Google has partly scrapped a browser plug-in project called O3D, instead throwing its full weight behind a 3D Web graphics technology called WebGL that got its start at Mozilla.
The move,, has the potential to simplify the effort to bring hardware-accelerated 3D graphics to the Web, an idea that has appeal to those trying to refashion it as a foundation for applications such as games. However, it also means the functioning--if experimental--O3D technology is going back to the drawing board for a while.
The overall idea of O3D, a higher-level interface than the 3D nuts and bolts provided by WebGL, will live on, though. Google is rebuilding it as a library of pre-build software others can use on top of a WebGL foundation, Engineering Director Matt Papakipos and programmer Vangelis Kokkevis announced the move on the final O3D blog post Friday.
Another tricky issue is that WebGL is a variant of the 3D graphics interface called OpenGL, which is used on Mac OS X and Linux as well as the iPhone and Android phones. On Windows, though, it's second fiddle to Microsoft's Direct3D interface. Google hopes to bridge that gap with its.
Said Papakipos and Kokkevis:
Asked in an interview this week about Microsoft's WebGL stance,, "I think it's different markup," meaning something not universally supported on browsers. "You're telling developers, 'Go write something else.'"
Google's answer:, a plug-in that sneaks a copy of Chrome into Internet Explorer, an idea .
Said Henry Bridge, another Google programmer involved in the effort:
Canceling O3D also ruffled some feathers.
As for IE, we could have gone down the route of making the O3D plug-in also ship WebGL for IE, but we realized that then developers would be lacking other APIs [application programming interfaces] that they need for their apps, like Audio or WebSockets. If we then added those APIs to the O3D plug-in also, well, it starts to look a lot like ChromeFrame: a plug-in implementation of Web APIs for IE. We're confident that ChromeFrame will be as easy if not easier to install than O3D, so it seemed better to focus the team's graphics expertise on making WebGL great in Chrome/ChromeFrame and on making the O3D library useful to developers.