The Modern Web: Multi-Device Web Development with HTML5, CSS3, and JavaScript

The Modern Web

Multi-Device Web Development with HTML5, CSS3, and JavaScript

Peter Gasston

Published by No Starch Press

Advance Praise for The Modern Web

This is a useful book, and it’s an important book. If you follow Peter Gasston’s advice, then test your sites across all browsers and on a variety of devices, you’ll impress your bosses and please your users. You’ll also be making the Web better and keeping it open.

—BRUCE LAWSON, AUTHOR OF INTRODUCING HTML5

Peter Gasston has now done for the modern web platform what he already did for CSS: write a consult-it-first compendium of information for web developers of practically any skill level.

—STEPHEN HAY, AUTHOR OF RESPONSIVE DESIGN WORKFLOW

Peter Gasston strikes a great balance between producing fantastic real-world code and staying right on top of the latest developments in web technology. He has a considerable gift for explaining difficult technical topics in a lucid and entertaining manner.

—CHRIS MILLS, DEVELOPER RELATIONS MANAGER, OPERA SOFTWARE AND AUTHOR OF PRACTICAL CSS3

Praise for Peter Gasston’s The Book of CSS3

"I can honestly say I will never need another book on this subject, and I doubt anyone else will either. The Book of CSS3 covers it all and covers it well."

—DEVON YOUNG, CSS3.INFO

One of the best technology books I’ve read.

—CRAIG BUCKLER, OPTIMALWORKS WEB DESIGN

This book deserves a place within easy reach of the developer’s keyboard and is a must have for anyone looking to join the visual revolution that CSS3 is bringing to the Web.

—C.W. GROTOPHORST, CHOICE MAGAZINE

There are a lot of neat things that you can do in CSS3, and this book is a great introduction to these features.

—STEVEN MANDEL, .NET DEVELOPER’S JOURNAL

An easy-to-read, easy-to-implement handbook of the newest additions to the Cascading Style Sheet specification.

—MIKE RILEY, DR. DOBB’S JOURNAL

For Dave, Jim, Morena, Nick, Rupert, Steve, and all of the other organizers of the London Web Standards group, who help to keep the London scene active and gave me my first opportunity in public speaking.

About the Author

Peter Gasston has been a web developer for over 12 years in both agency and corporate settings. He was one of the original contributors to CSS3.info, the leading online destination for CSS3. Gasston is the author of The Book of CSS3 (No Starch Press) and has been published in Smashing Magazine, A List Apart, and .net magazine. He gives talks about technologies at developer conferences and runs the web development blog Broken Links (http://broken-links.com/). Gasston lives in London, England.

About the Technical Reviewer

David Storey is an HTML5 evangelist at Plain Concepts, a founding member of the IE userAgents program, and a CSS Working Group member. Prior to this, he was the developer advocate manager on a top-secret skunk works project at Motorola. He also founded the developer relations team at Opera, product managed Opera Dragonfly, and worked at CERN, home of the World Wide Web. His passion is keeping the Web open for all.

Acknowledgments

Huge thanks are due to David Storey, the technical reviewer for this book; his deep knowledge of the field meant his feedback was invaluable to me. Although he corrected and guided me many times, if there are any errors in this book they’re entirely my responsibility.

Keith Fancher, Serena Yang, Bill Pollock, and the rest of the No Starch Press team provided fantastic support and guidance throughout the writing of this book. Their collective contribution is beyond measure.

Dimitri Glazkov helpfully answered a few questions on web components, and Bruce Lawson gave extra feedback on the new HTML5 elements in Chapter 2. His work as well as that of his fellow HTML5 Doctors was a constant reference during the writing of this book.

Although I’ve never met him, I’d like to thank David Walsh for maintaining an excellent website that I have used a lot.

Stephen Hay and Chris Mills have been generally useful in helping me to consolidate ideas, as well as incredibly nice people to know.

Great thanks to my friends and occasional colleagues Giles McCartney, Richard Locke, and Tom Shirley. Thanks also to all my other colleagues at Preloaded, Poke, Top10, Believe.in, and rehabstudio.

As always the biggest thanks must go to my wife, Ana, for her patience and support during the time I spent writing this book.

Introduction

We are in a time of unprecedented innovation on the Web. Not too long ago, one company, Microsoft, dominated the web landscape; in 2003, Internet Explorer was used on some 95 percent of computers worldwide. This domination brought the advantage of a stable market for developers, but there was also a serious drawback: Microsoft chose to end nonessential work on IE, and innovation on the Web stagnated—a consequence of a lack of competition and a closed environment.

Things could not be more different now. There are some four or five key browser vendors, about the same number of major operating systems, and more parties are getting involved all the time. Adobe has switched its focus from Flash and apps to the open web, and technology companies like Samsung and Nintendo are joining key players such as Google, Apple, and Microsoft in shaping the future of the Web and the way we build for it.

And the way we access the Web has changed enormously too—think devices. No longer are we limited to browsing the Web with a desktop or laptop. So many categories of devices exist now—including smartphones, tablets, and games consoles—that I’m running out of fingers to count them on.

This book is about front-end web development in this new web-everywhere era. It’s about learning methods to make first-class websites, apps, or anything built on open web technologies, with the multi-device world aforethought. This is not a book about how to make mobile websites or smart TV apps; it’s about learning the latest developments in current and near-future web technologies so you’ll be better able to build sites capable of offering the best experience everywhere.

I’ll return to what you’ll learn from this book in more detail at the end of the introduction, but first I want to talk about the bewildering array of today’s web-enabled devices.

The Device Landscape

The year 2008 was a landmark year and not only because of the theatrical release of Indiana Jones and the Kingdom of the Crystal Skull, which introduced the phrase nuking the fridge into our vernacular. This was the year that saw the number of Internet-connected devices exceed the number of people using them. This super-connectivity between devices is commonly known as the Internet of Things (IoT), and this book is aimed at an already substantial and fast-growing subset of those things, namely things with web browsers.

The range of web-enabled devices is enormous and getting broader by the day. As I write this, I’m the owner of the following devices that have a web browser: a desktop, a laptop, a tablet, a smartphone, an ebook reader, and a games console. I suspect that’s not an uncommon scenario to more affluent members of society; many people could add a smart TV and a portable games console to that list.

Of course, with this incredible range of web-enabled devices comes incredible opportunity for web developers, and this book is here to help you make the most of that. But first, let’s look at some of the core device categories and define some key terms so you’ll better understand what you’re dealing with.

Desktop/Laptop

The Web as we’ve known it until recently has primarily targeted the larger screens and more powerful processors of desktop or laptop computers (shorthand: desktops), where the operator is usually seated and using a fast Ethernet or Wi-Fi broadband connection. This mode is still the default for people at work in offices, so many business-to-business (B2B) sites are built to suit this configuration.

But desktop computer use is on the wane as many people nowadays, especially home users, tend toward mobile or tablet devices (which I’ll come to shortly). In the last quarter of 2012, global PC shipments were down almost 5 percent from the previous year. That said, research shows that people still tend to use desktops when performing tasks that involve significant amounts of text, multitasking across different tabs of a browser, or using programs that require close control of the content, such as image editing. Desktops are also used when security is a major concern, such as for Internet banking.

But even within this venerable and mostly stable group variation abounds. The 1024×768 screen resolution is slowly dying away but still prominent, while newer versions of Apple’s MacBook Pro sport a resolution of 2880×1800. Sites optimized for the former will look quite small and be somewhat lost on the latter.

The desktop browser world includes five major players: Chrome, Safari, and Opera, which share the WebKit engine (Opera had its own engine, Presto, which is now being phased out); Firefox; and Internet Explorer (version 9 can be considered semimodern and 10 fully modern). Thankfully they all tend to implement features in a standard way, and they all have a frequent or semifrequent update cycle (or are moving toward it), so users tend to get new features fairly quickly. Of course, a sizable percentage of users are still running older, less capable browsers, so you should always build with that in mind. (Much of the innovation in web standards comes to desktop browsers first because they’re the ones that developers tend to use on a daily basis. Therefore much of this book is written with desktop browsers in mind, although what you’ll learn can be applied anywhere.)

Mobile

When I say mobile, I generally mean phones. The range is wide: from older, cheaper hardware running rudimentary web browsers to mid-range feature phones with browsers optimized for speed over power to high-powered expensive smartphones with browsers on par with those on your main computer, as well as a series of apps that often use an embedded browser to view content.

In many countries, a majority of Internet users are on a mobile device; in India, for example, some 55 percent of web visits are made on mobile devices, and in Egypt nearly 70 percent of people rely on phones as their sole access to the Web. Worldwide estimates are that the number of mobile web users will exceed that of desktop computer users sometime between 2014 and 2015.

Across the developing world, the dominant mobile device is the feature phone, which has functions beyond basic phoning and texting but often lacks full web access. The rise of low-cost and secondhand smartphone sales, however, is already changing that market drastically. Within the next few years, we can expect to see smartphones dominate the mobile landscape. In 2011, smartphone sales were estimated to account for 27 percent of the global market; at the end of 2012, estimates were that 1 billion smartphones were in use, with that number predicted to double by 2015.

The sheer variety of mobiles and mobile browsers in the global market means that I’m selective about the coverage in this book, so for future-proofing I target phones with fully capable web browsers, that is, features on par with their desktop equivalents.

The WebKit engine dominates the smartphone browser market because it’s used on iPhone, Android, Blackberry, and more, powering many different browsers. Each OS uses a slightly different version of WebKit, but general interoperability is good.

Windows Phone OS from version 7.5 and up uses a mostly desktop-equivalent Internet Explorer as its browser; older models have IE9, and more recent models, IE10. Firefox is also available as an option for Android, and Firefox OS, a full operating system based around the browser, is set to launch in 2013. That launch may well impact the market in the long term.

Opera has a significant share in the global mobile browser market thanks to Opera Mini, a proxy browser that compresses requested pages and returns the data as a kind of image with very limited interactivity. Opera Mini is lightweight and fast, which makes it popular in countries with limited Internet connectivity, but its lack of full interactivity means I won’t give it much attention. As I write this, Opera has announced that it’ll be releasing new mobile browsers based on WebKit, although no details are available.

Today’s smartphones, although getting more powerful all the time, tend to have less available memory and storage, and lower potential connection speeds, than desktop or laptop computers. On the surface, this would seem to make them the poor cousin of web-enabled devices, but as Jonathan Stark says in The 10 Principles of Mobile Interface Design:

Smartphones are actually more powerful than desktops in many ways. They are highly personal, always on, always with us, usually connected, and directly addressable. Plus, they are crawling with powerful sensors that can detect location, movement, acceleration, orientation, proximity, environmental conditions, and more.

Obviously the advantage of mobile is just that: its mobility—the ability to find things around you, get directions, and look up information fast when out of the house or office. But increasingly, mobile devices are being used at home while watching TV or on the morning commute or when bored standing in line. Mobile is really less about being on the move and more about always being available.

The mobile space is changing faster than any sector of the market, as many users change devices on 12- to 18-month update cycles and hundreds (if not thousands) of new models are introduced every year.

Tablet

Tablet computers have been around for years, but it wasn’t until Apple came along with the iPad in 2010 that they became more of a leisure item than a desktop accessory. Many other devices have followed the iPad, with Android-based tablets doing especially well in the mid-sized 7" range and Microsoft pushing forward with the heavily tablet-focused Windows 8. From a market that was negligible at best in 2009, estimates are some 390 million tablets will be in use worldwide by 2015. That’s an incredible figure, and the growth is exponential.

Their larger size and reliance on Wi-Fi (in many cases) makes tablets portable rather than truly mobile; they exist in a space more akin to a laptop than a mobile device. Most people use them at home, and although they are carried around, they’re usually taken out and used only when the user is stationary, rather than walking down the street. That said, it’s not uncommon to see them used as somewhat awkward and ungainly cameras in public places!

Like smartphones, tablets are tactile, relying on touch input in most cases (although some also accept stylus input). Research shows they tend to be used for entertainment and browsing when time is less of an issue. The core browsers on tablets are essentially the same as those on mobile devices.

TV

The TV is still the most-used screen in the home, though mostly for passive viewing rather than interactivity. That’s changing as the new breed of web (or smart) TVs and media boxes begin to get a foot in the door. Although global figures are hard to come by, predictions are that there will be 100 million web-enabled TVs in Europe and North America by 2016.

Although many web TVs use applications rather than websites, they’re often built using web technologies and sometimes contain an embedded browser view. After all, if you’re making a Twitter app, you want your users to be able to open links.

The biggest drawback with web TV is that navigating with a TV remote is a horrible experience, made worse by some of the multibutton monstrosities that have been built to provide the level of interactivity required for Internet use—notably, a keyboard. Some manufacturers are making gesture- and voice-controlled TVs or remote controls that incorporate a trackpad, but the best solution seems to be to pair a mobile or tablet with the TV and use that as an interactive control.

Another obstacle to web browsing on the TV is that the television is inherently a communal device, viewed by many people at the same time—an experience that is fundamentally opposite to the personal Internet experience. Would you like to have your Facebook account broadcast on a big screen for everyone to see? Probably not.

The emerging common behavior is for web TVs to be used for video and interactive services, with a real-time social aspect happening concurrently on the user’s mobile or tablet. Recent research shows that 77 percent of people use another screen while watching (if that’s the right word) TV, and of that figure, only 5 percent or so are performing an activity that’s relevant or complementary to what’s on screen.

The Others

These four broad categories—desktop/laptop, mobile, tablet, and TV—cover the majority of the web-enabled device market, but they’re by no means exclusive. Many other devices have web browsers, even if they aren’t always used with great frequency.

Each of the current generation of home games consoles has a browser: The PlayStation 3 uses the WebKit-based NetFront, Microsoft’s Xbox 360 has Internet Explorer 9, and Nintendo’s Wii U has a custom build of WebKit. Likewise, portable games consoles have browsers: The PlayStation Vita uses NetFront, as do newer versions of the Nintendo 3DS (older versions use Opera). Bear in mind that each uses a slightly different control system. (For more on the topic, see Anna Debenham’s amazing work in the field as listed in Appendix A.)

And let’s not forget ebook readers, such as Amazon’s Kindle, the Kobo, Barnes & Noble’s Nook, and a series from Sony. (When I use the term ebook readers, I’m referring specifically to e-ink readers, as each company also has a full-color screen version that is more like a tablet.) The ebook browsers tend to be WebKit based, and their challenges come from low-powered processors, little memory, and very slow refresh rates. They wouldn’t be the first-choice browser of many, but as ebooks become more interactive, they may become more popular.

The In Betweeners

The iPhone 4 has a screen size of 3.5 diagonally, and the iPhone 5, 4. Samsung’s Galaxy S III has a diagonal length of 4.8; LG’s Optimus Vu, 5; and Samsung’s Galaxy Note II, 5.5. Google’s Nexus 7 is 7 diagonally, and Amazon’s Kindle Fire HD comes in 7 and 8.9 formats. The iPad mini is 7.9 diagonally and the iPad, 9.7.

In other words, the great variety of screen sizes makes the distinction between mobile and tablet quite hard to pin down. The only criteria I can think of to separate them are based on whether they have native telecom ability, can fit comfortably in a pocket, and can be used comfortably with one or two hands.

In fact, computing is experiencing a general convergence. Microsoft’s Surface is a tablet with an optional keyboard in the cover. When keyboard and tablet are attached, the Surface looks and behaves like a laptop. The Ubuntu Phone runs Android, except that when you dock it to a screen, it runs a full version of the Linux desktop. The PadFone by Asus is a phone that turns into a tablet when docked with a larger touchscreen. Your TV becomes a media center when you plug Apple TV or Roku into it.

This change is set to outdate all of our existing terminology. In the foreseeable future, the word tablet will be as meaningful as the floppy disk icon currently used to mean save in many software applications. One day, everything will be screens.

The Multi-screen World

The idea of a user being a mobile user or tablet user is somewhat misguided because people are much more likely to use many devices, often at the same time. Research conducted for Google in 2012 indicated that 81 percent of participants used their smartphone while watching TV, and 66 percent had it available while using a desktop. This use of multiple devices concurrently is known as simultaneous screening.

Meanwhile, 90 percent of participants started a task on one device and finished it on another; for example, they browsed shopping websites on a smartphone and then moved to a laptop at the time of purchase. This movement between consecutive devices depending on the task being performed is labeled subsequent screening.

In other words, visitors to your website will visit from two or three different devices, and although they’ll have slightly different aims each time, they’ll still want access to the same information.

Context: What We Don’t Know

The most important thing to bear in mind is how much we don’t know about the people using our sites. We don’t know where they are—I don’t mean geographically, as we can use GPS to locate them. I mean whether they’re at home, at work, on the bus, and so on. In essence, we have no idea of the context in which someone is using our site.

The common presumption is that people using mobile are on the move and in a hurry, often using low-bandwidth 3G connections, but that’s not always the case: Mobile is often used at home with a good fast Wi-Fi connection. But even the connection doesn’t tell the whole story; the signal may be poor or the bandwidth reduced because of congestion in the neighborhood. In fact, the 3G (or 4G in many countries) networks might actually provide a much better connection in many cases.

We also presume that mobiles are less powerful than desktops, but even that may not be true for much longer. Twin- or even quad-core phones with fast processors are making their way to the market, and within a year or two, who knows how much power they’ll have. And tablet/laptop hybrids like many Windows 8 devices are now more capable than laptops just a few years old.

There’s also the issue of size. The presumption is that for a large-screen device like a TV, viewers want a richer graphical environment, but a large screen doesn’t equal a powerful processor or a fast connection. Many web-enabled TVs have processors no more powerful than a smartphone, and the connection speed is open to the same vagaries as any other device. And with size comes the question of portability. The more portable the device, the less certain we can be of the context in which it’s being used.

Really, the cardinal rule is this: We cannot make presumptions. And having said that, I’m going to break my own rule.

Some Context Stereotypes

To avoid repetition in this book, I rely on a few shorthand contexts based on common scenarios. These won’t necessarily be the most common scenarios, but ones that I think are common enough to serve a useful purpose. One example: Mobile users don’t have a super-fast Internet connection. Often mobile users are using their devices from home with a mega-fast fiber-optic connection broadcast over clear Wi-Fi from 2 feet away, but the opposite is often just as true: They’re away from home and relying on a very weak 3G signal (as happens to me too often). Many smartphones are built with scenarios like this in mind; they limit the number of connections that can be made at any time in order to not gobble precious data from the user’s limited tariff.

Likewise, a user with a desktop computer will likely have a direct broadband connection, providing fast data transfer rates. That’s not always true, of course—many people in rural areas have extremely low broadband speeds or still use dial-up—but the first scenario is common enough that I can use it as a shorthand.

I use shorthands like these throughout the book simply to avoid constant clarification and repetition, but I can’t drum home enough the idea that these presumptions can’t, and shouldn’t, be foremost in your mind when building and planning websites or applications.

Fast Is the Only Context That Matters

You have a challenge. You don’t know who your users are, where they are, what they are doing, or which device they are doing it with. You can find out some (although not all) of that information, but their full context is completely unknowable and varies for each individual. The only thing you can reasonably know for certain is that either they want access to what you’re offering or they want to find out it’s not what they want. Either way, they want the answer quickly.

Performance is the only criterion that matters. Whether users are on a smartphone during rush hour and looking for information about the next train home or browsing through a shopping site while curled up on the sofa at home, they have a task that they want to complete as soon as possible, and completing this task using the nearest device will make them feel more efficient (this is known as found time).

Your site needs to be fast—and feel fast—regardless of the device it’s being displayed on. And fast means not only technical performance (which is incredibly important) but also the responsiveness of the interface and how easily users can navigate the site and find what they need to complete the task that brought them to you in the first place.

What You’ll Learn

I’ll be up-front and start by