Paul Kedrosky on the under-appreciation of mobile web platforms:

People just don’t get how good web apps on mobile can be. I get stuck in many conversations about native apps on mobile, which always strikes me as vestigial. The same way that I live in mostly browser-based apps on my desktop and laptop, I am convinced I will live in web apps on my mobile device. Pretending otherwise seems wrong-headed and self-defeating.
 

At MobileCamp Vancouver I pushed just this point on the industry panel. I’m not sure the other telco folks on the panel were convinced. But – what I found after the panel among the broader development community there, was intense interest in targeting mobiles through web-based runtimes – be they web sites, widgets, Adobe AIR, Google gears, etc.

There are really two issues to consider here.

1. Are web runtimes capable of delivering the necessary functionality and experience equal to native runtimes? The answer is yes although not everybody gets this yet. See my post on Adobe AIR / Google gears like platform for iPhone for an argument in favor of mobile runtimes over native runtimes.   
2. Recognition of another under-appreciated benefit of mobile web runtimes – they drastically expand the world of people that can develop mobile applications. In this sense web runtimes are poised to play a role similar to that played by Visual Basic for desktop apps in the 90′s. To be clear – I am not comparing the two in technical terms, but in terms of “developer reach”. Undeniably (and to many developers’ deep regret) VB drove a huge developer population.

Adobe’s AIR platform has shown that the web development model is fundamentally capable of delivering applications every bit as rich as native platform apps. Hence the outcry over the idea of web applications as the primary development model for the iPhone is overblown. Yes, the current iPhone development environment sucks but it’s important to keep in mind that we haven’t seen the true power of this model.

So what will Apple do if they are serious about the web development platform?

First up, a JavaScript bridge into native device capabilities on the iPhone. Right now in terms of access to device functions via web technologies, the iPhone offers nothing beyond the microbrowsers on feature phones. Yes, the AJAX support enables a quantum leap in usability, but we need more.

Top of the list – “the basics”

• file access
• camera access
• location
• address book
• calendar

But beyond these now familiar mobile device capabilities, there is an opportunity for Apple to provide powerful APIs that would provide an unparalleled mobile development environment, web or otherwise:

• SQLite local store
• Access to iTunes account-based charging
• Event subscriptions with persistent ECMAScript object handlers
  • timers
  • calendar events
  • location
  • support for installable URI handlers

Key to app usability is that these capabilities should be provided through a user consent system which does not take away the user’s choice to make use of device features with a given application, a la the restrictions in many J2ME runtimes:

On first load, present the list of “privacy affecting” services being requested by the application – and listen to what the user wants! Empower the user already – they bought the damn thing.

When can we expect such features? Who knows, but I would be very surprised if the first GPS-enabled iPhones, expected in Q1 2008, ship without at minimum a JavaScript bridge to location data following soon after.

Gut check: For anyone with development experience on J2ME, can you imagine going back to the underachiever that is the J2ME RMS storage API after being provided with a local SQL store?

The native app vs. web app argument is stale!

Jeff
Chief Software Architect, EQO