Google has today announced Google gears. It allows web applications to work offline. Google gears is a browser extension that can be downloaded and installed. It exposes Javascript classes which can be invoked by web applications to provide offline functionality.
While it appears cool at first look, it needs to be analyzed in detail. Here are some initial questions that needs to be answered:
1. Does the user have control over deciding which applications to make offline? How much data should be available offline?
2. Does the offline functionality run in-process (in the browser)? If it runs in-process is there a chance of browser bloating up in memory due to offline caching or run slow?
3. What about security? Is it possible for a rogue application to cache and use background threads to connect to their site just because the user visited the website and the Javascript accessing Google gears got loaded?
4. Does Google gear intercept all HTTP requests going out of the browser? If I am offline and just type google.com/mail how does it capture this request and present me with offline content?
5. It appears the offline functionality is going to work only for pure Javascript applications. How does it work when application has other tiers like Java, PHP, Ruby….?
6. Writing applications to handle connection status and gracefully transition is going to be a programming challenge (nightmare).
Interesting to note Adobe Apollo support for Google gears (or viceversa). “Kevin Lynch, senior vice president and chief software architect at Adobe, said his company will join Google in the effort to develop a standard cross-platform, cross-browser local storage capability. The Gears API will be available in Adobe’s Apollo tool that enables Web applications to run on the desktop, he added.”. I thought they did not accept this – Adobe Apollo – On A Collision Course With Web Browsers :)
Technorati Tags: Google gears, offline, Apollo