May App Store Report: HTML5 App Distribution

The May edition of the WIP App Store Report is now available. The App Store Report is a PDF version of our App Store Catalog, the most comprehensive online list of mobile app stores. It also features some additional analysis and commentary, reproduced here. This month, we take a look at HTML5 apps and some of the issues surrounding their development and distribution.

There is a growing swell of interest and support — and hype — behind HTML5 for mobile developers. This is largely driven by three factors:

  • Technical additions to the HTML standard that increase its functionality closer to that of native applications. In particular, the ability for offline data storage and access makes it possible to create a much more “app-like” experience than previously possible.
  • The relative ease of development with HTML and other web technologies vs. native app development.
  • The assumption that because it’s HTML, and fragmentation isn’t a particularly nasty issue on the desktop web, that it will ease cross-platform development and reduce fragmentation.

But the uptake of HTML5 for use in applications is hampered more by commercial concerns that technical ones, especially the idea that because HTML5 apps are “web apps,” the real triumph of the app economy — that is, app stores as an easy and successful distribution mechanism — won’t exist to drive downloads, usage and revenues. This, however, is not the case, with a wide range of app stores supporting the distribution of HTML5 apps in a number of different ways.

First, there are HTML5-specific app stores emerging, such as Zeewe, OpenAppMkt and the Premier App Shop. OpenAppMkt has gained some publicity as an alternative to iTunes for iOS users; discovery, download and payment are handled through a web browser, keeping the store within Apple’s guidelines, and apps are then saved to a device’s home screen as an icon.

Many stores support HTML5 apps through the W3C Widget or other standards with minor (or not so minor) differences. For instance, the WAC widget specs are converging with the W3C standard, while Nokia’s WRT implementation is slightly different. In other cases, such as with BlackBerry, a standard app needs to be put into a native app “wrapper” before it can be distributed.

This idea of a native wrapper is also used by some cross-platform tools, such as PhoneGap, to allow developers to build cross-platform apps using HTML5 and other web technologies, rather than porting native code. PhoneGap also adds the additional ability to take advantage of some device APIs and functionality, and outputs what appears to be a native application that can be distributed through normal means. Appcelerator’s Titanium has a similar output, but does something more akin to porting a web app to native code.

HTML5 certainly holds a lot of promise for mobile app developers while the gap between native and web development is growing ever smaller, and technical issues will continue to evolve and work themselves out. But the commercial concern for developers remains — will HTML5 apps mean a return to the pre-app store days, with difficult discovery and a tiny download and sales market?

The answer, of course, is a resounding no — such an outcome would not be in anybody’s interest. But two issues remain troublesome: the lack of clarity around the topic, and the morass of differing standards, wrappers and other mechanisms required for the wide (and cross-platform/operator/store) distribution of mobile web apps. Until web apps are able to offer an improvement, both commercially and technically, to native apps, their growth may remain stunted.