Sprint Building a Mobile Webapp Marketplace

Dan Hesse Keynote - Sprint Developers Conference

At the Sprint Developer Conference today, Sprint CEO Dan Hesse (image above) introduced what the operator is calling the “Browser-VAS” ecosystem. It’s supposed to let developers create and monetize browser based apps. Browser-VAS is built around Openwave‘s Integra cloud service.  Hopefully, the service will have a more user friendly name when it launches in the first half of 2011.

The apps appear in a toolbar that users can optionally add to the browser. They can then “install” free and paid apps from a “Storefront” that’s built into the toolbar. The toolbar is context aware. For example, it will only display a language translation app on pages that aren’t in the user’s native language and data intensive apps can be hidden when the user is roaming.

Browser-VAS apps are created using an Eclipse plug-in. JavaScript is used for both application logic and the presentation layer, there doesn’t seem to be any way to add HTML or CSS to an app using the plug in. Developers write the relevance rules, which define when an app appears in the toolbar,  in Java.

Openwave’s Integra web service sits in the data stream between the end user and the 3rd party web apps. It evaluates the relevance rules defined by the app against user data supplied by the network such as the user’s unique id, location, rate plan and roaming status. The user data is not shared with the apps and user must ot-in to the program and can define which sites the toolbar appears on.

The service is designed to be cross platform and cross browser but initially will only support the HTC Epic and EVO.

Browser-VAS will go live on the Sprint network in the first half of 2011. The Eclipse plug-in and a sandbox testing environment will be available soon to developers who register at success.openwave.com/developer-app.html

Browser-VAS seems like a promising idea. I’m all for anything that improves the discoverability and monetization of web apps. Building new apps using it seems to be pretty straight forward. Porting pure JavaScript web apps like bookmarklets and games written completely in JavaScript should be also trivial.

The biggest drawbacks to Browser-VAS currently is that it’s a single operator solution limited to a few devices.   I’m also not a fan of creating web apps entirely in JavaScript. Any HTML or CSS needed has to be either dynamically loaded with JavaScript or written to the DOM by JavaScript code, which means that most current web apps will require considerable modification.

Mobile apps are slowly but surely moving away from local native code to the web. But we aren’t there yet. To get there, three things are needed. First, browsers need to get better, particularly in JavaScript performance,  support for offline storage and access to core device features like location, contacts, calendar, messaging and the speaker and microphone. Second we need better ways for users to find the webapps they need. Finally, there need to  be better ways to make money on the mobile web. Browser-VAS is a first step toward solving the discoverability and monetization problems. There will be others, including Google and Mozilla who are working on web app stores, but Sprint and Openwave, along with others like OpenAppMkt are the pioneers in the mobile space.