It’s been a little quiet around here lately, but as Dylan’s last post points out, it’s because we’ve been so busy on working on the technologies we need to push the web forward. While Comet, cross-domain services, and native vector graphics are nearly “here”, “Mobile Ajax” is clearly a term in search of a definition.

I’m going to spend a little time recapping my recent excursions into the world of mobile web app development to share a little of what we’ve learned and where SitePen will be focusing its resources with respect to mobile web apps.

Recently, we’ve been hearing more and more about “Mobile Ajax”, mostly in vague terms. I suspect this stems mostly from no one knowing really what it means. Those who are clued into the size of the mobile phone market are, like me, scratching their heads as to why mobile web apps still suck. Mobile phone browsers have traditionally been under-powered to the extent that they can’t really support applications, but Moore’s Law solves that problem eventually. Coupled with significantly improved bandwidth to phones, you’d think that we’d be in a golden area of invention around mobile web apps. It hasn’t happened yet. Significant impedance mismatches between web app developers and the constraints of mobile phones still remain, and despite the high-end of the phone market being “there”, the platform as a whole isn’t. Here’s why:

  • Latency
  • Developer (mis)training
  • Browser issues
  • Carrier slow-wittedness and delusions of competence

All of these factors have contributed to a situation in which web apps on phones, nevermind ajax apps, generally suck. And it’s not for lack of trying. Using google or yahoo on a mobile device today is generally about as good as it gets, and they’re not bad. Unfortunately the text bias that’s built into the fabric of the “desktop web” hurts even these sites which have entire teams dedicated to figuring out how to reduce steps to action.

All of this matters to us at SitePen because it’s our goal to help push the web forward (and this case, outward). Through talks like the one I gave at EuroOSCON (and will give again at AjaxWorld) we’re starting to share the raw data that’s been available but not yet widely distributed. At the end of the day, a couple of things have become pretty obvious out of our research:

  • Ajax on mobile devices isn’t worth doing yet
  • Mobile browsers need to evolve before apps can really take off
  • Developers need to stop expecting to get text from clients
  • Developers need to start thinking about “stacking” interfaces instead of scrolling
  • Most big web properties are already “on it”, leaving smaller firms at a disadvantage in both experience and motivation

All of this generally means that the term “Mobile Ajax” is likely to be a code word for “snake oil” for the foreseeable future. Ajax isn’t the right answer for mobile devices. Other techniques will evolve to improve the utility web apps for mobile devices, and here at SitePen we’re already working to test out some of our hunches about how we can solve the existing UI problems.

If you’d like to learn more about what we’re working on in this area and how our expertise and perspective might improve your mobile apps, drop us a line.