Analysis: HTML5 on Mobile

Last week Peggy Albright on FierceDeveloper.com wrote an interesting analysis of mobile developer adoption of HTML5.

Peggy's audience is mobile developers at large, so she spends a bit of time introducing/evangelizing HTML5 to that superset. But if you're here in our HTML5 Zone, you probably already have some HTML5 chops, or maybe even some HTML5 mobile chops -- so this more focused group can zoom in on some of the more interesting details.

First, the numbers:

But 75% also said they will continue to support native apps.

In general the numbers indicate growing developer interest in HTML5, but considerable suspicion against the strong form of HTML5's 'write once, run anywhere' promise. What these numbers don't say, however, is whether HTML5 will gain more mobile ground if APIs for device-specific features continue to multiply -- or even whether open APIs directly from vendors, like AT&T's, will strengthen JavaScript on the telecom, rather than mobile hardware, side.

In fact, it's generally pretty hard for a developer to understand fully the complex network linking telecom giants, hardware manufacturers, OS vendors, standards bodies, and straight-up web developers, at both enterprise and hacked-together-a-cool-app levels. mobilehtml5.org will get you the quick-and-dirty facts on HTML5-on-mobile compatibility, but won't too much help you forecast the trends that you'll need to know to avoid tying your apps to a (who knows?) sinking technological ship.

My personal (perhaps naive) guess is that mobile device manufacturers have pretty much figured out what basic hardware is useful (camera, mic, vibration, accelerometer, notifications), at least in the short term (projectors and 3d gesture-recognition possibly excepted) -- which means that, one the JavaScript APIs give HTML5 apps access to all these existing features, native will have hardly any advantage beyond performance. (This is all leaving aside PhoneGap, of course -- which does a great job filling the space between JavaScript and native, but has some limitations, and won't be necessary if/when mobile browsers support extensive and standardized device APIs.

Peggy's full article is definitely worth reading, even if only as a solidification of trends you already suspected. But what do you think? how invested are you in the idea of the browser as a platform for mobile? and even if so, do you have any ideas for new APIs that might even more successfully bridge the gap between HTML5 and native?