In Mobile, It’s A WebKit World And We Just Browse In It. Is That Okay?
A decade ago, one browser platform had crushed competitors into rubble on its way to cementing a 95 percent share of the desktop market. That near-monoculture was unequivocably bad for the Web. Today, another browser platform is consolidating an even more dominant lead in mobile devices… and we’re not all sure if that’s a problem or not.
That former supremacy of Microsoft’s Internet Explorer and its Trident layout engine (the component that turns HTML code into text and media) and today’s dominance of the open-source WebKit engine in Apple’s Safari and Google’s Chrome and Android is not a case of the new boss being the same as the old boss.
The desktop market, even after years of decline by IE, remains fairly evenly split between Microsoft’s software, WebKit browsers and Mozilla Firefox. And it still dwarfs mobile, adding up to nearly 88 percent of all browsing in NetMarketShare’s stats.
But WebKit’s lock on the devices we carry almost all the time raises new issues of its own. Especially since that engine’s most durable competitor, Opera Software, announced two weeks ago that it would dump its own Presto library in favor of WebKit.
Before that move, WebKit held almost 87 percent of the mobile market in NetMarketShare’s research (to be exact, 61.02 percent Safari, 21.46 percent “Android browser,” and 4.33 percent Chrome). Add in Opera’s 9.84 percent–a steep drop from the quarter of the market it held back in March of 2011–and you get 96.65 percent.
It’s hard to think of any software market so one-sided, much less one so dominated by open-source software. (The Apache Web server peaked at nearly 71 percent in late 2005, by Netcraft’s data.)
The argument to welcome our new WebKit overlords starts and stops with those two magic words: “open source.”
WebKit’s GNU Library General Public License–a legacy of the open-source KHTML software that Apple built the first version of Safari on–owes nothing to the all-rights-reserved copyright of IE. WebKit’s terms, like Mozilla’s Public License, invite anybody to download, inspect and improve the code.
As a result, you can find WebKit code on BlackBerry phones and Amazon Kindles–the Silk browser on Kindle Fires owes a great deal to this engine.
(This diversity diminishes in iOS, where Apple’s security restrictions require alternative browsers such as Chrome to employ iOS’s WebKit framework–and where you can’t set another browser as your default anyway.)
So if nobody owns WebKit and everybody can use it on competing platforms, what’s the problem? Shouldn’t the addition of Opera’s engineering tablet make WebKit even better?
Unfortunately, the same lazy habits that led many Web developers to write only for IE a decade ago now invite today’s Web authors to put down their tools once their sites work with WebKit–even though they might need just a few lines of code to enable wider compatibility.
WebKit exclusivity looked bad enough a year ago for Daniel Glazman, co-chair of a key Web-standards working group at the World Wide Web Consortium, to denounce it in bold, all-cap, large type: “THIS MUST NOT HAPPEN.”
Some WebKit supporters have gone a step further by actively shutting out competing browsers. Two years ago, Apple posted an HTML5 demo that blocked even other WebKit-based browsers in favor of its own Safari; earlier this year, Google shut out Windows Phone users from the mobile version of its Google Maps site.
Apple and Google corrected those mistakes. But smaller companies continue to make the understandable business decision of optimizing for the overwhelming majority of today’s mobile users.
I’ve gotten a better sense of that as I’ve been spending a little more time with a loaner phone, HTC’s 8X, that runs Microsoft’s Windows Phone 8. I still hit sites that dumb themselves down by presenting a minimalist, feature-phone interface; the mobile version of baseball’s MLB.com looks awfully close to the one I had I bookmarked on my Palm Centro five years ago. Others make the opposite error, dumping me into a full-sized version that’s barely usable on such a small screen.
This isn’t about making room for Microsoft in a two-player smartphone market that may not want Redmond’s help. Other companies might have good ideas of their own: Today, I expect to see the first phone shipping with Mozilla’s Firefox OS at the Mobile World Congress trade show in Barcelona.
Mozilla supporters have felt particularly put upon since Opera’s move and have taken to blogs to contest the wisdom of WebKit everywhere.
Developer Steve Fink decried the notion of settling on a monoculture when the Web evolves so fast, sketching out ways independent developers of WebKit browsers could fall perpetually behind Apple and Google. Chief technology officer Brendan Eich argued that the Web needs a platform from a source with an explicitly non-commercial focus–as in, his own non-profit employer.
But you can find one of the most concise cases for Web-engine diversity from Opera itself–written two years ago. The author of this month’s WebKit announcement argued on the same blog that Google should fork WebKit into a separate Chrome architecture: “A more diverse browser market would also force more Web developers to adhere to standards because it would no longer pay off to optimize for just one or two browsers.”