given some visual indication that work
is occurring. In the XAML in Figure 1, a progress bar is added as
a child of the Map control and is made visible. The progress bar is
set up as indeterminate because the exact amount of time needed
isnt known. Once the center property of the map is calculated and
set into the Map control, the progress bar can be collapsed. Best
practices for the use of indeterminate progress bars indicate they
should always be placed at the very top of the page, and for most
scenarios, I agree. However, when using the Map control, I prefer
to see the progress bar displayed right across the center of the map.
The Map control tends to grab the users focus and for this reason a
progress bar at the very top of the page might be overlooked. Also,
64 msdn magazine

An instance of this class could be used to hold basic information

about a professional sports team. Notice that one of the properties
is a Geopoint property that contains the location of the stadium
used as the teams home field.
The XAML in Figure 4 sets up the Map control such that a collection of Team objects can be bound to it and the location of each
teams home field stadium will be marked with a small icon.
If youre familiar with collection binding in other controls,
nothing in Figure 4 should be unfamiliar. An Image control is
contained within a DataTemplate. The Image controls attached
Location property is bound to the StadiumLocation property of the
Team object. The Image control also has an attached NormalizedAnchorPoint property, which is bound to a property that centers
the image over the stadium location.
Note that when binding the Map control to a collection, the
NormalizedAnchorPoint property cant be hardcoded. It must
be bound to a value in the underlying object. If you try to set this
property like this:
Maps:MapControl.NormalizedAnchorPoint="0.5,0.5"

the XAML editor will not like this syntax and the value will never
get set. Microsoft is aware of this problem.
Figure 5 shows the Map control displaying the NFL stadium
locations. (This screenshot is from the code sample that accompanies this article. The code sample contains all the ancillary code
not shown here for the sake of brevity, such as instantiating the collection and properly formatting the control to fit nicely on a page.)
Figure 4 Setting Up the Map Control for Collection Binding
<Maps:MapControl
x:Name="myMapControl" Grid.Row="1"
MapServiceToken="{StaticResource MapServiceTokenString}" >
<Maps:MapItemsControl x:Name="MapItems" >
<Maps:MapItemsControl.ItemTemplate>
<DataTemplate>
<Image Source="{Binding ImageSource}"
Maps:MapControl.Location="{Binding StadiumLocation}"
Maps:MapControl.NormalizedAnchorPoint="{Binding NAP}"
Height="15" Width="15"
/>
</DataTemplate>
</Maps:MapItemsControl.ItemTemplate>
</Maps:MapItemsControl>
</Maps:MapControl>

Windows Phone 8.1

Scaling Controls
Its possible to size controls so they
represent a specified distance.
For example, you may want to
show a scale on your map. A scale
gives users a visual clue, allowing
them to guess distances between
objects on the map. Also, if youre
using the Map control to depict
a geofence, a good way to do this
is to add an ellipse to the Map
control, shape it into a circle,
and size the radius of the circle
such that the circle covers the
alert distance of the geofence.
Controls in Windows Phone
8.1
are sized by specifying width
Figure 5 The Stadiums of the
and
height in pixels. Therefore,
National Football League
scaling controls to represent a certain distance involves determining how much distance is represented
by a single pixel on the Map control. Heres the equation for determining the distance represented by a single pixel on the Map control:
const double BING_MAP_CONSTANT = 156543.04;
double MetersPerPixel =
BING_MAP_CONSTANT * Math.Cos(latitude) / Math.Pow(2, zoomLevel);

The latitude used should come from the Map controls center
property. It must be in radians, not degrees. Because latitude is
typically expressed in degrees, youll have to convert it to radians
by multiplying the degree value by Pi/180 (Math.PI/180). The zoom
level comes from the zoom level of the Map control and is a value
from 1 to 20. On Windows Phone 8.1, its expressed as a double.
A complete mathematical derivation of the equation shown here is
beyond the scope of this article; however, a few comments are in order.
At first glance the equation may seem incorrect. Why does the
latitude matter? Shouldnt the scale of a map simply be a function of
the zoom level? The truth is that latitude does matter and the zoom
level alone is insufficient when determining the scale of a map.
Bing Maps, which is the back end for the Map control, makes use
of the Mercator projection. The Mercator projection is a technique
for projecting the details of the surface of a sphere (in this case the
Earth) onto a square surface. This projection introduces inconsistencies in scale because the farther you go from the equator, the more
stretching there is to maintain a square fit. A simple experiment can
help you better understand the Mercator projection. Peel an orange
and save all the pieces. When the orange is completely peeled, put

the pieces back together on a flat surface and try to get them to
exactly fit into a square shape. Its impossible. The only way to get these
pieces into a square shape is to stretch the pieces that are not on the
equator of your orange. The closer you get to the poles, the more
stretching you need to do to get a square fit. This stretching creates a
difference in scale as compared to locations close to the equator. The
amount of stretching needed is a function of the distance from the
equator. Hence, the dependency on latitude. As an example of this
stretching in action on a Mercator projection of the Earth, consider
the two cities of Quebec, Canada, and Key West, Fla. A completely
zoomed-in view (Zoom Level of 20) of Quebec using the Map control will result in 100 pixels representing 33.5 feet. The same zoom
level of Key West will result in 100 pixels representing 44.5 feet. The
residents of Quebec enjoy a scale thats 11 feet better than the residents of Key Westa small consolation for the colder climate. For
more information on the Mercator projection, check out the MSDN
Library article, Understanding Scale and Resolution, at bit.ly/1xIqEwC.
The final fact to note on the equation is the Bing Map Constant.
The Bing Map Constant is based on the Earths radius and on the
equations Microsoft uses to determine the map view for a specified
zoom level. Its units are meters per pixel.
Figure 6 shows a TextBlock and a Rectangle added to a Map
control to be used as a distance scale. Figure 7 shows the ZoomLevelChanged event and the logic needed to create a distance scale.
(The code sample checks the RegionInfo.CurrentRegion.IsMetric
property and shows metric values for users who are more familiar
with the metric system.) Finally, Figure 8 show a screenshot from
the code samplea distance scale added to a Map control.

Downloading Maps for Offline Use

Maps can be downloaded for offline use. This is handy for applications that need to provide mapping functionality when the device
is without any form of Internet connectivity, such as while a user
Figure 7 ZoomLevelChanged Event
private void myMapControl_ZoomLevelChanged(MapControl sender, object args)
{
// Get the Map control's current zoom level.
double zoomLevel = sender.ZoomLevel;
// Use the latitude from the center point of the Map control.
double latitude = sender.Center.Position.Latitude;
// The following formula for map resolution needs latitude in radians.
latitude = latitude * (Math.PI / 180);
// This constant is based on the diameter of the Earth and the
// equations Microsoft uses to determine the map shown for the
// Map control's zoom level.
const double BING_MAP_CONSTANT = 156543.04;

Figure 6 Xaml TextBlock and Rectangle Used As a Distance Scale

// Calculate the number of meters represented by a single pixel.

double MetersPerPixel =
BING_MAP_CONSTANT * Math.Cos(latitude) / Math.Pow(2, zoomLevel);

<Maps:MapControl
x:Name="myMapControl" Grid.Row="1"
MapServiceToken="{StaticResource MapServiceTokenString}"
ZoomLevelChanged="myMapControl_ZoomLevelChanged">

// Aditional units.
double KilometersPerPixel = MetersPerPixel / 1000;
double FeetPerPixel = MetersPerPixel * 3.28;
double MilesPerPixel = FeetPerPixel / 5280;

<!-- Distance scale, which is located at the lower left of the Map control. -->
<TextBlock Name="tbScale" Text="Scale Text" FontSize="15" Foreground="Black"
Margin="24,530,24,6" Opacity="0.6" />
<Rectangle Name="recScale" Fill="Purple" Width="100" Height="6"
Margin="24,548,24,24" Opacity="0.6" />
</Maps:MapControl>

msdnmagazine.com

// Determine the distance represented by the rectangle control.

double scaleDistance = recScale.Width * MilesPerPixel;
tbScale.Text = scaleDistance.ToString() + " miles";
}

January 2015 65

is driving on a long highway

between urban areas or hiking
deep in the woods.
Maps cant be downloaded
programmatically without user
consent because they take up a
significant amount of storage.
The user must opt in to each
download by using a systemprovided UX that allows needed
maps to be selected and downloaded. This is accomplished
using the static class MapManager, which belongs to the
Windows.Services.Maps namespace. This class provides funcFigure 8 Distance Scale Added tions for both downloading
maps and updating maps.
to a Map Control Showing
The ShowDownloadedMapsUI
Sunset Pier in Key West, Fla.
function shown here will launch
the built-in Maps application and display the page shown in Figure 9:
MapManager.ShowDownloadedMapsUI();

The built-in maps application is named Maps and the page shown
in Figure 9 can also be accessed by tapping the Download maps
button from the Settings menu option. Because this function takes
the user to another application (the built-in maps application), he
will have to return to the original application manually.
The page shown in Figure 9 shows all the maps that have been
previously downloaded. It also contains a menu option that allows
you to delete any of the previously downloaded maps in case storage on the device needs to be freed or a map is no longer needed.
Tapping the add button (+) starts the map selection wizard,
displaying a page that lets you specify the continent containing
the map you wish to download. Once a continent is selected, the
page displaying all countries within the selected continent will be
displayed. You can select all the countries within a continent and

download maps for the entire continent, but this would require a
large amount of storage. For example, all the maps for the United
States require 3.4GB of storage. The entire continent of North and
Central America would require much more.
If you select a large country that contains multiple regions or
states, youll see a page that lets you choose specific regions or states.
You can select more than one region by using the Select button.
When a region is specified, youll be taken to the Download page
shown in Figure 10. This same page will also be shown if you
select a small country without any regions or states.
Before downloading any maps, its a good idea to encourage users to
check the available space on the device by going to the Settings application and selecting Storage sense. This will show the available space
on the device, as well as the storage used by each application on the device in case space needs to be freed up. At the time of this writing, there
are no WinRT APIs that allow you to programmatically determine the
amount of storage used on a device or the amount of storage available.
Its also a good idea to download maps while connected to an
unrestricted network, such as Wi-Fi. The code sample for this
article shows how to check the current network connection and
warn the user if theyre on a metered connection.

Updating Downloaded Maps

Maps are occasionally changed or modified. New roads may be added
to a city, street names may change, and regional borders are sometimes altered. Therefore, downloaded maps should be periodically
updated. The MapManager static class contains a function named
ShowMapsUpdateUI for updating downloaded maps:
MapManager.ShowMapsUpdateUI();

The ShowMapsUpdateUI function is similar to the ShowDownloadedMapsUI function in that it takes the user to the built-in maps
application. When this function is called, all downloaded maps
are checked to see if they need to be updated. If any previously
downloaded maps require updating, the user will be told the size
of the update and given the option to either cancel or proceed with
the download. This same functionality is also available from the
Settings menu option of the Maps application.
If your application encourages the downloading of maps by
utilizing ShowDownloadedMapsUI, its a best practice to also use
ShowMapsUpdateUI so that downloaded maps can stay current.

Wrapping Up
In this article, I showed you how to add application data to the
Map control for Windows Phone 8.1. This included adding XAML
controls to a map, data binding and drawing controls to scale.
I also showed how to take the Map controls underlying data
offline so the Map control can be set up to work seamlessly within
applications designed for offline use.
Q
KEITH PIJANOWSKI is an engineer, entrepreneur, and business guy. He has more
than 20 years of experience in the software industry, working for startups
and large companies in roles that have ranged from writing code to business
development. Reach him at keithpij@msn.com or twitter.com/keithpij.

Figure 9 The UI for

Downloading Maps
66 msdn magazine

Figure 10 Downloading a Map

THANKS to the following Microsoft technical expert for reviewing this article:
Mike OMalley
Windows Phone 8.1

Untitled-2 1

5/31/13 10:57 AM

MICROSOFT AZURE

Cross-Platform
Cloud Automation
with JavaScript
Steven Edouard
Not everyone knows Microsoft Azure has the Cross-Platform

The Cross-Platform CLI

Command-Line Interface (xplat-cli) for managing Azure Infrastructure-as-a-Service (IaaS) and Platform-as-a-Service (PaaS) services,
but it does. For years, Microsoft has advocated using Windows
PowerShell as a scripting language, which worked well for provisioning
cloud-based resources. But now that more developers are involved
in deploying and moving applications to the cloud, it makes more
sense to use JavaScript. The Azure Service Management REST APIs
make this cross-platform implementation easier. The official Azure
xplat-cli is implemented on top of Node.js, the same runtime often
used to build high-performance Web sites and APIs with JavaScript.
Building on top of Node.js makes scripting truly cross-platform.
It means you can run your scripts for managing virtual machines
(VMs), SQL Databases, Web sites and virtual networks from any
machine. This frees you to do automation scripts in a wider variety of
languages and for a wider variety of client computers (see Figure 1).
There are many advantages to scripting cloud resource provisioning. For example, the Web portal isnt repeatable and requires
human input. Another advantage is being able to source control the
steps to provision and deploy your service using different specific
configuration values. In this article, Ill walk through scripting the
process of creating VMs associated with an Affinity Group and a
virtual network using Node.js as our script automation environment.

Like the Azure PowerShell CLI, the cross-platform version offers

powerful script ability to manage IaaS and PaaS services. This lets
you describe your infrastructure with code, share it with your team
via source control, and automate the creation of development, test,
and production application deployments.
To get started with the xplat-cli, head over to Nodejs.org and install
the appropriate Node.js package for your system. After installing
Node, you can view the azure-cli package on npmjs.org/package/azure-cli.
The package listing gives you all the relevant documentation on
the CLI, as well as the simple command line to install the package.
This package is a CLI, so it makes sense to install it in the global
Node.js package store by specifying g, so you can make Azure

This article discusses:

How to automate configuration tasks with cross-platform scripting
Configuring and deploying groups of VMs
Confirming lists of provisioned VMs

Windows
PowerShell Scripts

JavaScript

Shell Script

Node.js xplat-cli

Azure Datacenter
Azure Management REST API
Databases
Compute
Storage

Technologies discussed:
Microsoft Azure, Windows PowerShell, Node.js
68 msdn magazine

Figure 1 Supported Scripting Environments

BatchScript

Access Your Subscription from the CLI

Not surprisingly, most of the Azure commands wont
actually work until you sign up for an Azure subscription
and tell the command-line tool what Azure subscription to use. Similar to the Windows PowerShell tools,
you need to give the CLI access to your subscription.
There are two ways to do thisvia an Active Directory
account using a Microsoft Organizational Account with
the login command or with a portable .publishsettings file
using the account import command. The latter method
is more portable as it doesnt require the Organizational
Account setup. Theres an easy way to download your
.publishsettings by executing:
Figure 2 The Azure xplat-cli

azure account download

This will actually open your default browser to a

link that will have you sign in to your Azure account
and download your .publishsettings file. After the file
downloads, you can use the account import command
to load up the file:
azure account import <PATH TO .publishsettings>

Figure 3 CLI Command to Create a Virtual Machine

command-line calls from any directory, instead of having to

explicitly invoke the CLI from a specific directory:
npm install azure-cli -g

After a few lines of output from the Node Package Manager, youll
have your azure-cli installed and you can start using it immediately.

Completely Self-Discoverable CLI

The great thing about the CLI is that its completely self-discoverable.
You can easily get started just by typing commands youd like to
know more about. Figure 2 shows what happens when you call
the base CLI command Azure.
Theres a plethora of possible commands you can do from the
base Azure command. Say you notice the vm command from the
help output in Figure 2. You can drill into the command to better
understand how to use it by just typing the azure vm command
(see Figure 3):
This same pattern applies to all commands in the CLI. It creates
an intuitive interface for exploring the command line.
Figure 4 Config.json Maintains Flexible Settings
{
"vnet_name":"azuretedvnet2",
"vm_count": "3",
"vm_image_name":
"b39f27a8b8c64d52b05eac6a62ebad85__Ubuntu-14_04-LTS-amd64-server20140416.1-en-us-30GB",
"vm_base_name":"sedouardmachine",
"vm_username":"sedouard",
"vm_password":"p@assW0rd",
"affinity_group":{
"name":"azureteddemoaffinitygroup",
"label":"sedouard",
"location":"West US"
}
}

msdnmagazine.com

This will set up the command line to use the Azure

subscription specified in the .publishsettings for any
Azure command called on the CLI. To see if the CLI
works, call the site list command to output all Web sites
running on your current subscription.

Create a Dev VM from the CLI

VMs are particularly useful for DevOps scenarios. You can easily
provision them through the xplat-cli.
Ill eventually use the CLI to automate these tasks by scripting
them with JavaScript on Nodejs. For example, Ill create a new
Ubuntu VM. To do so, I need to find a VM image. I can fetch a list
of available VMs by executing the command:
azure vm image list

The command output shows a list of available public and private

images you can use to provision a VM. The public images are curated
by Microsoft, but you can use your own uploaded images, as well.
To create a VM, you need a datacenter location. To get a list of
valid locations where you can place your VM, execute this command:
azure vm location list

You can use this location as input to thevm createcommand

to create a new VM:
azure vm create <yourdomainname> <imagename> <username> <password>
--location '<selected datacenter region>'

Now that youve successfully created a VM from the command

line, take a look at how easily you can do this using a script.

Script Away Azure Resources in JavaScript

More developers are trying to use JavaScript for just about everything. Provisioning resources in the cloud is no exception. This
supports other scripting languages, as well, such as Bash, Batch and
even Windows PowerShell. The choice of language boils down to
the platforms on which youll run and your language preferences.
Microsoft will soon release the Azure Resource Manager, which
will use a more declarative model instead of the programmatic
model. Until Azure Resource Manager is ready, you can use some
of these techniques to automate your provisioning to the cloud.
January 2015 69

If you browse closely on the azure-cli npm documentation

(npmjs.org/package/azure-cli), youll notice that theres a library dependent on azure-cli called azure-scripty. It turns out someone in the
open source Node.js community has already decided to write a
command-line adapter for the azure-cli for JavaScript. Because this
library is a command-line wrapper, you can use all the same commands as on the command line in JavaScript. Heres an example
that will list all of our VMs:
Figure 5 Create an Affinity Group
//////////////////////////////////////////////////////////////////////////////
//Creates an affinity group if the specified one doesn't exist
//////////////////////////////////////////////////////////////////////////////
function createAffinityGroup(cb) {
console.log('Getting affinity groups...');
scripty.invoke('account affinity-group list', function (err, result) {
// Check for errors.
if (err) {
return cb(err);
}
console.log('Current affinity groups');
// For debugging purposes, list out names of existing affinity groups.
for (var i in result) {
console.log(result[i].name);
}
// Get the name of the desired affinity group from the config.json file.
var affinityGroup = nconf.get('affinity_group').name;
var label = nconf.get('affinity_group').label;
// Verify that affinity group hasn't already been created.
for (var i in result) {
if (result[i].name === affinityGroup && result[i].label === label) {
// The affinty group to use in the config already exists.
return cb();
}
}
// Create affinity group because it doesn't exist.
console.log('Specified affinity group ' + affinityGroup +
' does not exist, creating new one...');
var cmd = {
command: 'account affinity-group create',
positional: [affinityGroup],
location: '\"West US\"',
label: label
};
scripty.invoke(cmd, function (err) {
if (err) {
cb(err);
}
return cb();
});
});
}

Figure 6 Create a Virtual Network

//////////////////////////////////////////////////////////////////////////////
//Creates the config specified vnet, if it doesn't already exist
//////////////////////////////////////////////////////////////////////////////
function createVirtualNetwork(cb) {
console.log('Getting networks...');
scripty.invoke('network vnet list', function (err, result) {
if (err) {
return cb(err);
}

// Get the list of existing VMs on this subscription

scripty.invoke('vm list', function (err, result) {
if (err) {
return cb(err);
}
for (var i in result) {
console.log(result[i].VMName);
}
});

To get started, take a look at how you can make a batch of VMs.
First, install the necessary packages:
# a
npm
# a
npm
# a
npm

command-line wrapper for the azure-cli

install azure-scripty
small library that helps manage asynchronous tasks
install async
library that makes using JSON configuration files easy
install nconf

To avoid hardcoding the configuration settings for staging, test

and production, use a configuration file to keep the scripts flexible
enough for different environments such as development, test and
production. Call this configuration file config.json (see Figure 4).
You can use that configuration file to set up a developer environment, complete with three Ubuntu VMs that are part of the affinity
group (azureteddemoaffinitygroup) and a corresponding virtual
network (azuretedvnet2). This approach is declarative, meaning
you only need to make changes to a configuration file to change
the way your development environments are set up. You can also
have additional versions for staging, test and productionand
source control each.
The custom JavaScript code will parse the configuration file so the
dependencies, such as affinity groups, will be provisioned with the correct names and values. Inside the config.json file, the script can read
the desired name for the affinity group (azureteddemoaffinitygroup).
From there, it can query check to see if the affinity group has already
been created by looping through the list of provisioned affinity groups
that already exist in the datacenter (see Figure 5).
Figure 7 Create a List of Virtual Machines to Provision
scripty.invoke('vm list', function (err, result) {
if (err) {
return cb(err);
}
// Read the desired VM name from config.json.
var baseName = nconf.get('vm_base_name');
var vmNames = [];
// Create the array of the computed VM names.
for (var z = 0; z < count; z++) {
vmNames.push(baseName + z.toString());
}
// Go through the list of existing VMs.
for (var i in result) {
for (var k in vmNames) {

var networkName = nconf.get('vnet_name');

if (result[i].VMName === vmNames[k]) {

// A VM we intend on creating already exists on this sub.
// Remove it on the list of VMs to create.
delete vmNames[k];
}

for (var i in result) {

console.log(result[i].name);
}
}

// Check if VNet name listed in config.json exists.

}
// If it doesn't, create the VNet.
});
}

70 msdn magazine

// vmNames now only contains the name of VMs that do not exist.
// Create them.

Microsoft Azure

Figure 8 Finish Provisioning Virtual Machines

var
var
var
var
var
var

domainName = nconf.get('dns_name');
userName = nconf.get('vm_username');
password = nconf.get('vm_password');
imageName = nconf.get('vm_image_name');
vmCreationTasks = [];
taskArguments = [];

for (var m in vmNames) {

if (vmNames[m]) {
var cmd = {
command: 'vm create',
positional: [vmNames[m], imageName, userName, password],
'affinity-group': nconf.get('affinity_group').name,
'virtual-network-name': nconf.get('vnet_name')
}
// Define the async task function that will create each VM.
var task = function (args, cb) {
console.log('Creating vm ' + vmNames[args[0]]);
scripty.invoke(args[1], function (err) {
if (err) {
console.log('Vm creation failed: ' + vmNames[args[0]]);
return cb(err);
}
console.log('vm creation of ' + vmNames[args[0]] + ' successful');
cb();
});
}
// Bind this function to this context and pass the index and VM create command.
task = task.bind(this, [m, cmd]);
vmCreationTasks.push(task);
}
}
// Execute each VM creation task using the async libray.
async.series(vmCreationTasks, function (err) {
if (err) {
return cb(err);
}
console.log('All VMs created successfully!');
cb();
})

Notice how the azure-scripty library lets you easily parse output
from the azure-cli via callback function result argument. It also lets
you create CLI commands by declaratively specifying the command,
positional and named parameters.
The pattern in Figure 6 is the same one you use to create your
virtual network and uses the azure-scripty library in the same way.
Some of the code in Figure 6 has been omitted for brevity, but
the same pattern exists. The needed resource is specified in the
configuration file. Youll need to provision it if it doesnt exist.

You can create a batch of VMs once youve created the dependencies (the affinity groups and networks). The number of VMs
youll create is specified in the vm_count field in the config.json
configuration file. Following the same pattern, first list the current
VMs on the subscription, check if any of the VMs youll create
already exist and create only the ones that dont exist.
Figure 7 will follow a simple algorithm to name the VMs using
the vm_base_name field in the config.json and append the numbers 0 (vm_count 1) to the end of the VM name. Given the
current config.json, create sedouardmachine0, sedouardmachine1
and sedouardmachine2 because none of those machines already
exist on the subscription.
Figure 7 confirms the list of non-existing VMs in the vmNames
array. Figure 8 will use the async library to manage the asynchronous tasks that will go off and create the VMs.
Figure 8 creates a task for each VM you need to create. Then
you need to add those tasks to an array, vmCreationTasks. Each
task function is bound to the necessary arguments in order to call
the Azure azure-cli with the correct VM name. Afterward, use
the async library to run each task in series and execute the callback with a null or an error parameter indicating if the VMs were
successfully created. Figure 9 confirms the output of the script and
lists out all the VM that have been created.
Each of your VMs has been successfully created. You can even
check back to your portal to confirm that the machines have been
created and are part of the specified virtual network. Use the GitHub
repository at bit.ly/azure-xplat-cli to get started scripting your Azure
setup using Node.js.

Wrapping Up
The Azure xplat-cli lets you completely automate the process of
configuring and deploying Azure resources in portable scripts
you can use on virtually any OS. It also lets you use source control
and configuration files to simplify sharing. You could have multiple versions of config.json, each one representing a deployment
model for test, staging and production.
You can also recreate the deployment process quickly and efficiently. The forthcoming Azure Resource Manager Microsoft will
release in 2015 will go above and beyond whats illustrated here, but this
approach is virtually identical to what the Azure Resource Manager
will dowhich is declaratively specify how resources are provisioned
in the Azure datacenter.
Q
STEVEN EDOUARD is a developer evangelist at Microsoft. Before that, he worked
on the .NET runtime team as a software
test engineer delivering products like the
Microsoft .NET Framework 4.5 and .NET
Native Compilation. Now his passion
resides in exciting people on the limitless
potentials of cloud computing services
through engaging technical demonstrations, online content and presentations.

Figure 9 The Final List of Virtual Machines

msdnmagazine.com

THANKS to the following Microsoft

technical expert for reviewing this
article: Bruno Terkaly
January 2015 71

Austin + San FraQcisco

NAVIGATE THE
.NET HIGHWAY
2 LRcatiRQ to ChR e frRm in JuQe!
Visual Studio Live! is hitting the open
road on the ultimate code trip to help you
navigate the .NET Highway. In June,
were pullin into Austin, TX, June 1 4
and then headed to San Francisco, CA,
June 15 18.
Both of these great locations will feature
four days of .NET developer training sessions,
workshops and networking events all
designed to make you better at your job.
Code with industry experts, get practical
answers to your challenges, and immerse
yourself in what's to come on the
.NET horizon. The full agenda and speaker
line-up will be available for both events
in early 2015. Stay tuned!

SUPPORTED BY

PRODUCED BY

GOLD SPONSOR

magazine

Untitled-4 2

12/10/14 4:30 PM

Austin

TRACKS
INCLUDE:

JUNE 1 - 4, 2015
HYATT REGENCY, AUSTIN, TX
vslive.com/austin

Visual Studio / .NET

JavaScript/HTML5

ASP.NET

Windows 8.1/WinRT

Cross-Platform Mobile Development

Database and Analytics

Cloud Computing

SharePoint/Ofce

REGISTER NOW
AND SAVE $300!
Scan the QR code to
register or for more
event details.

San FraQcisco
JUNE 15 - 18, 2015
THE FAIRMONT, SAN FRANCISCO, CA
vslive.com/sf

Use promo code VSLCODE2

vslive.com

JRin u Rn the UltiPate CRde Trip in 2015!

March 16-20

June 1-4

June 15-18

August 10-14

Sept. 28- Oct. 1

Nov. 16-20

FOLLOW US
twitter.com/vslive @VSLive

Untitled-4 3

facebook.com Search VSLive

linkedin.com Join the

Visual Studio Live group!

12/10/14 4:31 PM

TEST RUN

JAMES MCCAFFREY

Logistic Regression Classification

with Multi-Swarm Optimization
One of the most fundamental forms of machine learning
is logistic regression (LR) classification. The goal of LR
classification is to create a model that predicts a variable
that can take on one of two possible values. For example,
you might want to predict which of two candidates a
voter will select (Smith = 0, Jones = 1) based on the
voters age (x1), sex (x2) and annual income (x3).
If Y is the predicted value, an LR model for this problem would take the form:
z = b0 + b1(x1) + b2(x2) + b3(x3)
Y = 1.0 / (1.0 + e^-z)
Here, b0, b1, b2 and b3 are weights, which are just
numeric values that must be determined. In words, you
compute a value z that is the sum of input values times
b-weights, plus a b0 constant, then feed the z value to
the equation that uses math constant e. It turns out that
Y will always be between 0 and 1. If Y is less than 0.5, you
conclude the output is 0 and if Y is greater than 0.5 you
conclude the output is 1.
For example, suppose a voters age is 32, sex is male (-1),
and annual income in tens of thousands of dollars is 48.0.
And suppose b0 = -9.0, b1 = 8.0, b2 = 5.0, and b3 = -5.0.
Then z = -9.0 + (8.0)(32) + (5.0)(-1) + (-5.0)(48.0) = 2.0 and
so Y = 1.0 / (1.0 + e^-2.0) = 0.88.
Because Y is greater than 0.5, youd conclude the voter will
pick candidate 1 (Jones). But where do the b-weight values
come from? Training an LR classifier is the process of finding the values for the b-weights. The idea is to use training
data that has known output values and then find the set of b
values so that the difference between the computed output
Figure 1 Logistic Regression with Multi-Swarm Optimization in Action
values and the known output values is minimized. This is a
of birds. MSO maintains several swarms that interact with each
math problem thats often called numerical optimization.
There are about a dozen major optimization algorithms used in other, as opposed to PSO, which uses just one swarm.
A good way to see where this article is headed, and to get an idea
machine learning. For logistic regression classification training, two
of the most common algorithms are called iterative Newton-Raphson of what LR with MSO is, is to take a look at the demo program in
and L-BFGS. In this article, I present a technique called multi-swarm Figure 1. The goal of the demo program is to use MSO to create
optimization (MSO). MSO is a variation of particle swarm optimiza- an LR model that predicts Y for a set of synthetic (programmatition (PSO). In MSO, a virtual particle has a position that corresponds cally generated) data.
The demo program begins by creating 10,000 random data items
to a set of b-weight values. A swarm is a collection of particles that
move in a way inspired by group behavior such as the flocking where there are five predictor variables (often called features in ML
terminology). Each feature value is between -10.0 and +10.0, and
the Y-value, 0 or 1, is in the last column of the data set. The feature
Code download available at msdn.microsoft.com/magazine/msdnmag0115.
values dont correspond to a real problem.
74 msdn magazine

The 10,000-item data set is randomly split into an 8,000-item training set used to create the LR model, and a 2,000-item hold-out test set
used to evaluate the accuracy of the model after training. The demo
program creates an LR classifier and then uses four swarms, each
with three particles, to train the classifier. MSO is an iterative process
and the maximum number of iterations, maxEpochs, is set to 100.

A swarm is a collection of
particles that move in a way
inspired by group behavior such
as the flocking of birds.
The demo displays the best (smallest) error found by any particle,
every 10 epochs. After training is completed, the best weights found
are (4.09, 10.00, 8.43, 1.86, -9.27, 1.84). Using these weights, the
accuracy of the LR model is computed for the training data (99.98
percent correct, which is 7,998 out of 8,000) and for the test data
(99.85 percent correct, which is 1,997 out of 2,000). The accuracy
of the model with the test data gives you a rough approximation
of how well the model would do when presented with new data
that has unknown output values.
This article assumes you have at least intermediate programming skills, but doesnt assume you know anything about logistic
regression classification or multi-swarm optimization. The demo

program is coded using C#, but you shouldnt have too much
difficulty refactoring the code to another language such as Visual
Basic .NET or Python.
The demo code is too long to present here in its entirety, but
the complete source code is available in the code download that
accompanies this article. The demo code has all normal error
checking removed to keep the main ideas as clear as possible and
the size of the code small.

Overall Program Structure

The overall program structure, with a few minor edits to save space, is
presented in Figure 2. To create the demo, I launched Visual Studio
and created a new C# console application named LogisticWithMulti.
The demo has no significant Microsoft .NET Framework dependencies, so any recent version of Visual Studio will work.
After the template code loaded into the Visual Studio editor, in the
Solution Explorer window I renamed file Program.cs to the more
descriptive LogisticMultiProgram.cs and Visual Studio automatically renamed class Program for me. At the top of the source code, I
deleted all using statements that pointed to unneeded namespaces,
leaving just the reference to the top-level System namespace.
The program class has helper methods MakeAllData, MakeTrainTest, ShowData and ShowVector. All of the logistic regression
logic is contained in a single LogisticClassifier class. The LogisticClassifier class contains nested helper classes Particle, Swarm, and
MultiSwarm to encapsulate the MSO data and logic used during
training. These helper classes couldve been defined outside of the
LogisticClassifier class.

Figure 2 Overall Program Structure

using System;
namespace LogisticWithMulti
{
class LogisticMultiProgram
{
static void Main(string[] args)
{
Console.WriteLine("Begin demo");

Console.WriteLine("Setting numParticles = " + numParticles);

Console.WriteLine("Setting maxEpochs = " + maxEpochs);
Console.WriteLine("\nStarting training");
double[] bestWeights = lc.Train(trainData, maxEpochs,
numSwarms, numParticles);
Console.WriteLine("Training complete");
Console.WriteLine("Best weights found:");
ShowVector(bestWeights, 4, true);

int numFeatures = 5;
int numRows = 10000;
int seed = 0;

double trainAcc = lc.Accuracy(trainData, bestWeights);

Console.WriteLine("Accuracy on training data = " +
trainAcc.ToString("F4"));

Console.WriteLine("Generating " + numRows +

" artificial data items with " + numFeatures + " features");
double[][] allData = MakeAllData(numFeatures, numRows, seed);
Console.WriteLine("Done");

double testAcc = lc.Accuracy(testData, bestWeights);

Console.WriteLine("Accuracy on test data = " +
testAcc.ToString("F4"));

Console.WriteLine("Creating train and test matrices");

double[][] trainData;
double[][] testData;
MakeTrainTest(allData, 0.80, seed, out trainData, out testData);
Console.WriteLine("Done");

Console.WriteLine("End demo");
Console.ReadLine();
} // Main
static double[][] MakeAllData(int numFeatures,
int numRows, int seed) { . . }

Console.WriteLine("Training data: ");

ShowData(trainData, 4, 2, true);

static void MakeTrainTest(double[][] allData, double trainPct, int seed,

out double[][] trainData, out double[][] testData) { . . }

Console.WriteLine("Test data: ");

ShowData(testData, 3, 2, true);

static void ShowData(double[][] data, int numRows,

int decimals, bool indices) { . . }

Console.WriteLine("Creating Logistic Regression classifier");

LogisticClassifier lc = new LogisticClassifier(numFeatures);
int numSwarms = 4;
int numParticles = 3;
int maxEpochs = 100;
Console.WriteLine("Setting numSwarms = " + numSwarms);

msdnmagazine.com

static void ShowVector(double[] vector, int decimals,

bool newLine) { . . }
} // Program
public class LogisticClassifier { . . }
} // ns

January 2015 75

The Main method has a lot of WriteLine noise. The key calling
statements are quite simple. The synthetic data is generated, like so:
int numFeatures = 5;
int numRows = 10000;
int seed = 0; // Gives representative demo
double[][] allData = MakeAllData(numFeatures, numRows, seed);

Method MakeAllData creates random b-weight values between

-10.0 and +10.0, then for each data item, random x-values between
-10.0 and +10.0 are generated and combined with the b-weight values, which are then used to generate Y-values. The synthetic data
corresponds to a data set where the x-values have been normalized,
and where there are roughly equal counts of Y = 0 and Y = 1 values.
The data is split into training and test sets with these statements:
double[][] trainData;
double[][] testData;
MakeTrainTest(allData, 0.80, seed, out trainData, out testData);

The LR model is created and trained with these statements:

LogisticClassifier lc = new LogisticClassifier(numFeatures);
int numSwarms = 4;
int numParticles = 3;
int maxEpochs = 100;
double[] bestWeights = lc.Train(trainData, maxEpochs, numSwarms, numParticles);

And the accuracy of the model is evaluated with these two statements:
double trainAcc = lc.Accuracy(trainData, bestWeights);
double testAcc = lc.Accuracy(testData, bestWeights);

Understanding the MSO Algorithm

Heres the MSO optimization in high-level pseudo-code:
for-each swarm
initialize each particle to a random position
end-for
for-each swarm
for-each particle in swarm
compute new velocity
use new velocity to compute new position
check if position is a new best
does particle die?
does particle move to different swarm?
end-for
end-for

particle in any swarm. Term w is a constant called the inertia factor. Terms c1, c2 and c3 are constants that establish a maximum
change for each component of the new velocity. Terms r1, r2 and r3
and random values between 0 and 1 that provide a randomization
effect to each velocity update.

When a particle dies,

its replaced with a new particle
at a random position.
Suppose a particle is currently at (20.0, 30.0) and its current
velocity is (-1.0, -3.0). Also, the best-known position of the particle
is (10.0, 12.0), the best-known position of any particle in the swarm is
(8.0, 9.0), and the best-known position of any particle in any swarm
is (5.0, 6.0). And suppose that constant w has value 0.7, constants
c1 and c2 are both 1.4, and constant c3 is 0.4. Finally, suppose random values r1, r2 and r3 are all 0.2.
The new velocity of the particle (with rounding to one decimal) is:
v(t+1) = 0.7 * (-1.0, -3.0) +
(1.4 * 0.2) * ((10.0, 12.0) - (20.0, 30.0)) +
(1.4 * 0.2) * ((8.0, 9.0) - (20.0, 30.0)) +
(0.4 * 0.2) * ((5.0, 6.0) - (20.0, 30.0))
= 0.7 * (-1.0, -3.0) +
0.3 * (-10.0, -18.0) +
0.3 * (-12.0, -21.0) +
0.1 * (-15.0, -24.0)
= (-8.8, -16.2)

return best position found

The key part of the MSO algorithm is computing a

particles velocity, which is just a set of values that control to where a particle will move. For example, for a
problem with just two x-dimensions, if a particle is at
(6.0, 8.0) and the velocity is (-2.0, 1.0), the particles
new position will be at (4.0, 9.0).
Velocity is calculated so that a particle tends to move
in its current direction; tends to move toward its best
position found to date; tends to move toward the best
position found by any of its fellow swarm members;
and tends to move toward the best position found by
any particle in any swarm.
In math terms, if x(t) is a particles position at time
t, then a new velocity, v(t+1) is calculated as:
v(t+1) = w * v(t) +
(c1 * r1) * (p(t) - x(t)) +
(c2 * r2) * (s(t) - x(t)) +
(c3 * r3) * (g(t) - x(t))
Term p(t) is the particles best-known position. Term
s(t) is the best position of any particle in the particles
swarm. Term g(t) is the global best position of any
76 msdn magazine

Figure 3 The Multi-Swarm Optimization Algorithm Illustrated

Test Run

And so the particles new position is:

x(t+1) = (20.0, 30.0) + (-8.8, -16.2)
= (11.2, 13.8)
The graph in Figure 3 illustrates the MSO process for a problem with two x-values, three swarms with five particles each, and
where the optimal position is at (0, 0). The graph shows how the
first particle in each swarm closes in on the optimal position. The
spiral motion is characteristic of MSO.
In the MSO pseudo-code, a particle can die with some low probability. When a particle dies, its replaced with a new particle at a
random position. A particle can immigrate with some low probability.
When immigration occurs, a particle is swapped with another particle
from a different swarm. The death and immigration mechanisms add
an element of randomness to MSO and help prevent the algorithm
from getting stuck in a non-optimal solution.

...

Implementing Logistic Regression Using MSO

int[] sequence = new int[numParticles];

for (int i = 0; i < sequence.Length; ++i)
sequence[i] = i;

The definition of method Train begins as:

public double[] Train(double[][] trainData, int maxEpochs,
int numSwarms, int numParticles)
{
int dim = numFeatures + 1;
double minX = -10.0;
double maxX = 10.0;
MultiSwarm ms = new MultiSwarm(numSwarms, numParticles, dim);
...

The dimension of the problem is the number of predictor

features, plus one, to account for the b0 constant. Variables minX
and maxX hold the smallest and largest values for any single value
in a Particle objects position array. The LogisticClassifier class contains a nested MultiSwarm class. The MultiSwarm class constructor
creates an array-of-arrays of Particle objects, each of which initially
has a random position, and a random velocity. Because the logistic
regression Error method isnt directly visible to the nested Particle
definition, the MultiSwarm constructor doesnt supply error value
for each particle, so method Train adds error information. First,
each particle gets an error value:
for (int i = 0; i < numSwarms; ++i)
{
for (int j = 0; j < numParticles; ++j)
{
Particle p = ms.swarms[i].particles[j];
p.error = Error(trainData, p.position);
p.bestError = p.error;
Array.Copy(p.position, p.bestPosition, dim);
...

Next, the current swarms best error and the overall global best
error are computed:
Figure 4 New Velocity for Current Particle Is Calculated
for (int k = 0; k < dim; ++k)
{
double r1 = rnd.NextDouble();
double r2 = rnd.NextDouble();
double r3 = rnd.NextDouble();
p.velocity[k] = (w * p.velocity[k]) +
(c1 * r1 * (p.bestPosition[k] - p.position[k])) +
(c2 * r2 * (ms.swarms[i].bestPosition[k] - p.position[k])) +
(c3 * r3 * (ms.bestPosition[k] - p.position[k]));
if (p.velocity[k] < minX)
p.velocity[k] = minX;
else if (p.velocity[k] > maxX)
p.velocity[k] = maxX;
} // k

msdnmagazine.com

if (p.error < ms.swarms[i].bestError) // Swarm best?

{
ms.swarms[i].bestError = p.error;
Array.Copy(p.position, ms.swarms[i].bestPosition, dim);
}
if (p.error < ms.bestError) // Global best?
{
ms.bestError = p.error;
Array.Copy(p.position, ms.bestPosition, dim);
}
} // j
} // i

The main training loop is prepared, like so:

int epoch = 0;
double w = 0.729; // inertia
double c1 = 1.49445; // particle
double c2 = 1.49445; // swarm
double c3 = 0.3645; // multiswarm
double pDeath = 1.0 / maxEpochs;
double pImmigrate = 1.0 / maxEpochs;

The values for constants w, c1 and c2 are the result of some PSO
research. Interestingly, compared to many numerical optimization
algorithms, PSO and MSO are relatively insensitive to the values
used for internal magic constants (called free parameters or hyper
parameters). Theres little available research for the c3 constant,
which influences the tendency of a particle to move toward the
best-known position found to date by any particle in any swarm.
The value I use, 0.3645, is half the value of the inertia constant, and
has worked well for me in practice.
The sequence array holds indices of the training data. This array
will be scrambled so that in each swarm, particles will be processed
in a different order in each iteration of the main training loop.
The main training loop begins:
while (epoch < maxEpochs)
{
++epoch;
// Optionally print best error here
for (int i = 0; i < numSwarms; ++i) // Each swarm
{
Shuffle(sequence);
for (int pj = 0; pj < numParticles; ++pj) // Each particle
{
int j = sequence[pj];
Particle p = ms.swarms[i].particles[j];
...

Knowing when to stop training is one of the most difficult challenges in machine learning. Here, a fixed number, maxEpochs, of
iterations is used. This is a simple approach, but theres a risk you
might not train enough. Or you might train too much, which would
result in a model that fits the training data extremely well, but works
poorly on new data. This is called over-fitting. There are dozens of
techniques that can be used to combat over-fitting.
Next, the new velocity for the current particle is calculated, as
explained earlier (see Figure 4).
After the velocity is calculated, each of its component values
is checked to see whether the magnitude is large, and if so, the
value is reined back in. This prevents a particle from moving a
very great distance in any one iteration. For some ML training
tasks, eliminating the velocity constraint appears to speed up
the training, but there are no solid research results to give any
specific advice here.
January 2015 77

Next, the new velocity is used to calculate the current particles

new position and associated error:
for (int k = 0; k < dim; ++k)
{
p.position[k] += p.velocity[k];
if (p.position[k] < minX)
p.position[k] = minX;
else if (p.position[k] > maxX)
p.position[k] = maxX;
}

After the current particle has moved, its new error is calculated:
p.error = Error(trainData, p.position); // Expensive
if (p.error < p.bestError) // New best position for particle?
{
p.bestError = p.error;
Array.Copy(p.position, p.bestPosition, dim);
}

The current particles new error might be a new swarm best or

global best:
if (p.error < ms.swarms[i].bestError) // New best for swarm?
{
ms.swarms[i].bestError = p.error;
Array.Copy(p.position, ms.swarms[i].bestPosition, dim);
}
if (p.error < ms.bestError) // New global best?
{
ms.bestError = p.error;
Array.Copy(p.position, ms.bestPosition, dim);
}

After the current particle has moved, an option is to possibly kill

that particle and generate a new one. If a random value is below
some small threshold, a new particle is generated:
double p1 = rnd.NextDouble();
if (p1 < pDeath)
{
Particle q = new Particle(dim); // A replacement
q.error = Error(trainData, q.position);
Array.Copy(q.position, q.bestPosition, dim);
q.bestError = q.error;

Instead of using a fixed probability of death, an alternative is to

gradually increase the probability of death, for example:
double pDeath = (maxProbDeath / maxEpochs) * epoch;

After the replacement particle has been created, that particle may
be, by pure luck, a new swarm best or global best:
if (q.error < ms.swarms[i].bestError) // Best swarm error by pure luck?
{
ms.swarms[i].bestError = q.error;
Array.Copy(q.position, ms.swarms[i].bestPosition, dim);
if (q.error < ms.bestError) // Best global error?
{
ms.bestError = q.error;
Array.Copy(q.position, ms.bestPosition, dim);
}
}

The death mechanism concludes by exchanging the current particle for the replacement, effectively killing the current particle:
...
ms.swarms[i].particles[j] = q;
} // Die

Next, the (optional) immigration mechanism starts with:

double p2 = rnd.NextDouble();
if (p2 < pImmigrate)
{
int ii = rnd.Next(0, numSwarms); // rnd swarm
int jj = rnd.Next(0, numParticles); // rnd particle
Particle q = ms.swarms[ii].particles[jj]; // q points to other
ms.swarms[i].particles[j] = q;
ms.swarms[ii].particles[jj] = p; // the exchange
...

The immigration mechanism is quite crude. The code selects

a random swarm and then a random particle from that swarm,
78 msdn magazine

and then swaps the random particle with the current particle. As
coded, theres no guarantee the randomly selected particle will be
in a swarm different from the current particle.
The immigration mechanism finishes by checking to see if swapping
particles in two swarms resulted in one or two new swarm-best positions:
...
if (q.error < ms.swarms[i].bestError) // Curr has new position
{
ms.swarms[i].bestError = q.error;
Array.Copy(q.position, ms.swarms[i].bestPosition, dim);
}
if (p.error < ms.swarms[ii].bestError) // Other has new position
{
ms.swarms[ii].bestError = p.error;
Array.Copy(p.position, ms.swarms[ii].bestPosition, dim);
}
} // Immigrate

Knowing when to stop training

is one of the most difficult
challenges in machine learning.
Method Train concludes by returning the best position found
by any particle:
...
} // j - each particle
} // i - each swarm
} // while
return ms.bestPosition;
} // Train

The return value is a reference to an array representing a set of

weights of logistic regression. A minor alternative to reduce the
chances of an unwanted side effect is to copy the values to a local
array and return a reference to the local array.

Wrapping Up
This article and accompanying code should get you up and running if
you want to explore logistic regression classification using multi-swarm
optimization for training. MSO is really more of a meta-heuristic than
an algorithm. By that I mean MSO is a set of design principles that
can be implemented in many different ways. There are dozens of
modifications to the basic MSO design you can investigate.
Logistic regression is one of the simplest forms of machine
learning classification. For many classification problems, LR just
doesnt work well. However, many ML practitioners, including me,
usually begin investigating a classification problem using LR, and
then use more sophisticated techniques, if necessary.
Compared to older, calculus-based numerical optimization
techniques such as gradient descent and L-BFGS, based on my
experience, training using MSO tends to produce better ML models, but MSO is almost always an order of magnitude slower. Q
DR. JAMES MCCAFFREY works for Microsoft Research in Redmond, Wash. He has
worked on several Microsoft products including Internet Explorer and Bing. Dr.
McCaffrey can be reached at jammc@microsoft.com.

THANKS to the following Microsoft technical experts for reviewing this article:
Todd Bello and Alisson Sol
Test Run

Untitled-7 1

12/8/14 4:44 PM

DONT GET ME STARTED

DAVID S. PLATT

An Upstart Again
Its interesting to watch Microsoft as an underdog.
Not that long ago that underdog bestrode the computing world
like a colossus. You probably remember the joke, How many Microsoft programmers does it take to change a light bulb? The answer,
of course, was: None. Bill Gates just declares Microsoft Darkness
to be the new standard.
But while Microsoft concentrated on PCs (see msdn.microsoft.com/
magazine/hh224515), new business sectors exploded across the IT industry, namely mobile devices and the cloud. The most recent figures show Microsoft doing well in the latter, which can be seen as
the logical progression of the server side of its PC business.
Microsoft is having much more difficulty in the former.
As I wrote in my November 2013 column (msdn.microsoft.com/magazine/dn463796), Microsofts key to greater market share is making
more and better apps available in its ecosystem. This in turn means

The integration of Xamarin

with Visual Studio works exactly
the way existing Visual Studio
users would expect, and their
skills transfer about as seamlessly
as they could.
harnessing Microsofts army of developers [who] would love
to use their current skills to develop apps for this brave new device-and-tablet world. These developers need to support iOS and
Android. Theyd support Windows 8, too, if it didnt take much extra effort. Microsoft should develop a toolkit that covers all three
from the same code base, I wrote at the time. Either do it yourself
or buy Xamarin as a jump-start. Do it quickly, though, before the
army gets hungry and switches sides.
The purchase of Xamarin hasnt happened, but the companies
have moved closer. Last Aprils Build 2014 conference showcased
Xamarin front and center. And the recent Xamarin Evolve 2014
summit in Atlanta sold out in advance.
80 msdn magazine

The integration of Xamarin with Visual Studio works exactly

the way existing Visual Studio users would expect, and their skills
transfer about as seamlessly as they could. It hasnt yet reached the
level of, Click one button, and out come iOS, Android and Windows versions. The three major mobile platforms are different
enough that you cant abstract them completely. But Xamarin helps
you separate your mobile apps business logic, which doesnt vary
from one platform to another, from the presentation, which does.
And it lets you develop the whole package with your familiar languages, such as C#, and tools such as debuggers. Just at press time,
the company announced a new program making Xamarin free
for students. Its moving in the right direction, and pretty quickly.
Xamarin asked me a few years ago if I was interested in starting
up its developer training effort. I needed more cash than the company was willing to part with at the time, so the discussion never
progressed, but I took its online training class as a student last spring
and found it quite good. And if the instructors didnt quite have
my flair (and lets face it, who does?), there were certainly more of
them available live on multiple days in different time zones than
just one of me. Now I sort of wish Id held the price down to where
Xamarin could swing it.
As if to trumpet its arrival in the big leagues, Xamarin has
acquired former MSDN Magazine columnist Charles Petzold as
an in-house guru, describing him in the press release announcing
the hiring as a Turing-complete engine that converts APIs into
books (see bit.ly/1o9tfi7). His forthcoming book, Creating Mobile
Apps with Xamarin.Forms, is due in the spring of 2015. I learned
Windows from his famous book (Programming Windows) back
in 1990. Im looking forward to this one.
In Why Software Sucks (see pp. 193-194), I wrote about
the early 1990s, when Microsoft was seen as the brash upstart,
upending the established order of stodgy mainframes. I miss
those days, as I miss my now-teenage daughters first steps and
first school and first sleepover parties. Im looking forward to
seeing some of that again as a grandparent. Is that what Im seeing
in the device world today, the now-grown child Microsoft, and
the grandchild Xamarin?
Q
DAVID S. PLATT teaches programming .NET at Harvard University Extension School
and at companies all over the world. Hes the author of 11 programming books,
including Why Software Sucks (Addison-Wesley Professional, 2006) and Introducing Microsoft .NET (Microsoft Press, 2002). Microsoft named him a Software
Legend in 2002. He wonders whether he should tape down two of his daughters
fingers so she learns how to count in octal. You can contact him at rollthunder.com.

Untitled-10 1

12/5/14 1:59 PM

Untitled-7 1

11/5/14 4:36 PM

 6

/QPFC[(TKFC[/CTEJ

5

7%

%'

55

(7

.

;'

#4

%GPVTCN*CNN9GUVOKPUVGT.QPFQP

THE DEFINITIVE DEVELOPER CONFERENCE

9QTNFTGPQYPGF
URGCMGTUKPENWFKPI

ALLEN HOLUB

. H H S  \ R X U 
Z K R O H  G H Y H O R S P H Q W
W H D PV  V N L O O V  X S 
W R  G D W H  Z L W K  W K H 
O D W H V W  W H F K Q R O R J L H V  
E H V W  S U D F W L F H V 
DQGIUDPHZRUNV

KEVLIN HENNEY

DINO ESPOSITO

NEAL FORD

(7..&#;914-5*125
/+076'$4'#-1765'55+105
CPFOCP[OQTG
)1.&52105145

8
 EEA8O<H?:7O
)&@7DK7HO(&'+
I7L;KFJEw(&&

52105145

/'&+#2#460'45

7]_b[| 7hY^_j[Yjkh[ | 8?| 8_]:WjW| 9bekZ| :WjWXWi[| :[lEfi| ?eJ| B[WZ[hi^_f| CeX_b[ | CG| CIJ[Y^
FWjj[hdi| Fhe]hWcc_d]BWd]kW][i%J[Y^d_gk[i| I[Ykh_jo| Ie\jmWh[:[i_]d| J[ij_d]| K?%KN| M[X
Y Y Y F G X Y G G M  E Q O

DevWeek_2015_Brochure.indd 1

1/5/15 10:52 AM

&'89''-

/10&#;4&/#4%*
/#4%*

DONT MISS THE UKS LEADING

DEVELOPER CONFERENCE
&GX9GGMKUVJG7-UPWODGTQPGFGUVKPCVKQPHQT
UQHVYCTGFGXGNQRGTUCTEJKVGEVUCPFCPCN[UVU9KVJKPUKIJVU
QPVJGNCVGUVVGEJPQNQIKGUDGUVRTCEVKEGCPFHTCOGYQTMU
HTQOKPFWUVT[NGCFKPIGZRGTVURNWUJCPFUQPYQTMUJQR
UGUUKQPU&GX9GGMKU[QWTEJCPEGVQUJCTRGP[QWTUMKNNU
CPFGPUWTGGXGT[OGODGTQH[QWTVGCOKUWRVQFCVG

'
IXOOGD\ZRUNVKRSV
PLQXWH 
EUHDNRXWVHVVLRQV
H[SHUWVSHDNHUV
3OXVNH\QRWHSUHVHQWDWLRQV
IURP#NNGP*QNWDDQG 
-GXNKP*GPPG[

W
n

S
O
G
s
y
m
a
p
s
e

8'07'

?DJ>;>;7HJE<BED:ED
For five full days, from Monday 23 March
to Friday 27 March, DevWeek takes over
Westminsters iconic Central Hall.

Built in 1912, Central Hall is one of the UKs

oldest purpose-built conference centres.
Offering easy access to public transport and the
restaurants and theatres of the West End, the
venue fuses architectural grandeur with stateof-the-art facilities making it the ideal setting
for the UKs premier developer conference.
For more information about Central Hall,
visit www.c-h-w.com

D
o
d
L

D
d
o
a
c

0'#4'5667$'56#6+105
Westminster
(District, Circle & Jubilee)
St Jamess Park
(District & Circle)

T
a
c
+
c

2 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 2

"&GX9GGM

1/5/15 10:52 AM

V

'0,1;;174%10('4'0%';1749#;
With shareable tickets and online catch-up, DevWeek gives you all the information you
need, when you want it ensuring you and your team make the most of every session.
Share your ticket
Only have time to attend one day?
Getgreat value out of DevWeek by
sharing your ticket with others in
yourteam. That way, you can make the
most of five packed days of sessions
and workshops. Our online registration
page lets you add colleagues details
so we know who will be joining us for
eachday of DevWeek.

Topics
Agile
Architecture
BI
Big Data
Cloud
Database
DevOps
IoT
Leadership
Mobile

Never miss a session

So much content is packed into
DevWeeks five days, there are bound
tobe times when you wish you could
bein two places at once. But you
neednt miss out: all our sessions are
filmed (subject to speaker approval)
and as a registered delegate, youll
have exclusive access to the whole
event online to watch when you want.

MQ
MS Tech
Patterns
Programming Languages /
Techniques
Security
Software Design
Testing
UI/UX
Web

D

52105145*+2':*+$+6+10

he

eng

Do you provide services

or technologies to the
developer community?
Let DevWeek work for you
DevWeek is the UKs leading event for software
developers, DBAs and IT architects. As an exhibitor
or sponsor, we can can help you reach an engaged
audience of professionals with a package thats
custom-designed to suit your needs.
To discuss the wide range of sponsorship
and exhibition opportunities available,
contactChris Handsley
+44 (0)207 830 3634
chris.handsley@publicis-blueprint.co.uk

)1.&52105145

M>EM?BBOEKC;;J7J:;LM;;A5
Consultant

DBA

C-Level, Director
Developer,
Engineer,
Programmer

Analyst
Where are they from?

Architect

UNITED
KINGDOM

EUROPE
ROW

Development
Head, Manager,
Team Lead

Senior / Lead / Principle

Developer, Engineer, Programmer

52105145

/'&+#2#460'45

"&GX9GGM | YYYFGXYGGMEQO | 3

DevWeek_2015_Brochure.indd 3

1/5/15 10:52 AM

/10&#;/#4%*

24'%10('4'0%'914-5*125

&GX9GGMU
RTQITCOOG
QHRTG
EQPHGTGPEG
YQTMUJQRU
EQXGTUC
YKFGTCPIGQH
UWDLGEVUHTQO
#IKNGU[UVGOU
VQCRR
FGXGNQROGPV



$//(1
+2/8%

7=?B;%:;I?=D<HEC
IJ7HJJE<?D?I>

#NNYQTMUJQRU
TWPHQTC
HWNNFC[
HTQO
VQ
YKVJUJQTV
DTGCMUKPVJG
OQTPKPICPF
CHVGTPQQP
CPFC
NWPEJDTGCM
CV

( [ F H O O H Q W  P L [
RIVXEMHFWV
Z R X O G  U H F R P P H Q G
XQUHVHUYHGO\
IE<JM7H;:;L;BEF;H
(&'*:;B;=7J;

4 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 4

3($5/
&+(1

?DJ;HD;JJ?D=
OEKHJ>?D=I

6$+,/
0$/,.

@7L7I9H?FJ0
8;OED:J>;
87I?9I

Agile systems are, by

necessity, tightly coupled
to the users notion of
what the system is doing.
Without that connection,
the software cant stand
up to the stress of
constant change that all
Agile processes mandate.
Moreover, the process
you use influences the
architecture of your
system. Thats why hybrid
processes that mix Agile
and traditional practices
often fail. The hybrid
architectures that come
out of those processes
are unworkable.
In this workshop,
Allen will talk about both
process and architecture.
Well look at how Agile
processes work, and see
how both architecture
and low-level design
naturally fall out of those
processes. Specifically,
well examine the role of
stories: how theyre
created and developed,
and how they flow
through the process,
with a focus on
developing an optimal
architecture that closely
mirrors both the stories
themselves and the
assumptions that
underlie the stories.
Well also work
through a real-world
example of the process
from requirements
gathering and problemstatement definition,
tostory development
(use-case analysis)
andthe simultaneous
construction of
lightweight dynamic
andstatic models that
underlie the code.

From big consumer

success stories, such as
the Nest Thermostat, to
more offbeat monitoring
systems, such as
Botanicalls (which lets
your plant call you when
it is thirsty), your things
are finding their own
voice through small but
powerful embedded
microcontrollers. The
market for the Internet
of Things (IoT) and
wearables is exploding.
But what are your
options for getting
started with making
physical things when
youre more used to
writing software? In this
workshop, Pearl explores
a few hardware options
that are great for hobbyist
and rapid prototyping.
Youll get hands-on
time to choose from
some of the available
hardware platforms,
from a standard Arduino
to an Intel Edison,
MaKey MaKey, or
LightBlue Bean. Youll be
given a self-paced guide
on how to set up the
development
environment on your
computer and get a
Hello World program
running (typically an
onboard blinking light).
Once you have a feel for
the technology, well
gothrough a project
brainstorming exercise.
Finally, youll be let loose
to build your own
hardware-based
prototype in groups of
2-4. The workshop will
be capped at 30
participants so expect a
very team-oriented day.

In this workshop, Sahil

will teach you advanced
concepts in JavaScript,
allowing you to structure
your code well and learn
JavaScript beyond the
basics we have been
tinkering with.
This workshop
teaches you not to just
know JavaScript, but to
be good at JavaScript.
There are plenty of
hands-on labs, which
will walk you through
objects, prototypes,
scopes, this variable,
debugging, performance
and best practices.
We will begin with the
very basics of JavaScript:
variables, language
syntax, referencing
files, loops, conditions,
built-in functions and
custom functions,
arrays and so on. Well
move on to talk about
functions as expressions,
closures to structure
your code, the concept
of nested scopes, and the
confusion around this.
Well learn all about
objects and prototypes,
and how you can mimic
things in JavaScript that
are usually reserved for
higher-level languages.
Well solidify that
knowledge with some
best practices and
debugging tricks. Well
also explain how to avoid
common pitfalls, and
how to organise your
code in modules to keep
it maintainable and
understandable. Lastly,
theres performance
lets not forget
performance!

914-5*12
4'(&9

914-5*12
4'(&9

914-5*12
4'(&9

-$0(6
0217(0$*12

H;KI?D=OEKH
$D;J7D:9
IA?BBIJE:;L;BEF
D7J?L;7FFI<EH
M?D:EMI"?EI
7D:7D:HE?:

With .NET, Xamarin

and portable class
libraries, you can now
create native apps that
target iOS, Android
and Windows without
compromising on
performance, user
experience or developer
productivity. And you
can do it all within Visual
Studio IDE or Xamarin
Studio IDE forMac or PC.
James will introduce
the Xamarin platform,
including building native
iOS and Android apps
with C#. Well look
at the fundamentals
of how each platform
works, and deep-dive
into platform-specific
functionality. Designing
iOS and Android apps
could not be easier with
Xamarins integrated
iOS and Android
designers for both Visual
Studio and Xamarin
Studio. Youll find out
how to use both of these
designers to craft unique
user interfaces for each
platform. Well also take
a look at sharing code
with Universal Windows
apps, enabling you reach
all platforms from a single
shared C# code base.
Youll gain a full
understanding of the
Xamarin platform and
how to build iOS and
Android apps using
C#,as well as a solid
introduction to code
reuse techniques, plus
lots of sample code
totake home!

914-5*12
4'(&9

J>
H;

Re
col
sta
the
we
the
Bu
wo
so

Ro
wi
con
des
rev
of t
the
sor
per
thr
for
We
to u
ma
an
beh
top
int
def
un

wo
ple
ho
eff
rev
en

91
4'

"&GX9GGM

1/5/15 10:52 AM

C.

#.51'#%*&#;
&2))(( 5(*,675$7,21

52%(5760$//6+,5(
$867,1%,1*+$0

J>;FEM;HE<
H;L?;M

Review by peers,
colleagues, experts and
stakeholders is perhaps
the most effective tool
we have for improving
the quality of software.
But if review is so
wonderful, why is it used
so infrequently?
In this workshop,
Robert and Austin
will show you how to
conduct effective code,
design and requirements
reviews using a variety
of techniques from
the relatively informal
sort of reviews youre
perhaps doing already,
through to the most
formal inspections.
Well work together
to understand what
makes a good review,
and help you to identify
behaviours that lead
topoor outcomes,
inthe form of either
defective software or
unhappy colleagues.
Throughout this
workshop, youll receive
plenty of advice on
howyou can introduce
effective technical
reviews into your
engineering culture.

914-5*12
4'(&9

-8/(6
0$<

<0J>;B?JJB;
B7D=K7=;M?J>
7BEJE<8?J;

Since it was released

in 2008, F# has
inspired more and
more excitement. Its
a functional language,
like Haskell or ML, and
yet its built on top of
.NET, so it leverages all
the mature objectoriented power that the
platform provides. The
result is neither purely
functional, nor obviously
object-oriented, but is
something entirely new:
a unique and powerful
mixture of the two.
Those programmers
who have already
discovered F# find they
can write code that is
shorter, faster, vastly
more reliable, and
delightfully reusable.
This is no research
language its being
used on real, large-scale
projects, and it has
the backing of the F#
Software Foundation
and Microsoft.
If you want to
understand what the
fuss is about, this is
the day for you. In this
workshop, Jules will
touch on what makes F#
programming so special,
and show where it gets
its power from. Well
write some functional
code, well write some
of F#s unique take on
object-oriented code,
and well write some
code that only F# can do.

914-5*12
4'(&9

| &2))((%5($. |

,'2
)/$72:

I;9KH?D=7IF$
D;J7FFB?97J?EDI
7D:I;HL?9;I0
<HEC7#P

When you think of

ASP.NET security, the
first things that come
to mind are Windows
authentication and
forms authentication
using ASP.NET
Membership. For
years, those were the
common authentication
techniques for ASP.
NET applications and
services. But with the
new releases to the ASP.
NET Identity system,
those days are long gone.
For the enterprise,
ASP.NET broadened
its support from the
on-premises Active
Directory to include
Microsoft Azure
Active Directory. By
supporting external
identity providers, such
as Facebook, Microsoft
Account and Twitter,
the new ASP.NET
Identity system makes
the process of securing
an application less
scarythan ever.
In this workshop,
Ido will start from the
basics of getting to
know concepts such as
SSL, OAuth, OpenID
and claim-based
authorisation. From
there we will continue
to explore the various
scenarios of using
self-managed identities,
Active Directory and
ADFS, external identity
providers (Facebook,
Google, Microsoft) and
Microsoft Azure Active
Directory.

914-5*12
4'(&9

/81&+%5($. | &2))((%5($.

./$86
$6&+(1%5(11(5

',12
(6326,72

IGBI;HL;HGK;HO
JKD?D=<EH
:;L;BEF;HI

ED;:7OE<
JM?JJ;H
8EEJIJH7F

914-5*12
4'(&9

914-5*12
4'(&9

Are you a developer

writing T-SQL
queries for SQL
Server databases?
Maybe yourealready
mastered the basics
of T-SQL, but want to
reach a higher level in
T-SQL towrite better
performing queries?
Inthis workshop,
Klauswill take a full
day to talk about how
to improve your T-SQL
skills to solve complex
problems, and how to
further speed up your
queries byapplying a
good indexing strategy to
yourT-SQL queries.
In the workshop, well
cover four modules:
query processing basics
(set theory, predicate
logic, relational
models and logical
query processing),
physical query
processing (execution
plans, data access
paths, physical join
operators, aggregation
operators and spool
operators), temporary
data and aggregations
(temp tables, table
variables, common
table expressions,
and aggregations and
pivoting), and working
with windowing
functions (window
aggregate functions,
ranking functions,
distribution functions,
offset functions, query
tuning guidelines
and parallelism
optimisations).

6(%
526(

8::8O;N7CFB;

As a responsive web
framework, Twitter
Bootstrap is leading
the world of web
development today,
setting new standards
and capturing followers.
In this workshop, Dino
will provide a day-long
tour of the library and
delve deep into its
HTML chunks, CSS
styles and JavaScript
components.
Well focus on
facilities available for
building static and
responsive layouts,
rich input forms and
advanced features such
as auto-completion,
modals, tabs, carousels
and more. Well also
consider the downsides
of the framework,
missing pieces (ie. image
handling) and its overall
role in the broader
context of responsive
and device-friendly
sites. This workshop
is ideal for clearing up
a few things you may
already have heard
about Bootstrap, or just
for gaining an additional
perspective about it.
In any case, after this
workshop you should
be ready to get into it
at any level of further
complexity.

In this workshop, Seb

will provide a practical
introduction to using
examples to specify
software. Youll learn
to break down complex
business requirements
with your stakeholders,
using examples in their
own language, giving
you the tools you need
to explore their ideas
before you even write
any software.
This workshop is
for everybody involved
in the process of
developing software,
so please bring product
owners, testers and
architects along. As well
as describing what BDD
is (and isnt), well spend
a lot of time practising
collaborative analysis
to make sure that our
stories are appropriately
sized, easy to read and
unambiguous. Well
develop a ubiquitous
language, explore the
workings of the ThreeAmigos meeting, and
really get to grips with
the slippery interaction
between features,
stories, acceptance
criteria and examples.
Well use pens,
cards and other bits
of paper, so you wont
need to know any tools
in advance, or even
remember your laptop!
To build further on
these ideas, dont miss
Sebs post-conference
workshop: Applied
BDD with Cucumber,
Cucumber-JVM and
SpecFlow.

914-5*12
4'(&9

"&GX9GGM | YYYFGXYGGMEQO | 5

DevWeek_2015_Brochure.indd 5

1/5/15 10:52 AM

67'5&#;56#24+.
67'5&#;/#4%*

&GX9GGMUTUVFC[QHUGUUKQPUMKEMUQHHYKVJQWTMG[PQVGRTGUGPVCVKQPUIKXKPICNN
CVVGPFGGUVJGEJCPEGVQJGCTKPFWUVT[GZRGTVU-GXNKP*GPPG[CPF#NNGP*QNWDVCEMNG
VYQQHVJGDKIIGUVKUUWGUKPUQHVYCTGFGXGNQROGPV
&#;#)'0&#



7IOIJ;C?IDEJ7JH;;

.(9/,1
+(11(<

-';016'
24'5'06#6+10



52%(57
60$//6+,5(

KD:;HIJ7D:?D=
JH7DI:K9;HI
Transducers a
portmanteau of
transform reducers
are a new functional
programming concept
introduced into the
Clojure programming
language. Although
transducers are actually
pretty straightforward,
wrapping your brain
around them, especially
if youre not already a
competent Clojureist,
can be challenging. In
this session, Robert will
introduce transducers
by implementing
them from scratch in
everybodys favourite
executable pseudocode:
Python. Well start with
the familiar staples of
functional programming
and derive transducers
from first principles.

6 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 6

Trees. Both beautiful and useful. But were not talking about
the green, oxygen-providing ones. As abstract structures we
see trees all over the place file systems, class hierarchies,
ordered data structures, and so on. They are neat and tidy,
nested and hierarchical a simple way of organising things;
a simple way of breaking large things down into small things.
The problem is, though, that there are many things
from modest fragments of code up to enterprise-wide IT
systems that do not comfortably fit into this way of looking
at the world and organising it. Software architecture, design
patterns, class decomposition, performance, unit tests
all of these cut across the strict hierarchy of trees. In this
keynote, Kevlin will look at what this means for how we
think and design systems, whether large or small.

6$6+$
*2/'6+7(,1

7KJEC7J?D=
FHE8B;C7D7BOI?I
7D:JH?7=;
What do you do when
your application crashes
or hangs in production?
Nothing can compete
with a debugger or a full
process dump captured
on a production system.
But you cant always
afford the time to analyse
hundreds of crash
dumps. In this session,
Sasha will show you how
to perform automatic
dump analysis and triage
using Microsofts ClrMD,
a .NET library that can
explore threads, call
stacks and exceptions;
visualise threads and
locks to form wait chains
and detect deadlocks;
and walk heap memory
to inspect important
objects for your
application.

-2+1
.3$8/

;9C7I9H?FJ,
<EH7BBE<KI
Coming from a Java
background, there was
a time when JavaScript
was nothing but
annoyances. Now, even
after weve grown to love
the language, there are
dozens of times when we
feel the pain of missing
features that Java has
built in. ECMAScript 6
is changing all of that.
The next version of
JavaScript brings with
it an amazing standard
library that rivals that of
Java, Python and their
ilk. In this session, John
will explain some of
the great new additions
to the language and
demonstrate use cases
that take advantage of
ES6s elegance for clientside development.

'525
+(/3(5

D7L?=7J?D=J>;J::
7BF>78;JIEKF
TDD, BDD, ATDD are
all methodologies that
enable incremental
design that is suitable
for Agile environments.
It seems that every day a
new xDD methodology
is born with the promise
of being better than what
came before. Should
you use behaviourdriven tests or plain
old unit tests? Which
methodology is better?
And how exactly
would it benefit the
development life cycle?
In this session, Dror
will help to sort out the
various methodologies
explaining where they
came from, the tools
they use, and discussing
how and when to use
each one.

-'
24

$//(1
+2/8%

I;9KH?JO'&'07D
?DJHE:K9J?EDJE
IE<JM7H;I;9KH?JO
As more and more of
our applications move
on to the web, security
becomes even more
critical. Good security,
however, has to be built
in, not tacked on as an
afterthought. In this
session, Allen will give
you an overview of what
it means to make an
application secure. Hell
cover topics such as
security architectures,
code and design review,
penetration testing, risk
analysis and risk-based
testing, security-related
requirements, static
analysis, abuse cases,
security operations
andcrypto.

"&GX9GGM

1/5/15 10:52 AM

IG
?D#
:;
He
wo
go
OL
20
pe
tim
Kl
the
an
co
(M
on
is b
loo
tha
He
loc
lat
Se
fou
int
of
as
tra
be

#.51'#%*&#;
&2))(( 5(*,675$7,21

| &2))((%5($. |

/81&+%5($. | &2))((%5($.

-2,186)25'5,1.6$77+((1'2)'$<1(7:25.,1*'5,1.66321625('%<
DE;IJ?C7J;I

$//(1
+2/8%

-';016'
24'5'06#6+10

Estimates are always guesses and theyre always

wrong. Consequently, estimate-based planning is
foolhardy at best, and time spend creating them is a
waste. In spite of this fact, estimates are a central part
of most software-development processes, even some
Agile processes. Getting rid ofestimates doesnt
mean that you cant plan, but you do have to
go about planning in a more effective way.
Inthiskeynote, Allen willdiscuss both
$  J U H D W  P R W L Y D W L R Q D O 
the problems surrounding an estimation
L Q V S L U D W L R Q D O 
culture andhow to solve those problems
LQIRUPDWLYHZHHN
by usingactual measurements
:?H;9JEH
andpriority-based planning.
(&'*:;B;=7J;

./$86
$6&+(1%5(11(5

IGBI;HL;H(&'*
?D#C;CEHOEBJF
:;;F:?L;>;A7JED
Hekaton is the Greek
word for 100 and the
goal of In-Memory
OLTP in SQL Server
2014 is to improve query
performance up to 100
times. In this session,
Klaus will look inside
the case of Hekaton
and the multiversion
concurrency control
(MVCC) principles
on which Hekaton
is built. Hell start by
looking at the challenges
that can be solved by
Hekaton, especially
locking, blocking and
latching within SQL
Server. Based on this
foundation, hell move
into the principles
of MVCC, and how
a storage engine and
transaction manager can
be built on that concept.

'(-$1
6$5.$

?DJHE:K9?D=H7D:
7PKH;CB
R is a free software
programming
language and software
environment for
statistical computing,
data mining and
graphics. Azure
Machine Learning
(Azure ML) is the new
Microsoft cloud service
and environment for
advanced data analysis,
which utilises the R
algorithms intensively.
In this session, Dejan
will introduce both R,
with RStudio IDE, and
Azure ML.

$//$1
.(//<

:?7BE=K;I>;;JI
<EHH;JHEIF;9J?L;I
7D::?I9KII?ED
Retrospectives are a
key tool in the Agile
toolkit, but they arent
easy. In fact, its not
just retrospectives.
Teams need to learn
to talk, discuss and
reflect together over
many things. Good
conversation makes
for good software.
Retrospective
dialogue sheets can
help overcome these
problems. In this
hands-on session,
inwhich everyone
willget the chance to
work with a dialogue
sheet, discovering
whatone is and how
touse it, Allan will
discuss some of the ways
teams are using them
and look to the future.

0$5.
60,7+

:;I?=D?D=7:7FJ?L;
7FFB?97J?EDI<EH
J>;?EIFB7J<EHC
With the introduction
of the iPhone 6 and 6+,
we now have several
form factors to consider
when designing our
iOS applications. In this
session, Mark will focus
on the designer and
layout features Apple
has included in iOS to
help you create a single,
unified storyboard that
is capable of working
with all variations of
iOS. This will include
layout constraints (auto
layout), size classes,
unified storyboards
and the updated
UISplitViewController.

6(%
526(

0,&+$(/
.(11('<

B;II?ICEH;
7D?DJHE:K9J?ED
JEBEM#<?:;B?JO
7FFHE79>;I
In this session, Seb will
demonstrate some key
techniques that help
decompose large
problems. Decomposing
problems is a skill all
software developers
need, but were often not
very good at. Whether its
stories that take longer
than an iteration, or
features that cant be
delivered in the expected
release, weve all seen the
problems that tackling
an over-large problem
can cause. Well work
through two detailed
examples to demonstrate
the value of delivering
small, low-fidelity pieces
of work early rather than
prematurely focusing
onfully-polished
finalversion.

=;JJ?D=IJ7HJ;:
M?J>IM?<J7FFB;I
D;M;IJB7D=K7=;
Swift is Apples newest
language for building
native, high performance
applications for both iOS
and OS X. This session
will introduce you to
this exciting language.
Developers with a
background in either C#
or Python will see many
similarities. Almost
everyone will find Swift a
much more comfortable
and inviting language
when compared to
Objective C or C. Come
and learn why its now
fun to develop (natively)
for iOS and OS X.

"&GX9GGM | YYYFGXYGGMEQO | 7

DevWeek_2015_Brochure.indd 7

1/5/15 10:53 AM

67'5&#;/#4%*

&#;#)'0&#%106+07'&



$'$0
7251+,//

JH;7JOEKH9E:;7I
79H?C;I9;D;
Well never be able to
understand large-scale
systems from a single
snapshot of the code.
Instead, we need to
understand how the
code evolved and how
the people who work
on it are organised. We
also need strategies
that let us find design
issues and uncover
hidden dependencies
between both code and
people. Where do you
find such strategies if
not within the field of
criminal psychology?
In this session, Adam
will use this approach
to predict bugs, detect
architectural decay and
find the code that is most
expensive to maintain.



('
&2857(1$<

?DL;HI?EDE<
9EDJHEB'&'
The dependency
injection/inversion
of control design
pattern is an important
technique that helps
to write testable and
maintainable code. In
this session, Ed will
debunk the myth that
its hard to understand
or only for enterprise
development, and
demonstrate how to use
it in everyday code. This
demo-led session will
discuss the rationale
behind dependency
injection, demonstrate
injection with and
without a dependency
container, as well
as writing a simple
container from scratch.

8 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 8

72%,$6
.20,6&+.(

J>;FHEC?I;:B7D:
E<KN
In todays competitive
landscape, a stellar user
experience is a strong
product differentiator
and enabler of market
success. Companies
hope to get to this
Promised Land
where their own staff
deploys mature UX
design practices, their
customers are happy,
and their market share
and profits increase.
Noone ever said that
this journey to the
Promised Land is easy
a lot of companies
havetried and failed.
In this inspirational
overview session, Tobias
will showthe path and
identify what strategic
elements are critical
tosuccessful design.

3($5/
&+(1

?DJ;HD;JJ?D=
OEKHJ>?D=I
From big consumer
success stories, such as
the Nest Thermostat, to
more offbeat monitoring
systems, such as
Botanicalls (which lets
your plant call you when
it is thirsty), your things
are finding their own
voice through small but
powerful embedded
microcontrollers. The
market for the Internet
of Things (IoT) and
wearables is exploding.
But what are your
options for getting
started with making
physical things when
youre more used to
writing software? In this
session, Pearl will go
through a few hardware
options that are great
for hobbyists and rapid
prototyping.

6+$<
)5,('0$1

7D=KB7H@IJ>;ED;
<H7C;MEHAJEHKB;
J>;C7BB
In the last couple of years,
weve seen the rise of
client-side JavaScript
frameworks. From almost
nothing, now we have at
least a dozen to choose
from. One of the new kids
in the block, AngularJS,
comes straight from the
Google offices and tries
to stand out from the
crowd with a complete
set of tools and utilities.
Some are very powerful
and tryto ease your
way towards your SPA
web site. In this session,
Shay will go through
the different features
of AngularJS and see
what makes it one of the
most popular JavaScript
frameworks out there.

3+,/
/(**(77(5


F7JJ;HDI7D:
FH79J?9;I<EH
8K?B:?D=;DJ;HFH?I;
I97B;>JCB+7FFI
Developing large apps
is difficult. Ensuring
that code is consistent,
maintainable, testable
and has an architecture
that enables change is
essential. When it comes
to large server-focused
apps, solutions to some
of these challenges have
been tried and tested.
But, how do you achieve
this when building
HTML5 single-page
apps? In this session,
Phil will highlight
signs to watch out for
as your HTML5 SPA
grows, andpatterns
and practices to help
you avoid problems.
He will also explain
thearchitecture that
their HTML5 apps
havethat is core to
ensuring they scale.

5$/3+
'(:$5*1<

9E:?D=J>;
F7H7BB;B<KJKH;
?D9!!
Many-core processors
and computing
platforms will be
ubiquitous in the future:
from multi-core and
many-core CPUs to
integrated GPUs to
compute clusters in
the cloud, its the new
parallel universe for
software developers.
In this session, you
will learn the tools,
techniques and best
practices available to
C/C++ developers to
make sure your code is
ready today to run with
maximum performance
and reliability in this
new parallel universe.

25(1
58%,1

J;IJ7KJEC7J?ED
:ED;H?=>J0
J>;>EBO=H7?B
E<9EDJ?DKEKI
:;FBEOC;DJ
In this session, Oren
will explain all there is
to know about end-toend test automation.
Starting with the basics
and a comparison to
unit testing, he will then
drill down into what is
considered uncharted
territory for many
developers. You will
learn the best practises
and design patterns,
common pitfalls, and
most importantly
the full ecosystem
and how it connects to
your existing toolchain.
You will learn about
different approaches to
UI verifications, and see
real industry use cases
and bugs.

',12
(6326,72

7IF$D;J?:;DJ?JO"
9B7?CI7D:IE9?7B
7KJ>;DJ?97J?ED
ASP.NET Identity is the
new and comprehensive
membership system for
the whole ASP.NET
platform, including Web
API and SignalR. Similar
in many ways to the
popular simple
membership provider,
ASP.NET Identity goes
well beyond in a number
of aspects: replaceable
storage, flexible
representation of user
profiles, external logins,
claims-based
authentication and role
providers. Dino will use
the Identity API to set
upsocial authentication
via Facebook and
Twitter, and to collect
any user information
made available by
socialnetworks.

I9
M
Th
No
be
ou
up
M
co
sca
gen
mo
No
Mo
wh
to
see
of
(re
geo
No
da
ou
on
wi
sca
ac
lev

6$6+$
*2/'6+7(,1

M>7JID;M?D
L?IK7BIJK:?E(&'+5
In this session, Sasha
will lead us through
a sample of the new
Visual Studio 2015
features, ranging from
developer productivity
to low-level C++ code
optimisations. In a
series of quick-paced
demos, we will show
how Visual Studio 2015
makes key diagnostics
experiences easier,
improves IntelliTrace
analysis, helps
developercollaboration
and productivity, and
produces higher-quality
and faster code for
both managed (with
.NET Native) and C++
applications. If youre
using Visual Studio,
youcant afford to
missthis talk!

"&GX9GGM

1/5/15 10:53 AM

C;
F7
Th
wh
me
be
ap
as
in
me
In
wo
pu
pro
sp
exp
ne
to
op
ses
sev
int
can
me
loo
me
gu
tec

#.51'#%*&#;
&2))(( 5(*,675$7,21

| &2))((%5($. |

/81&+%5($. | &2))((%5($.

-2,186)25'5,1.6$77+((1'2)'$<1(7:25.,1*'5,1.66321625('%<
0,&+$(/
.(11('<

I97B?D=J>;DEIGB
M7OM?J>CED=E:8
The great promise of the
NoSQL databases has
been their ability to scale
out rather than scaling
up. In this session,
Michael will look at a
concrete example of
scaling one of the most
generally useful and
most widely deployed
NoSQL database:
MongoDB. Hell explore
why you might need
to scale out, and youll
see the full spectrum
of choices for scaling
(replication, sharding,
geo-replication, etc).
NoSQL document
databases typically
outperform RDBMSes
on single servers but
with the ability to
scale out they can truly
achieve an entirely new
level of performance.

0,.(
:22'

C;II7=?D=
F7JJ;HDI
There are many reasons
why asynchronous
messaging should
be introduced in
applications, as well
as many approaches
in incorporating
messaging subsystems.
In some cases, intensive
workloads need to be
pushed to back-end
processing, or perhaps
specialised (and often
expensive) resources
need to be utilised
to perform certain
operations. In this
session, Mike will cover
several scenarios where
introducing messaging
can help, discuss a few
messaging patterns, and
look at abstracting your
messaging subsystem to
guard against evolving
technology and designs.

,4%$/
.+$1

B;7HD>EMJE
I97B;$D;J7FFI?D
C?9HEIE<J7PKH;
M?J>:?IJH?8KJ;:
979>?D=
Discover the scalability
bottlenecks for your
.NET applications in
Microsoft Azure, and
how you can improve
their scalability with
distributed caching.
This session provides
a quick overview of
scalability bottlenecks,
and answers some key
questions: What is
distributed caching and
why is it the answer in
Microsoft Azure? Where
in your application can
you use distributed
caching? What are some
important features in
a distributed cache?
Youll also see hands-on
examples of using a
distributed cache.

*$5<
6+257

:7J7I9?;D9;<EH
<KD7D:FHE<?J
Make no mistake: data
science can be hard,
but it can also be fun. In
this session, Gary will
introduce you to classic
and Bayesian statistics
and machine learning, all
through the medium of
predicting horse-racing
results. Hell explore a
number of techniques
for making such
predictions and finish by
combining them into a
powerful mixed model
prediction engine thats
sure to pick the next big
winner. This session
wont only improve your
knowledge, itll improve
your bank balance too!
(Note: Session may not
improve bank balance.)

1($/
)25'

8K?B:?D=
C?9HEI;HL?9;
7H9>?J;9JKH;I
Inspired by the success
of companies such as
Amazon and Netflix,
many organisations
are moving rapidly
towards microservice
architectures. This
style of architecture is
important because its
the first architecture
to fully embrace the
Continuous Delivery
and DevOps revolutions.
In this session, Neal will
cover the motivations for
building a microservice
architecture, some
considerations you must
make before starting
(such as transactions
versus eventual
consistency), how
to determine service
partition boundaries,
and ten tips for success.

$867,1
%,1*+$0

>?=>#GK7B?JO
:;9?I?EDC7A?D=
M?J>EF;D:;I?=D
FHEFEI7BI
Making complex
decisions in software
design involves
balancing many factors,
and maintaining
that balance can be
challenging. By opening
up the decision process
for evolution, we can
harness the insight
of fellow developers,
communicate plans and
designs more effectively,
and produce a useful
record of the work we do.
In this session, Austin
will look at a specific
technique: Open Design
Proposals. Hell examine
implementations of this
approach, see why its
effective, and show how
development teams can
use it to manage their
own decision making.

1812
*2',1+2

;L;DJ>K8I"7PKH;
IJH;7C?D=7D:
7PKH;?II
Recently, there has
been a lot of talk
around IoT, M2M,
big data and similar
topics. Its important
to understand how we
can take advantage of
these concepts, and how
they can help us achieve
our goals. Fortunately,
Microsoft has some
solutions for us. These
are Azure Event Hubs,
Azure Stream Analytics
and Azure Intelligence
Systems Service (ISS).
In this session, Nuno
will explore these three
topics, demonstrate
their interconnectivity,
and show how they
provide the perfect
answer for our nextgeneration solutions
andinteractions.

-$0(6
0217(0$*12

?8;79EDI7D:
9EDJ;NJK7B
BE97J?ED
7M7H;D;II?D?EI
7D:7D:HE?:7FFI
iBeacons are taking the
world by storm from
retail stores to major
sporting events, youll
soon be finding iBeacons
just about everywhere.
This gives you the ability
to enable any number of
device proximity-based
scenarios that were
never before possible.
Inthis session, James
will explain what an
iBeacon is, how they
work, how you would
want to use them, and
how to get started
making apps in both
Android and iOS. All
demonstrations will be
coded in C#, but will be
applicable to any iOS or
Android developer in
any language.

-8/(6
0$<

6$+,/
0$/,.

?<9EDI?:;H;:
>7HC<KB0>EMJE
;H7:?97J;/+E<
7BBOEKH8K=I?D
ED;I?CFB;IJ;F
In 1968, CACM
published a letter from
Edgar Dijkstra, called
Go To statement
considered harmful.
In it, he explained why
most bugs in programs
were caused by Gotos,
and appealed for
Goto to be expunged
from programming
languages. But Goto has
a twin brother, which is
responsible for nearly
every bug that appears
in our programs today.
That twin is If. In this
session, Jules revisits
Dijkstras original
explanation to show
why If and Goto have
thesame pathology, and
how you can avoid it.

.(9/,1
+(11(<

JEF'&
@7L7I9H?FJJ?FI
JavaScript, the lingua
franca of the web, is
incredibly freeform and
therefore hard to get
right. Weve all hacked
JavaScript, but what
do you need to know
when you are doing big
and complex JavaScript
projects? This isnt your
Dads browser, yknow!
So you have written
JavaScript, but want to
go beyond the basics? In
this session, Sahil will
show you the JavaScript
concepts that every
modern JavaScript
developer needs to know.

6$+,/
0$/,.

<F*EEF<JM
Although not yet fully
mainstream, functional
programming has
finally reached a critical
mass of awareness
among developers.
Theproblem, however,
is that many developers
are up against an even
greater critical mass
of existing code. Much
of this code purports
to be object oriented,
but beyond theuse
of the class keyword,
falls somewhatshort
of putting the
OO to gooduse.
Manytechniques
identified as functional
have broader
applicability. In this
session, Kevlin will
explore how some
FP habits can be
incorporated into
OOPpractice.

JEF'&@7L7I9H?FJ
:;8K==?D=JH?9AI
Weve all been writing
lots of JavaScript code
lately. But JavaScript
is incredibly free
form, and that sharp
double-edged sword can
also make debugging
JavaScript errors a lot
more difficult. The new
operating system is the
browser, and complex
JavaScript pages cannot
ignore performance,
or their unpredictable
behaviour under
different bandwidths.
In this session, Sahil
will showyou some
really useful debugging
techniques and
demonstrate how to
use each browser for
itsbest capabilities.
Ninety minutes spent
here willsave you hours
inyour day job.

"&GX9GGM | YYYFGXYGGMEQO | 9

DevWeek_2015_Brochure.indd 9

1/5/15 10:53 AM

9'&0'5&#;/#4%*

&C[QH&GX9GGMEQPVKPWGUYKVJCRCEMGFRTQITCOOGQHYQTMUJQRUCPFDTGCMQWV
UGUUKQPUCPFFQPVHQTIGVVJCVKH[QWECPVOCMGKVVQCP[QHVJGVCNMU[QWTGKPVGTGUVGFKP
CUCTGIKUVGTGFFGNGICVG[QWNNDGCDNGVQECVEJWRNCVGTQPNKPGCVVJG&GX9GGMYGDUKVG
&#;#)'0&#



$'$0
7251+,//

9E:;J>7J<?JI
OEKH8H7?D
People think, remember
and reason in a very
different way from that in
which code is presented.
So how should code
look to make it both
easier to understand and
maintain? To see what
really works, we need to
look across languages
and paradigms. In this
session, Adam will
start with common
problematic constructs
such as null references,
surprising cornercases and repetitive
code, and discuss the
cognitive costs and
consequences of each.
Hell then apply ideas
from object-orientation,
functional programming
and lesser-known array
languages to explore
better approaches.



3$9(/
6.5,%7629

9ECFKJ?D=B?A;
J>;8H7?D07D
?DJHE:K9JEHO
=K?:;JE7?
Every now and again,
every professional
developer faces a
program that he or she
has trouble writing. Try
to imagine an algorithm
that has to differentiate
a dog from a cat. They
come in different shapes
and sizes, and there is
no single feature that
could discriminate
between the two. Any
attempt to code that
algorithm manually
using deep-nested
if/else branches is
doomed. Human beings,
on the other hand, have
no trouble with this task.
In this session, Pavel will
introduce the basics of
an artificial intelligencebased approach to
solving these problems.

10 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 10

72%,$6
.20,6&+.(

L?IK7B:;I?=D<EH
DED#:;I?=D;HI0
?JIDEJ@KIJ
78EKJ9EBEKH
In an ideal world,
front-end developers
dont need to worry
about visual design
because they get
specs and assets from
professionally trained
designers. The reality
is that developers often
need to make their own
decisions about how to
make the UI attractive.
In this session, Tobias
will provide a solid
base knowledge about
what constitutes
attractiveness and what
design principles can
be applied to boost the
visual appeal of UIs. By
the time you leave, youll
be ready to step up to the
mark when designers
are nowhere to be found!

-2(
1$72/,

J>?DA<?HIJ0M>O
=H;7JKNIJ7HJI
8;JM;;DOEKH;7HI
7D:DEJEDJ>;
I9H;;D
When developers are
tasked with improving
UX, their focus tends
to be on the screen:
elements, interactions,
workflow, often
accompanied by the
worrying cry, Im
not a UI designer!
Fortunately, Joe has
good news: you can
still design great user
experiences without a
shred of visual design
talent. In this session,
Joe will show you
how changing the
way you think about
features, functions and
implementations can
make a massive, positive
change in the experience
people have with your
UIand your product.

3+,/
/(**(77(5

M>OOEKI>EKB:
8;KI?D=M;8
9ECFED;DJI
7D:>EM
Web Components are
touted as the future of
web development. In this
session, Phil will explain
what Web Components
are, the state of native
support in web browsers,
what your options are for
building componentised
web apps right now
using AngularJS, Ember,
Knockout or React, and
why Web Components
probably are the future of
web development. Hell
also cover the benefits
of a component-based
architecture and how
it helps when building
JavaScript apps, as well
as how components
can communicate
in a loosely coupled
way,andwhy.

$//(1
+2/8%

ADE9AEKJ07D
?DJHE:K9J?ED
The Knockout
framework is
a standalone
implementation of the
MVVM (Model-ViewViewModel) pattern,
which is one of the
best user-interface
architectures for
web applications. It
provides an alternative
to AngularJS more
limited in scope but
smaller and, in some
contexts, faster. In this
session, Allen will look at
Knockouts architecture
and how to leverage
that architecture to
build highly interactive
web-application user
interfaces. The session
will include several code
(JavaScript) examples.

('
&28571(<

7D?DJHE:K9J?ED
JEJOF;I9H?FJ
JavaScript is the
scripting glue that
holds the web together
largely because of
its flexibility. This
flexibility also means
that it can be difficult to
manage, especially in
large-scale applications.
In this session, using
real-world examples,
Ed will explore some
of the problems with
client-side JavaScript
development as a
motivating example and
introduce TypeScript as
a way of solving some of
these issues. Hell also
explore how existing
JavaScript codebases
can be targeted by
TypeScript and how it
can be integrated into
build systems.

'525
+(/3(5

87JJB;E<J>;
$D;JCE9A?D=
<H7C;MEHAI
Writing unit tests is
hard, isnt it? You need an
entire set of tools just to
start. One of the crucial
decisions when building
this set is picking up a
mocking framework.
But beware the
mocking framework
you choose has the
ability to make or break
you! In this session,
Dror at one time a
mocking framework
developer will cover
the capabilities and
functionality of the
leading frameworks,
showing the good and
the bad of the different
options (both free
and commercial), and
making them battle to
the death!

6$6+$
*2/'6+7(,1

9EEBB?8H7H?;I
<EHCE:;HD9!!
The C++ standard
library dates back to the
1990s, but that doesnt
mean there arent new
and exciting frameworks
to use in your C++
application. Inthis
session, Sasha will look
at some brand-new
and some existing C++
libraries that can speed
up C++ cross-platform
development rapidly.
Some of the libraries
we might cover include:
Casablanca (C++ REST
SDK), Cinder (creative
coding), Boost (generalpurpose), Google Test
(unit testing), SOCI
(modern database
access) and many others.
Theres something
foreveryone!

J>
8;
BE
Yo
Lo
Se
fle
da
sh
yo
the
co
Bo
ca
pe
tha
CP
In
wi
un
Bo
ho
SQ
ou
he
ab
pe
the

6+$,
5(=1,.

8K?B:FHE:K9J?ED#
H;7:O@7L7I9H?FJ
7FFIM?J>=HKDJ
In this session, Shai will
deliver an overview of
the steps required in
order to build JavaScript
apps and get them
ready for deployment.
Hell cover build
theory,asking Why
build in JS? Hell also
talk about the build steps
and then jump to Grunt,
explaining what it is and
providing a live demo.
Finally, Shai will cover
the scaffolding tool,
Yeoman. This session
is intended to be both
funny and informative,
so get ready to have a
good time while picking
up some essential tips to
make your day job that
much easier.

"&GX9GGM

1/5/15 10:53 AM

7F
M?
7D
No
in t
tod
is N
use
app
ses
dig
dat
the
the
the
(ke
doc
for
exp
app
No
dat
han
the
and
doc
Mo

#.51'#%*&#;
&2))(( 5(*,675$7,21

| &2))((%5($. |

/81&+%5($. | &2))((%5($.

s.

./$86
$6&+(1%5(11(5

J>;:7D=;HEKI
8;7KJOE<8EEAC7HA
BEEAKFI
You know Bookmark
Lookups in SQL
Server? You like their
flexibility to retrieve
data? If you do, then you
should be warned that
you are dealing with
the most dangerous
concept in SQL Server!
Bookmark Lookups
can lead to massive
performance losses
that will devastate your
CPU and I/O resources!
In this session, Klaus
will provide a basic
understanding of
Bookmark Lookups and
how they are used by
SQL Server. After laying
out the foundations,
helltalk in more detail
about the various
performance problems
they can introduce.

0,&+$(/
.(11('<

7FFB?;:DEIGB
M?J>CED=E:8
7D:FOJ>ED
NoSQL is a hot topic
in the tech industry
today. But what exactly
is NoSQL and should I
use it to build my next
application? In this
session, Michael will
dig into why NoSQL
databases are sweeping
the industry and discuss
the trade-offs between
the various types
(key-value stores vs
document databases,
for example). He will
explore the most broadly
applicable variant of
NoSQL, document
databases, through
hands-on demos with
the most popular
and successful of the
document databases,
MongoDB.

'(-$1
6$5.$

:7J7C?D?D=
7B=EH?J>CIM?J>
IGBI;HL;H7D:H
F7HJ'
Data mining is gaining
popularity as the most
advanced data analysis
technique. With modern
data mining engines,
products and packages,
such as SQL Server
Analysis Services
(SSAS) and R, data
mining has become a
black box. Its possible to
use data mining without
knowing how it works,
but this can lead to many
problems, such as using
the wrong algorithm for
a task, misinterpretation
of the results and more.
In this session (and Part
2, at 11.30), Dejan will
explain how the most
popular data mining
algorithms work and
when to use each one.

'(-$1
6$5.$

:7J7C?D?D=
7B=EH?J>CIM?J>
IGBI;HL;H7D:H
F7HJ(
Data mining is gaining
popularity as the most
advanced data analysis
technique. With modern
data mining engines,
products and packages,
such as SQL Server
Analysis Services
(SSAS) and R, data
mining has become a
black box. It is possible to
use data mining without
knowing how it works,
but this can lead to many
problems, such as using
the wrong algorithm for
a task, misinterpretation
of the results and more.
In this session (following
on from Part 1, at 09:30),
Dejan will explain how
the most popular data
mining algorithms
workand when to
useeach one.

1($/
)25'

9EDJ?DKEKI
:;B?L;HO<EH
79>?J;9JI
Yesterdays best practice
is tomorrows antipattern. Architecture
doesnt exist in a
vacuum: a painful
lesson developers who
built logically sound
but operationally
cumbersome
architectures
learned. Continuous
Delivery is a process
for automating the
production-readiness
ofyour application every
time a change occurs
to code, infrastructure
or configuration. In
this session, Neal will
take a deep dive into
the intersection of the
architect role and the
engineering practices in
Continuous Delivery.

',12
(6326,72

:::
C?I9ED9;FJ?EDI
For too long, domaindriven design (DDD)
has been sold as the
ideal solution for very
complex problems
that only a few teams
are actually facing.
While technically
correct, this statement
sparked a number of
misconceptions. In
fact, DDD is only an
approach to the design
of software systems and
is driven by the domain
of the problem. In this
session, Dino will clear
the ground around
DDD, emphasising
the theoretical pillars
of theapproach:
ubiquitouslanguage
andbounded context.

-$0(6
0217(0$*12

?EI7D:7D:HE?:
:;L;BEFC;DJ<EH
9:;L;BEF;HM?J>
N7C7H?D
As the mobile landscape
continues to expand
and evolve, managing
multiple codebases in
different programming
languages and
development tools
can quickly become a
nightmare. Wouldnt
you love to build native
UIs for iOS, Android and
Windows Phone from
a single codebase? In
this session, James will
show how to leverage the
awesome features of C#
and combine them with
Xamarin technology to
create beautiful, native,
cross-platform, mobile
apps from a shared C#
codebase, with the tools
that you love.

0,.(
:22'

+B?=>JM;?=>J
C?9HEIE<J7PKH;
<;7JKH;I<EH<7IJ#
CEL?D=CE8?B;:;LI
Mobile development
has exploded, and
everyone has an idea
they want to try out.
But bootstrapping a
mobile app doesnt
always seem that easy.
Consumers demand
slick user experiences
and the ability to share
data across a plethora of
devices and platforms,
while were trying to
get a minimal viable
product out the door
totest our ideas as fast
as possible. Thankfully,
Azure has powerful
features available to
help. In thissession,
Mike willtake a practical
lookat five features of
Azure that are useful
formobile developers
ofany platform.

6$+,/
0$/,.

$1'<&/<0(5
5,&+$5'%/(:(77

B;7HD7D=KB7H@I0
J>;HE7:JE
FEM;H<KB"
C7?DJ7?D78B;
7FFB?97J?EDI
JavaScript, by its nature,
makes it difficult to
write maintainable code.
HTML, by its nature,
is loosely structured.
AngularJS fixes both of
those. Itsastructural
framework for dynamic
web apps, allowing you to
extend HTMLs syntax,
enabling you to write
powerful, maintainable
applications succinctly.
In this workshop,
Sahil will build on your
existing knowledge of
JavaScript and teach
you the ins and outs of
AngularJS. There are
plenty of examples,
which will walk
you through a basic
introduction, models,
controllers and views in
Angular, templates and
databinding, services
and dependency
injection, directives,
routing and single-page
applications.

IEB?:7IOD9
FHE=H7CC?D=
?D$D;J
In this special two-day
workshop, Andrew and
Richard will take you
through the core skills
required to successfully
develop async and
multithreaded code,
both in the .NET and
web worlds. Not only do
we cover the core APIs,
but also how they are
used effectively, tested
and debugged.
For a full description
ofthe workshop, please
seePage 20

For a full description

ofthe workshop, please
seePage 21

10'&#;914-5*12


691&#;914-5*12

914-5*12
4'(/%

914-5*12
4'(/%

"&GX9GGM | YYYFGXYGGMEQO | 11

DevWeek_2015_Brochure.indd 11

1/5/15 10:53 AM

9'&0'5&#;/#4%*

&#;#)'0&#%106+07'&



('
&2857(1$<

8;>7L?D=B?A;7
=?J"7D:=;JJ?D=
7M7OM?J>?J
Although Git has rapidly
become almost a de facto
standard in recent years,
it can be intimidating
or confusing for those
transitioning from
other systems or those
new to using source
control. In this session,
Ed will explain how to
use Git effectively, how
to navigate your way
around a repository, and
how to work as part of a
team. Hell attempt to cut
through the mystique
and demonstrate how
easy it can actually be
to use. Then hell go on
to show some of the
more advanced ways
ofworking with Git.



.(9/,1
+(11(<

FHE=H7CC?D=
M?J>=KJI
These days, testing
is considered a sexy
topic for programmers.
Whod have thought
it? But what makes
for good unit tests
(GUTs)? Theres more
to effective unit testing
than just knowing the
assertion syntax of a
testing framework.
Testing represents a
form of communication
and, as such, it offers
multiple levels and
forms of feedback,
not just basic defect
detection. Effective
unit testing requires
an understanding of
what forms of feedback
and communication
are offered by tests. In
this session, Kevlin will
explore exactly what
makes a good unit test.

12 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 12

+2:$5'
'(,1(5

9EDJ?DKEKI
:;B?L;HO0J>;
M>OI"M>7JI"
7D:>EMI
DevOps is commonly
believed to be
accomplished by having
the development staff
collaborate more closely
with the operations
staff. Thats definitely
necessary, but woefully
inadequate to achieve
the goal of faster and
better delivery in the
last mile of an Agile
shop. In this session,
Howard will discuss
the rationale behind
Continuous Delivery,
along with specific
practices to get you
started on making your
sprints toward customer
satisfaction less tiring
and more enjoyable for
everyone involved.

-8/(6
0$<

J;7C8K?B:?D=
No programmer is
an island. Modern
programs are created
by teams of developers.
And everybody knows
you need great teams to
build great products so
you need to build your
teams carefully. But
what, exactly, makes
a great programming
team? Great
programming skills?
Great interpersonal
skills? Working-allnight-because-theboss-has-thrown-a-fit
skills? Turns out, its
none of these. In this
session, Jules will reveal
that what makes a
programming team great
is exactly the same stuff
that makes any other
team great and most
programming teams
dont have it.

6$6+$
*2/'6+7(,1

C7A?D=$D;J
7FFB?97J?EDI
<7IJ;H
Speed is king on mobile
devices, embedded
systems, and even
run-of-the-mill
desktop applications
that need to start up
quickly and deliver
good performance on
low-power machines.
In this session, Sasha
will review a collection
of practical tips you
can use today to make
your .NET applications
faster. Hell talk about
choosing the right
collection, improving
start-up times, reducing
memory pressure, and
many other techniques
for quickly improving
your apps performance.

,'2
)/$72:

7IF$D;JLD;NJ0
H;?C7=?D?D=
M;87FFB?97J?ED
:;L;BEFC;DJ
?D$D;J
ASP.NET vNext is being
designed from the
bottom up to be a lean
and composable .NET
stack for building web
and cloud-based
applications. Envision
an ASP.NET stack
where MVC, Web API,
and web pages are all
merged into the same
framework, where you
have a server-optimised
version of ASP.NET
with a smaller memory
footprint. This is the new
ASP.NET vNext. In this
session, Ido will explore
the ecosystem of ASP.
NET vNext, its new
project system and
configuration system,
and how to use it to build
exciting web applications.

0,&+$(/
+$%(50$1

KD?JJ;IJ?D=7D:
;(;J;IJ?D=KI?D=@I#
87I;:<H7C;MEHAI
Unit testing and end-toend (e2e) testing are the
tools to enforce stability
on applications. They
create an environment
that ensures our code
does what it was
designed to do. Recently,
web application
developers are looking
to identify the best
testing option, as their
applications are getting
increasingly large and
more complex. In this
session, Michael will
review two methods for
testing web applications
in different JS-based
frameworks: the
unit-testing approach
and end-to-end
testing. He will also
review thebenefit of
combiningthe two.

6+$<
)5,('0$1

9>HEC;:;L;BEF;H
JEEBI7:;;F:?L;
Every developer needs
a set of tools, especially
web developers that
bend under the pressure
of multiple languages,
environments, IDEs
and what not. One of the
most comprehensive
toolsets out there today
is Chrome Developer
Tools. It contains so
many amazing features
beyond the common
ones, and its just a
shame most developers
dont know about
them! In this session,
Shay will tell you all
about the known and
less-known features
of Chrome Developer
Tools, and youll see
how your everyday
web development can
become easier with just
a few simple steps.

$17+21<
61(('

I;9KH?D=M;87F?i
J>;D;MM7OM?J>
EM?D7D:A7J7D7
Sometimes the
technology landscape
is changing so fast, it
feels like youre standing
on quicksand. That is
certainly the case with
ASP.NET Web API,
the new OWIN hosting
model and Microsofts
Katana implementation.
In this session, Anthony
will show how to
correctly apply security
at the transport level to
ensure confidentiality,
integrity and server
authentication, as well
as the nuts and bolts of
configuring SSL for both
web and self-hosted
web APIs using the new
OWIN hosting model.

P;
H7
C;
7=
I9
M
ess
hig
dy
ser
the
pa
inf
ser
wo
far
lev
for
da
log
far
int
tha
Al
fro
pe
pr
Ra

6$6+$
*2/'6+7(,1

FH79J?97B9,
7D:8;OED:
Visual Studio 2015, .NET
2015 and C# 6 are just
around the corner. The
new language features
have been out ofthe
bag for a while now,
but how do you apply
them effectively? How
do you refactor existing
code to be shorter and
sweeter? In this fastpaced session, Sasha
will lead us through
experiments with the
new language features,
including expressionbodied members,
enhancements to
automatic properties,
null propagation,
stringinterpolation
andmany others.

"&GX9GGM

1/5/15 10:53 AM

KD
7I
?D
Is i
use
as
Se
lot
DB
the
by
clu
In
wi
ofu
wh
bad
goo
dis
the
you
cri
If t
ne
wi
pra
res
wi
un

#.51'#%*&#;
&2))(( 5(*,675$7,21

$//(1
+2/8%

P;HECG7D:
H788?JCG0
C;II7=?D=<EH
7=?B?JO7D:
I97B78?B?JO
Messaging is an
essential technology in
high-volume,
dynamically scalable
server applications. Its
the most effective way to
pass non-time-critical
information between
servers, and to distribute
work within a server
farm. At the inter-server
level, messaging is ideal
for use with remote
databases, monitoring,
logging and so on, and a
far better solution to
intra-server data sharing
than a shared database.
Allen looks at messaging
from an architectural
perspective, with
practical examples using
RabbitMQ and ZeroMQ.

./$86
$6&+(1%5(11(5

KD?GK;?:;DJ?<?;HI
7IFH?C7HOA;OI
?DIGBI;HL;H
Is it good practice to
use uniqueidentifiers
as primary keys in SQL
Server? They have a
lot of pros for devs, but
DBAs just cry when
they see them enforced
by default as unique
clustered indexes.
Inthis session, Klaus
will cover the basics
ofuniqueidentifiers:
why they are sometimes
bad and sometimes
good; and how to
discover if they affect
the performance of
your performancecritical database.
If they are having a
negative impact, you
will also learn some best
practices you can use to
resolve those limitations
without changing your
underlying application.

*$5<
6+257

>7:EEF
A?9AIJ7HJ;H<EH
C?9HEIE<J:;LI
Big data is the new
shiny thing right now,
and if you read the
blogosphere youd be
forgiven for thinking
it was a tool just for
Linux devs or worse,
only for those annoying
hipsters with their shiny
Macs. Nothing could
be further from the
truth. Windows makes
an excellent platform
for Hadoop and, in this
session, Gary will show
you everything you need
to know to get started.
From downloading and
installing, to writing your
first map-reduce job,
using both the streaming
API and the SDK. This
session will cover it all,
so come along and join
the big data wave!

'$1
&/$5.

7KJEC7J?D=II?I
F79A7=;9H;7J?ED
M?J>8?CB
(BIML) is a powerful
XML-based markup
language that allows
you to generate
SSIS packages
programmatically. Using
BIML along with C#,
you can create metadatadriven packages, greatly
reducing development
time and increasing
consistency across the
team. In this session,
Dan will show you how
to automate your SSIS
package creation using
the power of BIML and
C#. Youll see how to
create a template for
loading dimension tables
that will greatly increase
your productivity.

| &2))((%5($. |

6$1'(5
+22*(1'2251

?D:?L?:K7BI7D:
?DJ;H79J?EDI
EL;HFHE9;;
7D:<EEBI
The first statement in
the Agile Manifesto
favours individuals,
teams, interaction
and collaboration over
processes and tools.
But there are two sides
to every story. When
it comes to tools, the
Agile Manifesto is often
misinterpreted, in the
sense that its wrong
to use tooling in Agile
projects. Despite this,
more and more vendors
are trying to jump on the
Agile bandwagon and
sell their tools as being
the most Agile toolset
available. In this session,
Sander shines a critical
light on the sense and
nonsense of tools in
theAgile field.

6(%
526(

9K9KC8;H7D:
IF;9<BEM7I
F7HJE<OEKH
:;L;BEFC;DJ
FHE9;II
Behaviour-driven
development (BDD) and
specification by example
(SBE) are quite recent
additions to the software
development toolbox.
Sometimes it feels like
were using a hammer
to drive in a screw.
So,inthis session, Seb
will explore what theyre
good for and when to
use them. Hell also look
at what problems they
dont help with and when
not to use them. Bythe
end of this session, youll
know enough to decide
whether your problems
are more like a screw
or a nail and whether
Cucumber/SpecFlow
isthe right hammer.

/81&+%5($. | &2))((%5($.

0$5.
60,7+

=;JJ?D=OEKH
CE8?B;7FFIH;7:O
<EHJ>;MEHB:
Localising your
applications can open up
a whole new audience of
users for your software.
In this session, Mark will
take a look at how to get
your application ready
for localisation and
how to then utilise the
built-in services of iOS,
Android and Windows
Phone to display proper
information for different
cultures and regions.

6$+,/
0$/,.

$1'<&/<0(5
5,&+$5'%/(:(77

B;7HD7D=KB7H@I0
J>;HE7:JE
FEM;H<KB"
C7?DJ7?D78B;
7FFB?97J?EDI
One-day workshop
continues from the
morning session.
For a full description
ofthe workshop, please
seePage 21

IEB?:7IOD9
FHE=H7CC?D=
?D$D;J
Two-day workshop
continues from the
morning session.
For a full description
ofthe workshop, please
seePage 20

1812
*2',1+2

?EJC(C0>EM7H;
J>;O9>7D=?D=J>;
MEHB:M;B?L;?D5
Internet of Things (IoT)
is here, and every day
a new sensor or device
starts to generate more
data. With that, more
and more machineto-machine (M2M)
communications start
to happen, which make
our solutions behave
differently and face new
issues. In this session,
Nuno will look at how
both of these new buzz
words are changing the
world we live in, from
fitbit to Google Glass
and smart watches. How
can we prepare for this?
How can we anticipate
and get some business
opportunities from it?
Join us and find out.

10'&#;914-5*12

691&#;914-5*12

914-5*12
4'(/%

914-5*12
4'(/%

"&GX9GGM | YYYFGXYGGMEQO | 13

DevWeek_2015_Brochure.indd 13

1/5/15 10:53 AM

6*745&#;/#4%*

&GX9GGM&C[UGGUVJGEQPENWFKPIRCTVQHQWTVYQFC[YQTMUJQRQPCU[PEJTQPQWU
RTQITCOOKPIKP0'6RNWUCPGYYQTMUJQRQP'PVKV[HTCOGYQTMRNWUOQTG
DTGCMQWVUGUUKQPUEQXGTKPICYKFGTCPIGQHVQRKEU
&#;#)'0&#



0,&+$(/
.(11('<

FOJ>ED0
7D7C7P?D=I;9ED:
B7D=K7=;<EH
$D;J:;L;BEF;HI
The modern software
development landscape
isa terrain of many
platforms and
technologies. Gone are
the days where knowing
one technology really
well was enough to stay
on the cutting edge. Even
as we know we should
learn more and branch
out, that choice is
increasingly difficult as
the technology options
explode. In this session,
Michael offers one very
solid choice: Python. It
may seem like a very
different language and
ecosystem from .NET
but beneath the surface,
there are many more
similarities than
differences.



3$9(/
6.5,%7629

:;;FB;7HD?D=0
J>;>7D:9H7<J;:
9E:;A?BB;H
In this session, Pavel
will reveal how a new
direction in artificial
intelligence, called deep
learning, is gradually
reducing demand for
hand-crafted code for
intellectual data analysis,
primarily in the area
of feature extraction.
He will explain why
the internet giants
(Google, Microsoft
etc) are interested in
deep learning, and the
connection with big data
projects. He will also
cover practical examples
of applying existing
deep-learning software
frameworks to an imagerecognition problem.

14 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 14

',12
(6326,72

J>;:H7C7J?95
?CF79JE<KN
EDIE<JM7H;
7H9>?J;9JKH;
7D::;I?=D
Always neglected in
favour of domain
analysis and modelling,
the presentation layer of
applications receives
little attention. But
whether your application
is web, mobile or desktop,
the presentation layer is
the face it shows to users.
Dino will discuss a
design approach that
starts from requirements
and builds the system
from top to bottom,
focusing on use-cases,
screens and overall user
experience measured by
a new professional figure
the UX architect and
backed by new, but
partially green, tools such
as UXPin and Balsamiq.

$0<
&+(1*

:H7M?D=M?J>
@7L7I9H?FJ0
7D?C7J?EDI7D:
?D<E=H7F>?9I
A number of libraries
and frameworks allow
developers to use
JavaScript to create
engaging visuals without
switching programming
languages. So lets
explore the visual (and
fun) side of JavaScript!
In this session, Amy will
provide a whirlwind
tour of a few libraries
and frameworks that let
you create animations,
simple drawings and
infographics with
JavaScript. First, shell
examine the advantages
(and disadvantages)
of using JavaScript for
graphics and animations.
Then shell go through
Hello World examples
of a few JavaScriptbased libraries.

&+5,6726
0$76.$6

C;;JJ>;D;M
A?:EDJ>;8BE9A0
C?9HEIE<J7IF$D;J+
Imagine if you could
write an ASP.NET
application using your
favourite text editor,
compile it and run it on
Mac OS X. Imagine if
you could mix and match
Web Forms, MVC, Web
API and SignalR within
a single project. How
would it feel to create a
faster, leaner and more
memory-efficient ASP.
NET application that
has been freed from the
shackles of Windows,
and all you need are your
coding skills, a couple of
NuGet packages and
your imagination? In
this session, Christos
provides an intro to
Microsofts ASP.NET 5
the new kid on
theblock.

6$1'(5
+22*(1'2251

?DJHE:K9?D=
7D:;NJ;D:?D=
8EEJIJH7F
Bootstrap is by far
the most popular web
framework of all, with
many ready-to-use styles
and components in CSS
and JavaScript. In this
session, Sander will
show you how to build a
basic web site, leveraging
the many components
of the Bootstrap
framework. He will then
go on to show the use of
additional frameworks
and libraries to add
drop-down support,
icons and date pickers
to your web pages, and
how to build additional
reusable components
using Razor syntax,
in JSF, and applying
Angular directives. Of
course, Sanders talk will
be illustrated with many
coding demos.

'525
+(/3(5

KD?JJ;IJ?D=
F7JJ;HDI<EH
9ED9KHH;DJ9E:;
Getting started with
unit testing is not hard,
the only problem is that
most programs are more
than a simple calculator
with two parameters
and a return value thats
easy to verify. Writing
unit tests for multithreaded code is harder
still. In this session,
Dror will demonstrate
useful patterns that he
has discovered over
the years, and that
have helped him to test
multi-threaded and
asynchronous code and
enabled the creation of
deterministic, simple
and robust unit tests.
Hell also point out the
pitfalls to avoid.

$//(1
+2/8%

:89
:;I?=D8O9E:?D=
Design by Coding
(DbC) is a way to
develop an architecture
incrementally as you
code. It builds on testand behaviour-drivendevelopment techniques,
but adds a focus on the
story thats central to
all Agile processes. The
process answers the
question of how you can
build a coherent Agile
system incrementally,
without a formal upfront design process.
In this session, Allen
will explain how DbC
eliminates the need for
a separate design phase
in the development
process, since your code
is effectively your design
artefact.

(2,1
:22'6

IOIJ;CI;9KH?JO
8;OED:J>;
B?8H7H?;I
Security is now
important to all of us, not
just people who work
at Facebook. But its a
complicated domain,
with a lot of concepts
to understand. In any
technical ecosystem,
there is a blizzard of
security technology,
as well as generic
concepts such as keys,
roles, certificates, trust,
signing and so on. Yet
none of this is useful
unless we know what
problem were really
trying to solve. In this
session, Eoin will dive
into the fundamentals
of system security to
introduce the topics we
need to understand in
order to decide how to
secure our systems.

>;
=K
:;
IG
SQ
loc
pro
asp
As
wo
dea
a
for
ses
loo
loc
in S
on
you
the
dea
in S
tro
and
the
que
stra
dat

721,
3(75,1$

7M;IEC;J>?D=I
OEK97D:E
M?J>HEIBOD
Roslyn, the revamped
compiler for C# and
Visual Basic.NET,
goes beyond a mere
black-box compiler
and gives us limitless
possibilities. Besides
enabling a new era for
C# as a language, it gives
everyone a chance to
utilise compiler powers
for building custom
tools. It acts as a CaaS,
or Compiler as a Service,
which allows you to plug
in at any point in the
compilation process.
But what can you do
with it? In this session,
Toni will show you,
demonstrating how you
can build Visual Studio
extensions, create your
own editors or host
C# compiler to form a
scripting environment.

"&GX9GGM

1/5/15 10:53 AM

H7
DE
Th
is a
dat
sup
of d
sca
que
bes
for
bec
cla
aL
ses
sho
dat
sto
how
effi
con
wil
you
mil
rev
and
sup
ver
cap

#.51'#%*&#;
&2))(( 5(*,675$7,21

./$86
$6&+(1%5(11(5

>;7:79>;
=K7H7DJ;;:0
:;7:BE9A?D=?D
IGBI;HL;H
SQL Server needs its
locking mechanism to
provide the isolation
aspect of transactions.
As a side-effect, your
workload can run into
deadlock situations
aguaranteed headache
for any DBA! In this
session, Klaus will
look into the basics of
locking and blocking
in SQL Server. Based
on that knowledge,
you will learn about
the various kinds of
deadlocks thatcan occur
in SQL Server, how to
troubleshooting them,
and how you can resolve
them by changing your
queries, your indexing
strategy and your
database settings.

6$6+$
*2/'6+7(,1

H7L;D:80J>;$D;J
DEIGB:7J787I;
The big data hype
is all about NoSQL
databases that can
support huge amounts
of data, replication,
scaling and super-fast
queries. RavenDB is the
best NoSQL database
for .NET developers,
because it has a firstclass .NET client with
a LINQ API. In this
session, Sasha will
show you how to model
data as documents for
storage in RavenDB;
how to query the data
efficiently; and how to
construct indexes that
will help you get the data
you need in just a few
milliseconds. Well also
review full-text search
and query suggestions
support, which make it
very easy to add search
capabilities.

'(-$1
6$5.$

:7J7
;NJH79J?ED7D:
JH7DI<EHC7J?ED
M?J>FEM;H
GK;HO7D:C
Power Query, a free
add-in for Excel 2010
and 2013 and part of the
Power BI suite in Excel
365, is a powerful tool.
Dejan will show how you
can use Power Query to
gather all kinds of data,
from databases to web
sites and social media,
inside Excel data models.
In this way, you can
make Excel an analysing
engine for structured
and unstructured data.
In addition to the queries
you can create through
the UI, there is a fully
functional language,
called M, behind the
scenes. This session
introduces both Power
Query and M.

'$1
&/$5.

9H;7J?D=IEB?:
FEM;HF?LEJ
:7J7CE:;BI
Self-service business
intelligence is gaining
popularity among
business analysts
today. It greatly
relieves the problems
created by traditional
data warehouse
implementations.
Using tools such as
Microsofts Power Pivot,
Power Query and Power
View alters the process
significantly. In this
session, Dan will take
you through the process
of creating a solid data
model in Power Pivot.
You will learn how to
import data from various
sources, combine
these in a scalable data
model, use DAX to
create measures, and
incorporate time-based
analysis.

| &2))((%5($. |

$//(1
+2/8%

C?9HEI;HL?9;I0
797I;IJK:O
In this session, Allen
will take a deep dive
into a micro-service
implementation.
Hell look at both the
architecture and the
implementation of
authentication and
comment-management
micro-services suitable
for use in a blog or
similar application. The
core system is written
in Java, so youll need
to know Java, C++, C#,
or equivalent to follow
along easily. Auxiliary
technologies include
Mongo, JavaScript,
AngularJS and
Bootstrap, so this session
provides a real-world
example of how those
technologies work. You
dont need to be familiar
with any of them, though.

+2:$5'
'(,1(5

=;JJ?D=F7IJ
J>;+&!-&3'(&
97B9KB7JEH?D
9K9KC8;H0'(J>?D=I
JEMEHAED
With the best intentions,
people have flocked to
behaviour-driven
development by way of
Cucumber over the past
few years, and thats a
great thing! But often,
BDD can fall by the
wayside due to the
pressure to deliver more
and more functionality,
sprint after sprint. In this
session, Howard will
explore 12 of the most
important issues, such
as imperative versus
declarative style, and
how to keep Gherkindriven Selenium
WebDriver tests
working dependably
through the use of
advanced ExpectedCondition techniques.

/81&+%5($. | &2))((%5($.

6$6+$
*2/'6+7(,1

CE:;HD?I?D=
9!!9E:;
The C++ standard
library dates back to
the 1990s, but that
doesnt mean there
arent new and exciting
frameworks to use in
your C++ application. In
this session, Sasha will
look at some brand-new
and some existing C++
libraries that can speed
up C++ cross-platform
development rapidly.
Some of the libraries
we might cover include:
Casablanca (C++ REST
SDK), Cinder (creative
coding), Boost (generalpurpose), Google Test
(unit testing), SOCI
(modern database
access) and many others.
Theres something for
everyone!

1812
*2'+,12

7H9>?J;9JKH;
8;IJFH79J?9;IED
M?D:EMI7PKH;
When new technologies
and paradigms appear,
its essential to learn
them quickly and
well. But this can be
difficult, since some
things are only learned
with experience. Thats
why best practices are
so important. In this
session, Nuno will look
at some architecture
best practices that
will help us make our
solutions better across
several levels, including
performance, cost,
integration, security
and so on. By doing this,
youll gain the knowledge
needed to quickly start
using the technology and
paradigms that can help
improve your business.

$17+21<
61(('

$1'<&/<0(5
5,&+$5'%/(:(77

IEKFJEDKJI0
:;L;BEF?D=H;7B#
MEHB:8KI?D;II
7FFIM?J>;DJ?JO
<H7C;MEHA7D:
7IF$D;JM;87F?
Performance. Scalability.
Maintainability.
Testability. Security.
Todays application
developers need to build
systems that are
designed to achieve
these goals from the
outset. In this in-depth
workshop, Anthony will
take you beyond the
basics, to learn how to
build RESTful services
that are robust, scalable
and loosely coupled, using
dependency injection
with repository and unit
of work design patterns.
But theres more
to building loosely
coupled systems than
applying a set of design
patterns. Anthony
will show you how to
harness the power
of code generation
by customising T4
templates for reverse
engineering Code First
classes from an existing
database, in order to
produce entities with
persistence concerns
that are completely
stripped away. Youll also
learn ninja techniques
for handling cyclical
references with codebased configuration
and using efficient
binary formatters,
all without polluting
your entities with
mapping, serialisation or
validation attributes.
This workshop will
focus on developing
real-world business
apps using the Entity
framework and ASP.
NET Web API.

IEB?:7IOD9
FHE=H7CC?D=
?D$D;J
Two-day workshop
continues from the
previous days session.
For a full description
ofthe workshop, please
seePage 20

10'&#;914-5*12


691&#;914-5*12

914-5*12
4'(/%

914-5*12
4'(/%

"&GX9GGM | YYYFGXYGGMEQO | 15

DevWeek_2015_Brochure.indd 15

1/5/15 10:53 AM

6*745&#;/#4%*

&#;#)'0&#%106+07'&



3(7(5
2+$1/21

8K?B:?D=D7JKH7B
KI;H?DJ;H<79;I
M?J>H;7BI;DI;
:;L?9;I
The Intel RealSense
SDK is the powerhouse
behind the perceptual
computing cameras
powering the next
generation of Ultrabook
devices. In this session,
Peter will take a look at
how RealSense devices
can provide unique
ways to interact with
applications, explaining
the advantages, and
offering tips and tricks
to building compelling
applications using
RealSense devices. You
will learn how to easily
create applications that
use gesture and facial
recognition, emotion
detection, as well as
speech recognition and
speech synthesis.



.(9/,1
+(11(<

=?L?D=9E:;7
=EE:D7C;
Code is basically made
up of three things:
names, spacing and
punctuation. With
these three tools, a
programmer needs to
communicate intent,
and not simply instruct.
But if we look at most
approaches to naming,
they are based on the
idea that names are
merely labels, so that
discussion of identifier
naming becomes little
more than a discussion
of good labelling. A
good name is more than
a label; a good name
should change the way
the reader thinks. Good
naming is part of good
design. In this session,
Kevlin will look at why
and what it takes to get a
good name.

16 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 16

$1'5(<
$'$029,&+

=HEELO:;LEFI
?DJ>;9BEK:
In this session, Andrey
will focus on a set of
tools to automate the
provisioning of (cloud)
servers using Groovy
libraries and Gradle
plug-ins. He will explore
how to leverage those to
create an infrastructure
for building, configuring
and testing the
provisioning of boxes in
the cloud elegant and
groovy. This session will
help those Java/Groovy
developers interested
in reusing their existing
skills for infrastructure
provisioning and
learning more about
problems encountered
during system
operations.

-2(
1$72/,

J>;8?=B?;0M>O
<EHC:E;IDJ7D:
I>EKB:DJ<EBBEM
<KD9J?ED
The prescriptive
interpretation of this
axiom has guided the
work of engineers,
programmers,
developers and even
designers for a very
long time. The result
of this has been sites,
software and systems
that exhibit poor
usability, frustrating
user experiences and a
marked failure to deliver
expected business
results. In this session,
Joe will show you why
pure function is rarely
the single or most
important component of
success. He will explain
how every force at play
in any project is what
really evolves form (and
dictates function).

6+$,
5(=1,.

'.J?FI<EHJ>;
7D=KB7H7H9>?J;9J
So your company is
planning to build a largescale web application,
and has chosen to do it in
Angular.js. That raises a
lot of questions: Where
do I start from? What
tools should I use? And,
basically, how do I avoid
making mistakes and do
the job efficiently? Shai
has worked with more
than 20 companies,
helping them with their
struggles migrating to
Angular and avoiding
crucial mistakes in the
process. In this talk,
Shai will present useful
time-saving architecture
tips that will help you
prepare for scalability,
write cleaner code, and
even make your life
happier.

',12
(6326,72


C7A?D=7D;N?IJ?D=
M;8I?J;7CE8?B;#
<H?;D:BOM;8I?J;
Do you feel frustrated
every time you run
across a web site
that doesnt adjust to
the viewport of your
current phone browser?
In some cases, for a
better experience, you
have to know that a
mobile version of the
site exists somewhere
with a different URL.
Theres no reason to
further delay plans to
make your primary web
site display nicely on
small-screen devices,
including smartphones.
In this session, Dino will
lead the discussion and
explore pros, cons and
technologies that could
make each option viable.

,'2
)/$72:

:;8K==?D=J>;M;8
M?J><?::B;H
Every web developer
needs to see what goes
on in the wire, whether
it is a jQuery call from
JavaScript, a WCF
service call from a client
app, or a simple GET
request for a web page.
With Fiddler, the most
famous HTTP sniffer,
this is simple enough
to do. But Fiddler is
more than just a sniffer.
With Fiddler you can
intercept, alter and
record messages, and
even write your own
message visualiser. In
this session, we will
learn how to use Fiddler
from bottom to top to
debug, test and improve
web applications.

*,/
),1.

J>;9>7H79J;H?IJ?9I
E<7IK99;II<KBIF7
Single-page applications
(SPAs) are web
applications that are
built using a single page,
which acts as a shell to
all the other web pages,
with a rich JavaScript
front-end. As opposed
to traditional web
applications, most of
the SPA development is
done on the front-end.
The server, which once
acted as a rendering
engine, provides only
a service layer to the
SPA. In this session,
Gil will explain the
characteristics and
building blocks that form
the foundation of any
successful SPA.

0,&+$(/
+$%(50$1

<HECN7CB%9
JE>JCB%@I
In recent years, the
mobile evolution caused
many developers
to find themselves
migrating from desktop
applications to web
applications. In this
session, Michael will
explore how to make the
transition from XAML
and C# to HTML5 and
JavaScript. He will
review how to port the
MVVM design pattern
to the web environment,
and go on to tackle
important architecture
concepts, such as
dependency injection
and modularity.

:7
;H
IE
Th
is n
wit
alo
pla
im
Lin
up
thi
tal
.NE
wh
cor
and
goi
app
als
and
of t
an
to m
com
tre

1($/
)25'

8K?B:OEKHEMD
J;9>DEBE=OH7:7H
ThoughtWorks
Technical Advisory
Board creates a
technology radar
twice a year: a working
document that helps
the company make
decisions about what
technologies are
interesting. This is a
useful exercise both for
you and your company.
In this session, Neal
will describe the radar
visualisation, how to
create litmus tests for
technologies, and the
process of building a
radar. Attendees will
leave with tools that
enhance your filtering
mechanisms for new
technology and help you
(and your organisation)
develop a cogent strategy
to make good choices.

"&GX9GGM

1/5/15 10:53 AM

@E?
#7I
Ha
an
pe
tw
yo
a
act
tw
Se
In
wi
ho
ha
Ini
the
joi
dig
joi
exe
att
yo
to
var
us

#.51'#%*&#;
&2))(( 5(*,675$7,21

6$6+$
*2/'6+7(,1

:7MDE<7D;M
;H707DEF;D#
IEKH9;$D;J
The .NET framework
is now open source,
with the CLR to follow
along and a crossplatform reference
implementation for
Linux and OS X to show
up during the year. In
this session, Sasha will
talk about the future of
.NET in this new era,
what it means for the
core stack on the server
and desktop, and how its
going to affect our .NET
applications. We will
also see how to build, test
and run our own version
of the .NET framework
and CLR, and how
to make changes to
components previously
treated as a black box.

./$86
$6&+(1%5(11(5

@E?DI?DIGBI;HL;H
#7I;7IO7I7895
Have you ever looked at
an execution plan that
performs a join between
two tables? And have
you ever wondered what
a Left Anti Semi Join
actually is? Joining
two tables using SQL
Server is far from easy!
In this session, Klaus
will take a deep dive into
how join processing
happens in SQL Server.
Initially, he will lay out
the foundation of logical
join processing, then
dig deeper into physical
join processing in the
execution plan. After
attending this session,
you will be well prepared
to understand the
various join techniques
used by SQL Server.

'(-$1
6$5.$

L?IK7B?I?D=
=H7F>?97B7D:
J;CFEH7B:7J7
M?J>FEM;HC7F
Power Map is a new 3D
visualisation add-in
for Excel, used for
mapping, exploring
and interacting with
geographical and
temporal data. Power
Map exists as free
preview add-in for Excel
2013 and in the Power
BI suite in Office 365.
In this session, Dejan
will explain how to
use Power Map to plot
geographic and temporal
data visually, analyse
that data in 3D, and
create cinematic tours to
share with others.

*$5<
6+257

JHEBB>KDJ?D=ED
J>;?DJ;HD;J
With so many people
on social media these
days, almost inevitably
not a day goes by
without some tragedy
befalling someone. As
if that wasnt horrible
enough, these poor
souls and their families
can then become
victims of the perverse
behaviour of the
trolls. In this session,
Gary will examine
this problem from a
data scientists point
of view, showing how
to use computational
linguistics to ensure that
such posts never reach
peoples streams, and
network theory to trace
and expose the trolls so
that they no longer have
the shield of anonymity
to hide behind.

| &2))((%5($. |

-8/(6
0$<

FHE8B;CIF79;
7D7BOI?I
How do you design
a large system? The
architecture of any
system is crucial to
its success get this
wrong, and the project
may never recover. And
yet, we are expected
to deliver designs
that can last five, 10,
sometimes 30 years into
an unknowable future.
Problem space analysis
is a technique that
informs and documents
system designs by
anticipating and defining
the variabilities of a longlived, evolving system.
In this session, Jules will
explain the principles
of the method, give an
outline of the benefits,
and demonstrate its
power with some
illustrative examples.

6(%
526(

CED7:7JJ>;
>OF;H8EB;7D:
EJ>;H7M;IEC;
IJEH?;I
In this session, Seb will
help you to be a better
software developer. As
software developers, we
have to deliver
something useful to our
customers. We have to
produce it in a manner
that acknowledges their
requirements and
context. And usually, we
need to be able to work
as part of a team. Are my
customers more likely to
be satisfied if Im
awesome? In what
circumstances would
monads (or any other
implementation level
detail) be a critical part
of a successful solution?
Seb will analyse
real-world examples
ofprojects that
succeeded and failed.

/81&+%5($. | &2))((%5($.

786+$5
6+$50(5

:E;IOEKH:;I?=D
IC;BB5
In this session, Tushar
will propose a unique
approach to developing
high-quality software
design. Borrowing
a phrase from the
healthcare domain,
agood doctor is one who
knows the medicines
but a great doctor is
one who knows the
disease, the proposed
approach is grounded
on the philosophy that
a good designer is one
who knows about the
design principles but
a great designer is one
who understands the
problems (or smells)
with the design, their
cause, and how they
canbe addressed by
applying proven and
sound design principles.

$17+21<
61(('

$1'<&/<0(5
5,&+$5'%/(:(77

IEKFJEDKJI0
:;L;BEF?D=H;7B#
MEHB:8KI?D;II
7FFIM?J>;DJ?JO
<H7C;MEHA7D:7IF$
D;JM;87F?
One-day workshop
continues from the
morning session.
For a full description
ofthe workshop, please
seePage 21

IEB?:7IOD9
FHE=H7CC?D=
?D$D;J
Two-day workshop
continues from the
morning session.
For a full description
ofthe workshop, please
seePage 20

,'2
)/$72:

C?=H7J?D=
7FFB?97J?EDIJE
C?9HEIE<J7PKH;0
B;IIEDIB;7HD;:
<HECJ>;<?;B:
How much time will
it take us to move to
Azure? Can we just Lift
& Shift our servers?
Will my load-balancer
work in Azure? Should
I use SQL Databases
or an SQL Server VM?
These are just some of
the questions customers
ask when they consider
migrating their
applications to Azure. If
youre evaluating Azure,
come to this session,
where Ido will explain
what to do, what not to
do, what to avoid and
what to embrace when
moving your apps to
Azure. These are not
general best practices;
these are lessons learned
from the field.

10'&#;914-5*12


691&#;914-5*12

914-5*12
4'(/%

914-5*12
4'(/%

"&GX9GGM | YYYFGXYGGMEQO | 17

DevWeek_2015_Brochure.indd 17

1/5/15 10:53 AM

(4+&#;/#4%*

2156%10('4'0%'914-5*125

&GX9GGM
EQPENWFGU
YKVJQWTPCN
FC[QHRQUV
EQPHGTGPEG
YQTMUJQRU



1($/
)25'

#UQP
RTGXKQWUFC[U
CNNYQTMUJQRU
TWPHQTC
HWNNFC[
HTQO
VQ
YKVJUJQTV
DTGCMUKPVJG
OQTPKPICPF
CHVGTPQQP
CPFCNWPEJ
DTGCMCV


9EDJ?DKEKI
:;B?L;HO

* U H D W  O H F W X U H V  
Y H U \  L Q I R U P D W L Y H  
)D Q W D V W L F  Y H Q X H
@KD?EHM;8:;L;BEF;H
(&'*:;B;=7J;

Getting software released

to users can be painful,
risky and timeconsuming. Here, Neal
sets out the principles
and practices that enable
rapid, incremental
delivery of high-quality,
valuable new functionality
to users. By automating
the build, deployment
and testing process, and
improving collaboration
between developers,
testers and operations,
delivery teams release
changes in a matter of
hours or even minutes.
Neal will look at the
differences between
related topics such as
continuous integration,
continuous deployment
and continuous delivery,
and explore the new
technical artefact that
continuous delivery
introduces: the
deployment pipeline.
Hell discuss the various
stages, how triggering
works, and how to
pragmatically determine
what production
ready means, before
covering the role of
testing and the testing
quadrant, including
the audience and
engineering practices
around different types
of tests. This is followed
by version control usage
and offering alternatives
to feature branching,
such as toggle and
branch by abstraction.
Neal will then go on to
cover operation, DevOps
and programmatic
control of infrastructure,
using tools such as
Puppet and Chef.
914-5*12
4'(&9

18 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 18

-2(
1$72/,

=;D;H7J?D=
C;7D?D=<KB
H;GK?H;C;DJI

Ask any group of people

what they want or
need and youll find no
shortage of opinions or
answers. Clients and
stakeholders will always
have a voluminous
laundry list of features
and functions, all
of which they will
insist are equally
important. Your clients,
employers, project
stakeholders and users
all share something very
important in common:
theyre all human beings.
And we human beings
all have a fundamental
flaw: we often make very
confident but equally
false predictions about
our future behaviour.
So the requirements
that will actually be most
useful and most valuable
the ones that will
increase user adoption
or sales; the ones that
will make or save money
are almost never
surfaced in traditional
requirements sessions.
In this workshop, Joe
will show you how to
change that, along with
how to tell the difference
between what people say
they need and what they
actually need. Finally,
hell show you how to
uncover the things they
dont know they need
(but absolutely do).

914-5*12
4'(&9

-8/(6
0$<

FHE8B;CIF79;
7D7BOI?I

How do you design a

large system? We know
Waterfall doesnt work
very well; yet we also
know that Agile scales
poorly. Various proposals
have been made (BDUF,
domain-driven design,
prototyping) but none
really solves the problem.
The key to managing
a large system is
managing change. No
specification ever
survives its own
implementation: as a
system takes shape,
everyone developers,
architects, stakeholders
change their minds. In
any non-trivial project,
goalposts are constantly
in motion. A robust
architecture is one that
anticipates those changes,
and a good design is one
that accommodates them
cheaply and efficiently.
Problem space
analysis is a technique
that simply and clearly
anticipates, documents
and defines the changes
that can affect a project.
It informs the
architectural design so
that it can accommodate
those changes, and it
delivers a changetolerant ubiquitous
language to unify and
coordinate the
development effort.
Jules introduces the
principles of problem
space analysis and how
they translate into
architectures and
working systems, even
while the goalposts are
moving. The technique
will be actualised using a
real-life design problem.
914-5*12
4'(&9

786+$
6+$50$

79>?;L?D=
:;I?=D7=?B?JO
8OH;<79JEH?D=
:;I?=DIC;BBI

Software design plays

an important role in
adapting changing
requirements for
software development.
Software products
that follow Agile
methodologies are
no exception. Agile
methodologies welcome
rapid requirement
changes and yet promote
timely delivery. Often,
software teams follow
these practices (to some
extent) but ignore the
software design and its
quality due to negligence
or ignorance. Design
agility suggests that
the software design has
to be Agile if the team
intends to follow Agile
practices in a true sense.
In the absence of agility
of design, its difficult to
achieve the benefits of
being Agile.
Tushar emphasises
the importance of
design agility by
exposing design smells
in software systems. He
explains design smells
and connects them to
their impact on software
design, design agility,
bug-proneness and
delivery schedule of the
software.
This workshop offers
an understanding of the
importance of software
design quality and design
agility. Youll also learn
the vocabulary of smells,
using a comprehensive
classification, and
practical refactoring
strategies to repay
technical debt.

914-5*12
4'(&9

J>
JE
JH
7I
7F

Is y
wo
an
you
app
wo
AS
you
thi

aw
tric
des
It r
fro
an
bet
sin
are
Jav
cod
to u
acc

wi
of t
we
dem
use
tra
an
the
ho
we
pro
an
to t
the
NE
an
.NE
IIS
tip
Jav

91
4'

"&GX9GGM

1/5/15 10:53 AM

#.51'#%*&#;
&2))(( 5(*,675$7,21

,'2
)/$72:

J>;;II;DJ?7B
JEEB8EN<EH
JHEK8B;I>EEJ?D=
7IF$D;JM;8
7FFB?97J?EDI

6$6+$
*2/'6+7(,1

C7A?D=J>;CEIJ
E<9!!''%'*

Is you web application

working slower than
anticipated? Have
you rewritten your
application code, but still
wonder if there is some
ASP.NET or IIS trick
you can use to boost
things up?
Testing and profiling
a web application is
trickier than with
desktop or mobile apps.
It requires testing both
front-end and back-end,
and the network inbetween the two. And
since web applications
are a mix of HTML, CSS,
JavaScript and .NET
code, you often need
to use several tools to
accomplish this task.
In this workshop, Ido
will go over the process
of testing and profiling
web applications, and
demonstrate how to
use various tools of the
trade both for front
and back ends. Ido will
then go into detail on
how to speed up your
web applications, by
providing important tips
and tricks you can apply
to the different parts of
the application: ASP.
NET (Web Forms, MVC
and Web API), general
.NET best practices,
IIS server, networking
tips, HTML/CSS and
JavaScript.

In this workshop, Sasha

will look at the most
important C++ language
features that improve
system performance and
developer productivity,
and see how to apply
them to existing code.
The C++11 standard
is already behind us,
and C++14 is just
around the corner.
With a huge variety
of language features
such as lambdas, rvalue
references, auto and
decltype, and variadic
templates, its easy to
get lost in C++. In fact,
it often seems like a
completely new and
foreign language. We
will make the most of
Visual C++ 2013 and
give a special focus
to converting and
refactoring code to use
modern C++ idioms.
Specifically, we will
look at how to best use
STL algorithms with
lambda functions,
when to use each kind
of smart pointer class,
how to convert macros
and non-generic code
to templates, and a
variety of best practices
concerning concurrency
in C++ applications.
This will be a
particularly relevant
workshop for C++
developers who watched
the C++ 11/14 train
passing by and werent
able to apply all the best
practices of modern C++
to their applications
just yet.

914-5*12
4'(&9

914-5*12
4'(&9

| &2))((%5($. |

'(-$1
6$5.$

8?M?J>
C?9HEIE<JJEEBI0
<HEC;DJ;HFH?I;
JE7F;HIED7B
B;L;B

Microsoft has really

been investing a lot in
business intelligence
(BI) over the last 10
years. The result is
a huge number of
analytical tools and
services. When building
a BI solution, many
companies make basic
mistakes and choose
inappropriate tools for
the problem they are
trying to solve.
In this workshop,
Dejan will help you put
the building blocks into
the right context. You
will learn about data
warehousing with SQL
Server, reporting with
SQL Server Reporting
Services (SSRS), Power
View and Power Map,
on-line analytical
processing (OLAP) with
SQL Server Analysis
Services (SSAS),
Multidimensional,
Tabular, Power Pivot for
Excel, and Power Pivot
for SharePoint Server,
data mining with SSAS,
Excel, R and Azure
Machine Learning
(ML), and about the
extract-transform-load
process with SQL Server
Integration Services
(SSIS) and Power Query.

914-5*12
4'(&9

/81&+%5($. | &2))((%5($.

*,/
),1.

6(%
526(

$//(1
+2/8%

8K?B:?D=
I97B78B;
@7L7I9H?FJ7FFI

J>;IM?<J
FHE=H7CC?D=
B7D=K7=;

Swift is Apples
new programming
language. It represents a
significant improvement
over both Objective C
and C++; incorporating
many contemporary
language features
(such as duck typing
and closures) without
abandoning object
orientation. It promises
to become the dominant
language of the Apple
platform.
Swift is, unfortunately,
a mixed bag. It supports
many important
language features, but it
also omits features that
are essential for longterm maintainability
(privacy, for example).
In this workshop,
geared to programmers
who already know an
OO language (C++,
Java, C#, etc.), Allen
will present all the
interesting parts of the
Swift language. Well
gloss over the basic
stuff (declarations,
flow control, etc.) and
focus on those parts
of the language that
will be new to you
(lambdas, subscripts,
the inheritance model,
extensions and chaining,
etc.), with considerable
emphasis on places
where the language can
get you into trouble.

Its all very well reading

books, but nothing beats
actually getting practical
experience. In this
workshop, working in
your choice of Java, C#
or Ruby, Seb will drive
out the implementation
of a simple utility by
specifying its behaviour
in Cucumber. The
tyrannical Product
Owner will regularly
change his mind, so well
need to keep our code
well factored and easy to
modify.
This session is
designed for developers
and testers. By the end
of the day, youll be
comfortable working
with Cucumber in your
chosen development
environment. Youll have
seen, first hand, how to
use Cucumber to drive
out valuable features
for your customers
and how that can help
keep your stakeholders
engaged in the software
development process.
It will also be clear how
BDD interacts with TDD.
Bring a laptop
with your chosen
development
environment installed,
and try to pre-install
your chosen Cucumber
variant before you come
(instructions available).
But dont worry, we
can install on the day if
necessary.
If this subject
interests you, but youd
prefer a gentler intro,
dont miss Sebs earlier
workshop on Monday:
BDD by example.

914-5*12
4'(&9

914-5*12
4'(&9

914-5*12
4'(&9

Building and
maintaining large
andscalable JavaScript
web apps isnt easy.
So how do you build
your front-endoriented applications
without being driven
to madness? Using
and combining proven
JavaScript patterns will
do the trick.
In this workshop,
Gil will discuss the
patterns behind some
of the largest JavaScript
apps, such as Gmail
and Twitter, and well
explore how to apply
them in your own apps.
Well start from object
patterns and discuss
how to write more
object-oriented-like
code in JavaScript.
Then, we will focus
on module patterns
and asynchronous
module definitions
(AMD). Wewill also
discuss patterns such
as promises, timers and
mediator. At the end of
the day, we will combine
the patterns and see how
to use them to build your
next scalable JavaScript
webapp.

7FFB?;:8::M?J>
9K9KC8;H%@7L7
7D:IF;9<BEM%9

"&GX9GGM | YYYFGXYGGMEQO | 19

DevWeek_2015_Brochure.indd 19

1/5/15 10:53 AM

/#+0%10('4'0%'

/10&#;4&/#4%*
914-5*125

&GX9GGMUOCKPEQPHGTGPEGFC[UPQYQHHGTOQTGEJQKEGVJCPGXGTDGHQTG#NQPIUKFGVJG
DTGCMQWVUGUUKQPU[QWECPEJQQUGHTQOCURGEKCNVYQFC[YQTMUJQRVJCVTWPUCETQUUDQVJ
9GFPGUFC[CPF6JWTUFC[CPFQPGFC[YQTMUJQRUVQEJQQUGHTQOQPGCEJFC[

691&#;914-5*12

9'&6*745
Collections package, now available on NuGet,
which provides another way to model data for
async applications.

$1'5(:
&/<0(5

5,&+$5'
%/(:(77

IEB?:7IOD9FHE=H7CC?D=
?D$D;J
In this special two-day workshop, Andrew
and Richard will take you through the core
skills required to be successful developing
async and multithreaded code, both in the
.NET and web worlds. Not only do we cover
the core APIs but also how they are used
effectively, tested and debugged.
Tasks
When the Parallel Framework Extensions (PFx)
were first announced, it looked as though they
were going to target a narrow set of requirements
around parallelising processor intensive code.
Over time, the scope of the library has grown
significantly, such that it will become the main
model for building asynchronous code. The
pivotal type enabling this transition is the Task
class. This is a functionally very rich type,
allowing the creation of both short- and long-lived
asynchronous work, Tasks can have dependencies
on one another and support cancellation. In this,
the first of the PFx modules we look specifically
at how this class gives us a unified framework for
building multithreaded code.
Thread safety
Asynchronous programming requires careful
attention to detail, since most objects are
not designed with multithreaded access
in mind. This module introduces the
importance of Interlocked and Monitor-based
synchronisation.

$  Z R Q G H U I X O  Z D \
W R  U H W K L Q N  Z K D W
L W  P H D Q V  W R  E H  D
JUHDWGHYHORSHU
IE<JM7H;:;L;BEF;H
(&'*:;B;=7J;

20 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 20

Concurrent data structures

Ever since its inception, .NET has had support
for a number of synchronisation primitives (such
as Interlocked, Monitor and Mutex). However,
on their own, these primitives do not provide
support for more complex synchronisation
situations, and so people have had to use them as
building blocks to build things such as efficient
semaphores. PFx finally brings to the library a set
of richer primitives, such as lazy initialisation, a
lightweight semaphore and a countdown event.
But more than this, it also introduces a set of
high-performance concurrent data structures
that allow you to use them without you having
to provide your own synchronisation logic
around them. We also look at the new Immutable

Parallel
The initial goal of PFx was to simplify the
parallelisation of processor intensive tasks and this
remains a key feature. This part of its functionality is
focused on the Parallel class and its For and ForEach
members. In this module, we look at the simplified
model but also highlight that parallelising algorithms
is never as simple as it might first seem we show
you some of the pitfalls that you should be aware of
when trying to parallelise functionality using the
Parallel class.
async/await
C# 5 builds on the Task API, introducing async
and await keywords, which bring asynchronous
execution as a first-class concept in the C# language.
These new keywords create a very elegant model for
all sorts of async work and this module explains not
only how to use them but also how they work under
the covers.
Server-side async
The nature of server-side applications often means
they are asynchronous by their very design, servicing
many clients at the same time on different threads.
But as you dig deeper, you often find these threads
performing long-running blocking operations
particularly in terms of IO resulting in consuming
more threads (an expensive resource) than
necessary. This module focuses on a range of serverside technologies and demonstrates how to perform
maximum concurrency for least number of threads.
TPL Dataflow
TPL Dataflow is a downloadable addition to the
TPL (Task Parallel Library) that ships with the
.NET framework. TPL Dataflow provides an
alternative approach to define concurrency. Instead
of just simply throwing threads at a synchronously
structured program and having to deal with all the
thread safe and race conditions that introduces; we
have the concept of many autonomous objects, each
with its own thread of execution. These autonomous
objects co-operate with other such objects through
asynchronous message passing. In this module, we
will see how TPL Dataflow can greatly reduce the
complexity normally associated with asynchronous
programming.
Rx
Reactive Framework is a new library that uses the
.NET 4.0 IObservable interface and LINQ to create a
compelling new programming model that allows you
to build event-based code with declarative LINQ
statements. This module introduces the Reactive
Framework and shows how it can greatly simplify
your code.

"&GX9GGM

1/5/15 10:53 AM

#.51'#%*&#;
&2))(( 5(*,675$7,21

| &2))((%5($. |

/81&+%5($. | &2))((%5($.

J
10'&#;914-5*125

9'&0'5&#;

B;7HD7D=KB7H@I0
J>;HE7:JEFEM;H<KB"C7?DJ7?D78B;7FFB?97J?EDI
JavaScript, by its nature,
makes it difficult to write
maintainable code. HTML, by
its nature, is loosely structured.
AngularJS fixes both of those.
6$+,/
It is a structural framework
0$/,.
for dynamic web apps,
allowing you to extend HTMLs
syntax, enabling you to write powerful, maintainable
applications succinctly.
In this workshop, Sahil will build on your existing knowledge
of JavaScript and teach you the ins and outs of AngularJS.
There are plenty of examples, which will walk you through a
basic introduction, models, controllers and views in Angular,
templates and databinding, services and dependency injection,
directives, routing and single-page applications.
An introduction to AngularJS
AngularJS is a structural framework for dynamic web apps.
It lets you use HTML as your template language and lets
you extend HTMLs syntax to express your applications
components clearly and succinctly. This module gets you
started with AngularJS showing the basic syntax and some
quick starts to get us running.
Models, controllers and views
Controllers in AngularJS are a fundamental building block
of Angular. AngularJS encourages better code architecture
by encouraging you to use MVC-based patterns. There is
rich support for controllers and views, and this module will
familiarise you with those.

Templates and databinding

When working with complex forms, or even reactive user
interfaces, it really helps to leverage concepts such as
databinding. Databinding in JavaScript is especially valuable
but can be difficult to implement, unless you use something like
Angular.
Services in Angular, dependency injection
Services in Angular are substitutable objects that are wired
together using dependency injection. Fancy! What on Earth
does that even mean? Well for one, services are a classic way
of bundling together reusable code across controllers in your
application, and they can be lazily instantiated or singletons.
There are services that Angular provides, such as $http, or ones
you can create. And then there is something called interceptors.
This module gets in in the deep of AngularJS services.
Directives
Directives in AngularJS are markers on DOM elements that
tell the AngularJS HTML compiler to attach some specified
behaviour to that particular DOM element or its children. There
are many directives, such as ngBind, ngModel, ngClass, and this
module shows you the most important directives.
Routing and single-page applications (SPAs)
AngularJS routes enable you to create different URLs for
different content in your application. Having different URLs for
different content enables the user to bookmark URLs to specific
content. Single-page applications allow you to write HTML and
JavaScript code that more or less performs and behaves like
a thick client application. Its pretty neat trust me, youll be
using this.

9H U \  L Q I R U P D W L Y H   F R Y H U L Q J  D  E U R D G  U D Q J H  R I  W H F K Q R O R J \ 
D Q G  D  J R R G  R S S R U W X Q L W \  W R  H Q J D J H  Z L W K  L Q G X V W U \  H [ S H U W V
IE<JM7H;7H9>?J;9J"(&'*:;B;=7J;
6*745&#;

IEKFJEDKJI0:;L;BEF?D=H;7B#MEHB:8KI?D;II7FFI
M?J>;DJ?JO<H7C;MEHA7D:7IF$D;JM;87F?
Performance. Scalability.
Maintainability. Testability.
Security. Todays application
developers need to build
systems that are designed
$17+21<
to achieve these goals from
61(('
the outset. In this in-depth
workshop, Anthony will take
you beyond the basics, to learn what it takes to build
RESTful services that are robust, scalable and loosely
coupled, using dependency injection with repository and
unit of work design patterns.

But theres more to building loosely coupled systems than applying

a set of design patterns. Anthony will show how you can harness
the power of code generation by customising T4 templates for
reverse engineering Code First classes from an existing database,
in order to produce entities with persistence concerns that are
completely stripped away. Youll also learn ninja techniques for
handling cyclical references with code-based configuration and
using efficient binary formatters, all without polluting your entities
with mapping, serialisation or validation attributes.
This workshop will focus primarily on developing real-world
business apps using the Entity framework and ASP.NET Web API.

"&GX9GGM | YYYFGXYGGMEQO | 21

DevWeek_2015_Brochure.indd 21

1/5/15 10:53 AM

/10&#;4&/#4%*
/''6;17452'#-'45

&GX9GGMUURGCMGTUCTG
CEMPQYNGFIGFGZRGTVU
KPVJGKTGNF4GEQIPKUGF
KPVGTPCVKQPCNN[VJG
URGCMGTHCEWNV[
EQORTKUGURTQHGUUKQPCN
EQPUWNVCPVUVTCKPGTU
KPFWUVT[XGVGTCPU
VJQWIJVNGCFGTUCPF
RWDNKUJGFCWVJQTU(KPF
QWVOQTGCDQWVVJGOJGTG
$0<
&+(1*
Amy is a web developer
at the Brooklyn Museum
of Art in New York City,
working to increase
dialogue between the
museum and its visitors
through technology, and
has been a mentor for
the non-profit Girls Who
Code. She is interested in
using code to create art.

,'2
)/$72:
Ido is a senior architect
and trainer at SELA
Group, a Microsoft ASP.
NET/IIS MVP, and an
expert on Microsoft
Azure and web
technologies such as
WCF, ASP.NET and IIS.
He has co-authored a
number of books and
official Microsoft courses.

6$1'(5
+22*(1'2251
Sander is the author
of the best-selling
book ThisIs Agile. An
independent mentor,
trainer, programmer,
architect, speaker
and writer, Sander
is a catalyst in the
innovationof software
development at many
international clients.

22 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 22

'$1
&/$5.
Dan is a senior business
intelligence (BI)/
programming consultant
specialising in Microsoft
technologies. A former
physics teachers, he has
written several books and
numerous articles on
.NET programming and
BI development, and is a
regular conference speaker.

1($/
)25'
Neal is director,
softwarearchitect
andmeme wrangler at
ThoughtWorks, a global
IT consultancy focusing
on end-to-end software
development and
delivery. Hes the author
of applications, articles,
and books on a variety of
subjects and technologies.

$//$1
.(//<
Allan has held just about
every job in software,
before joining Software
Strategy to help teams
adopt and deepen Agile
practices. He has written
books including Xanpan
Team-centric Agile
Software Development,
and is a regular speaker
and journal contributor.

52'#-'4241(+.'5
$1'5(<
$'$029,&+
Andrey is a software
craftsman with years of
experience. His true love
is the JVM ecosystem,
and applying it to his
companys DevOps
initiatives. He is one
of the authors of the
Groovy 2 Cookbook,
and a frequent speaker
at conferences.

$1'5(:
&/<0(5
Andy is a co-founder of
Rock Solid Knowledge,
creating Kiosk-based
solutions on Windows
Embedded with
.NET. He cut his teeth
programming on a host
of platforms at various
start-ups, and now
consults and teaches for
adiverse range of clients.

6+$<
)5,('0$1
Shay is a Visual C#/
IronRuby MVP and the
author of IronRuby
Unleashed. With more
than 10 years experience
in the software industry,
he is the co-founder of
CodeValue, a company
that creates products for
developers, consults and
conducts courses.

0,&+$(/
.(11('<
Michael is an author,
instructor and technical
curriculum director
at DevelopMentor,
and lead developer
for its online training
platform, LearningLine.
He has been building
commercial applications
with .NET since its initial
public beta in 2001.

./$86
$6&+(1%5(11(5
Klaus provides
independent SQL Server
Consulting Services
across Europe and the
US. He has worked
with SQL Server
2005/2008/2012/2014
from its very beginning,
and has also written
the book Pro SQL
Server2008.

('
&2857(1$<
Ed is an experienced
software developer and
technical evangelist who
has been programming
professionally for
more than 25 years.
He currently works for a
major manufacturer and
retailer in the UK, leading
the team responsible for
its ecommerce web site.

1812
*2',1+2
Nuno is Director of
Cloud Services, Europe
at Aditi Technologies,
and has more than
16 years experience in IT.
His specialities include
enterprise architecture
and solution architecture,
cloud computing,
development
and training.

72%,$6
.20,6&+.(
Tobias is Senior Director
of User Experience at
Infragistics, Inc., and
has worked in user
experience for more than
10 years. He specialises
in Human Factors
Engineering, which is
rooted in his academic
background in cognitive
psychology.

$867,1
%,1*+$0
Austin is a founder of
Sixty North, a Norwaybased software consulting,
training and application
development company.
Developer of industryleading oil reservoir
modelling software
in C++ and Python,
he is anexperienced
presenterand teacher.

+2:$5'
'(,1(5
Howard is a software
consultant and educator
who specialises in Agile
process and practices.
With a career spanning
more than 30 years, hes
been a developer, analyst,
team lead, architect and
project manager, and is a
long-standing member
of the ACM and IEEE.

6$6+$
*2/'6+7(,1
Sasha is the CTO of
SelaGroup, a Microsoft
C# MVP and Azure
MRS, a Pluralsight
author, and an
international consultant
and trainer. The author
of two books, Sasha is
a prolific blogger and
author of numerous
training courses.

,4%$/
.+$1
Iqbal is the President
and Technology
Evangelist of software
developer Alachisoft,
maker of NCache,
the industrys leading
open-source distributed
cache for .NET. NCache
is also available for
Microsoft Azure.

5,&+$5'
%/(:(77
Richard has worked on
distributed systems,
including as middle-tier
architect on the UK
national police systems.
He focuses on technologies
that enable developers to
build large-scale systems
on the Microsoft platform,
such as WCF, BizTalk,
Workflow and Azure.

',12
(6326,72
Dino is a trainer,
speaker, consultant and
author. CTO of Crionet,
a company providing
software and mobile
services to professional
sports, Dino is also
technical evangelist
for software developer
JetBrains, focusing on
Android and Kotlin.

0,&+$(/
+$%(50$1
Michael (MCT, MCPD)
is a senior consultant and
lecturer specialising in
rich client technologies
such as WPF, Windows
Phone, XNA and
HTML/JS. He has
helped to develop
complex infrastructures
using Prism, MVVM
andAngular.

3+,/
/(**(77(5
Phil is a Developer
Evangelist at Caplin
Systems, working on
the BladeRunnerJS
open source project.
Hewrites frequently
andspecialises in
JavaScript development
and real-time web
technologies.

3($5/
&+(1

-'

Pearls cross-disciplinary
approach ranges from
Android to Arduino,
HTML to LEDs. Her
work has taken her from
Facebook campaigns for
Google Chrome to projects
that turn payphones into
gumball machines or
dynamically create origami
from SMS messages.

*,/
),1.

&
0

Gil is a web development

expert, ASP.NET/
IIS Microsoft MVP
and the founder of
sparXys. He consults for
various enterprises and
companies, where he
helps to develop web and
RIA-based solutions,
and conducts lectures
and workshops.

Ch
eng
10
ma
the
wo
inc
Str
Am
Ign
and

6+
5(

'525
+(/3(5
Dror is a senior
consultant at software
company CodeValue,
with a decade of
experience ranging from
Intel and SAP to small
start-ups. He evangelises
Agile methodologies and
test-driven design in his
work, at conferences
andas a consultant.

Sha
cons
ente
help
and
proj
foun
ano
web
fron
with

5
6

6$+,/
0$/,.
Sahil, the founder
and principal of
Winsmarts.com, has
been a Microsoft MVP
and INETA speaker
for 11 years. Author of
books and articles about
Microsoft technologies,
iOS and JavaScript,
Sahil helps make the
most difficult topics fun.

"&GX9GGM

1/5/15 10:53 AM

Rob
dire
a so
con
Nor
in s
tech
role
dea
of in
from

#.51'#%*&#;
&2))(( 5(*,675$7,21

Kevlin is an independent consultant

and trainer based in the UK. His
development interests are in patterns,
programming, practice and process.
He is co-author of two volumes
in the Pattern-Oriented Software
Architecture series, editor of the book
97 Things Every Programmer Should
Know, and a columnist for various
magazines and web sites.

om
for
jects
nto

ment

for
nd

and

om
l
ises
and
his

VP

f
out
ies,

fun.

/81&+%5($. | &2))((%5($.

-';016'52'#-'45

nary
m

gami

| &2))((%5($. |

.(9/,1
+(11(<

&+5,6726
0$76.$6
Christos is a software
engineer with more than
10 years experience
mainly focusing on
the .NET stack. He has
worked with big names
including MarkIT,
Strathclyde University,
Amor/Lockheed Martin,
Ignis Asset Management
and Barclays.

-8/(6
0$<
Jules is a software
architect with a particular
interest in languages
(both for programming
and discourse), presently
active in web and mobile
convergence. He has been
writing, teaching and
speaking for 25 years,
andis the originator of
Problem Space Analysis.

6+$,
5(=1,.
Shai is an AngularJS
consultant working with
enterprise companies,
helping with migration
and building large-scale
projects. He recently
founded HiRez.io,
anonline training
web site teaching
front-end architecture
with humour.

6(%
526(
Seb focuses on helping
teams adopt and refine
their Agile practices. The
founding trainer with
Kickstart Academy, he
has more than 30 years
industry experience
(including IBM Rational
and Amazon), and is
a popular speaker at
international meetings.

52%(57
60$//6+,5(
Robert is a founding
director of Sixty North,
a software product and
consulting business in
Norway. He has worked
in senior architecture and
technical management
roles, providing tools for
dealing with the masses
of information flowing
from todays energy sector.

0$5.
60,7+
Mark runs the curriculum
team at Xamarin
University, building and
managing the growing
course catalogue
used to train Xamarin
developers all over the
world. He is a Microsoft
MVP, Wintellect
author andXamarin
Consultingpartner.

-$0(6
0217(0$*12
James is a Developer
Evangelist at Xamarin.
He has been a .NET
developer for more than
a decade, working in
industries including
games development,
printer software and web
services, with several
published apps on iOS,
Android and Windows.

25(1
58%,1
Oren has more than 16
years experience with
IBM, Cadence, Wix and
others, and is the founder
of Testim.io. He regularly
speaks about new
technologies in web
development and test
automation, and teaches
at the Techion, Israel
institute of Technology.

$17+21<
61(('
Tony is a course author,
instructor and consultant
for Wintellect, specialising
in robust, scalable and
maintainable applications
using Entity framework,
WCF, Windows Identity
Foundation and ASP.NET
Web API. He is the
authorof two popular
open-source frameworks.

Allen is an internationally
recognised consultant, trainer,
speaker and author. He specialises
in lean/Agile processes and culture,
Agile-focused architecture and
cloud-based web-application
development. He has written a
dozen books, hundreds of magazine
articles, and currently blogs on
Agilefor Dr Dobbs Journal.

$//(1
+2/8%

-2(
1$72/,
Joe has been preaching
and practising the gospel
of user and customer
experience to Fortune
100, 500 and government
organisations for more
than 25years. As founder
of Give Good UX, he
offers coaching, training
and product audit
programmes.

'(-$1
6$5.$
Dejan, MCT and
SQL Server MVP,
is an independent
consultant, trainer and
developer focusing on
database and business
intelligence applications.
He specialises in topics
like data modeling, data
mining and data quality,
and has written 13 books.

$'$0
7251+,//
Adam combines degrees
in engineering and
psychology for a different
perspective on software.
An architect and
programmer, he writes
open-source software in a
variety of languages, and
is the author of Your Code
as a Crime Scene.

3(7(5
2+$1/21
Peters fascination
withnew technologies
has seen him writing
articles for CodeProject,
blogging and contributing
to open-source projects.
He was recently made an
Intel Software Innovator
for his work with
RealSense technology.

786+$5
6+$50$
Tushar is a technical
expert at the Siemens
Research and Technology
Center in Bangalore,
India. His research into
software design, design
smells and refactoring
has resulted in several
patents, research papers
and tools.

5$/3+
'(:$5*1<
Ralph has spent 10 years
focusing on the software
development market at
Intel, helping companies
and institutions to
maximise application
performance and move
code from serial to parallel.
He has spoken at TechEd,
TechDays, OOP and other
conferences across Europe.

721,
3(75,1$

-2+1
.3$8/
John is the VP of
engineering at Penton
Media and former lead
technical architect of
Cond Nasts platform
engineering team.
Healsoorganises the
NYC HTML5 meetup
group, and contributes
toa number of opensource projects.

3$9(/
6.5,%7629
Pavel is a graduate of
the Moscow Institute of
Physics and Technologies
(MIPT), with a PhD
in Neurocomputer
application for the
representation of static
and dynamic 3D data. He
is also the founder, CEO
and ideological leader of
Pawlin Technologies Ltd.

Toni is a Microsoft
MVP for C#, developer,
speaker, blogger and
technology enthusiast.
With years of
professional experience
working on range of
technologies, his recent
focus has been on
Windows Phone and
Windows 8 as a platform.

*$5<
6+257
Gary is a freelance data
science practitioner
and trainer. He has a
deep understanding
of the fullHadoop and
HDInsight environment,
as well as an interest in
Social Network Analysis,
(UCINet and Pajek)
and computational
linguistics(NLTK).

(2,1
:22'6

0,.(
:22'
Mike is a Technical
Evangelist for Red Gate
Software, on the Cerebrata
Team. He describes
himself as a problemsolving, outdoorsy, user
group founding, dog-loving,
blog writing, solutioncreating, event planning,
married, technologyspeaking, father-of-one.

Eoin is a lead architect in

the Operations Technology
group at UBS. Prior to
UBS, he spent 20 years in
software engineering at
Bull, Sybase, InterTrust
and BGI. His main
interests are software
architecture, distributed
systems, computer security
and data management.

"&GX9GGM | YYYFGXYGGMEQO | 23

DevWeek_2015_Brochure.indd 23

1/5/15 10:53 AM

4')+56'44&
/10&#;
/#4%*
24+%+0)126+105

$11-
019

:;LM;;A(&'+FH?9?D=IJHK9JKH;
$OOSULFHVLQFOXGHUHIUHVKPHQWV
EXIIHWOXQFKDQGVHVVLRQ
QRWHVEXWH[FOXGHWUDYHO
DQGDFFRPPRGDWLRQ

8EEA8O<H?:7O
'/:;9;C8;H(&'*

I7L;KFJEw)&&

8EEA8O<H?:7O
)&@7DK7HO(&'+

I7L;KFJEw(&&

8EEA8O<H?:7O
,C7H9>(&'+

I7L;KFJEw'&&

8EEA7<J;H
<H?:7O
,C7H9>(&'+

KD?L;HI7BF7II
7BB+:7OI

c 8#6

c 8#6

c 8#6

c 8#6

*#:7OF7II
C7?D9ED<;H;D9;!
FH;%FEIJMEHAI>EF

c 8#6

c 8#6

c 8#6

c 8#6

)#:7OF7II
C7?D9ED<;H;D9;EDBO

c 8#6

c 8#6

c 8#6

c 8#6

(#:7OMEHAI>EFF7II
C7?D9ED<;H;D9;
(#:7OMEHAI>EF

c 8#6

 8#6

c 8#6

c 8#6

c 8#6

c 8#6

c 8#6

c 8#6

'#:7OMEHAI>EFF7II
FH;%FEIJC7?D
9ED<;H;D9;MEHAI>EF

$FFRPPRGDWLRQLVQRWLQFOXGHGLQWKHSULFHRIDWWHQGLQJ'HY:HHN,I\RXZRXOGOLNHWRERRNDKRWHOURRP
SOHDVHFKHFNWKHOLVWRIUHFRPPHQGHGKRWHOVRQWKH'HY:HHNVLWHZZZGHYZHHNFRPDERXW

 6 1  4 ' ) + 5 6 ' 4  
8 + 5 + 6  & ' 8 9 ' ' -  % 1 /
)RUIXUWKHULQIRUPDWLRQ

GHYZHHN#EVLFRXN

6'4/5#0&%10&+6+1052NGCUGPQVGVJCVVJGQPNKPGUWDOKUUKQPQHCEQORNGVGFTGIKUVTCVKQPHQTOEQPUVKVWVGUCTODQQMKPIUWDLGEVVQVJGHQNNQYKPIVGTOU
CPFEQPFKVKQPU#P[ECPEGNNCVKQPUTGEGKXGFCHVGT(TKFC[,CPWCT[YKNNKPEWTCCFOKPKUVTCVKQPHGG%CPEGNNCVKQPUOWUVDGOCFGKPYTKVKPICVNGCUV
FC[UDGHQTGVJGEQPHGTGPEGQTVJGHWNNHGGYKNNDGEJCTIGF9GCTGJCRR[VQCEEGRVUWDUVKVWVKQPUKHVJG[CTGUWDOKVVGFKPYTKVKPIDGHQTGVJGEQPHGTGPEGDGIKPU
6JGQTICPKUGTUTGUGTXGVJGTKIJVVQOCMGEJCPIGUVQVJGRTQITCOOGCPFURGCMGTUYKVJQWVPQVKEGKHVJKUKUWPCXQKFCDNG+HFGNGICVGUCTGWPCDNGVQCVVGPFHQT
CP[TGCUQPVJCVKUDG[QPFVJGEQPVTQNQHVJGQTICPKUGTUUWEJCUVTCPURQTVRTQDNGOURGTUQPCNKNNPGUUDGTGCXGOGPVKPENGOGPVYGCVJGTVGTTQTKUOQT#EVQH)QF
KVYKNNPQVDGRQUUKDNGVQOCMGCP[TGHWPFUQHEQPHGTGPEGQTYQTMUJQRHGGU
24 | YYYFGXYGGMEQO |

DevWeek_2015_Brochure.indd 24

"&GX9GGM

1/5/15 10:53 AM