Open Standards for Real-Time 3D Communication

I first heard about VRML back in 1996 from a local (Fargo, ND) web-developer who described it to me as “HTML for 3D”. While I’m sure he got this from someone who described VRML to him, the main point that came across to me was that VRML was as easy to use as HTML, and it would one day become as prevalent. Possibly, one day, a content author could place both HTML and VRML text within the same text document. Skipping a head ten years we still wait for Web3D technology to equal HTML as a means for conveying information over the web, and we still need seperate plug-ins to display our X3D content. Its not as easy as the 1996 hype first made it out to be.

Recently, however, two major HTML Browser manufactures, Opera and Mozilla, have begun working together under the banner of the WHAT Working Group (http://www.whatwg.org/) to develop 3D rendering capabilities for the <canvas> tag. The <canvas> tag was first introduced by Apple in the Safari 1.3 HTML browser for the purpose of dynamically rendering 2d vector graphics.

3D capabilities of the <canvas> tag are a very recent development, and not much is known about the plan for implementation. However, such a tag brings us back to one of the original promises of web-based 3D, that is, it could be handled as easily as HTML.

I first learned of the 3D plan for the <canvas> tag through Kurt Cagle’s recent blog recap of the 2006 AJAXWorld Conference. Kurt specifically covered Håkon Lie’s (CTO Opera) presentation Why Browsers (still) matter, and its here where the OpenGL accelerated 3D canvas is mentioned.

I was Intrigued by the possibilities of a 3D <canvas> so I decided to contact Mr. Hakon Lie to see if I could find out more about it, and he graciously agreed to provide some answers. What follows are his responses to my questions:

Interview
Aaron: Where is the pressure to add 3D functionality to the tag originating from? What is the source of the demand for the functionality that a 3D ui could provide?

Mr. Håkon Lie: This feature is driven by Opera developers who see exciting opportunities when the browser is coupled with a 3D engine. As you know, many systems ship with 3D engines and they even start appearing in mobile devices. Given that the engines are there, is seems natural to try using them for web content.

Aaron: What are Opera’s goals for 3D functionality in the tag? How are the members of the WHAT Working Group working together meet common goals for the tag?

Mr. Håkon Lie: Opera isn’t alone playing with 3D canvas, Mozilla does, too. We collaborate with Mozilla and Safari developers in WHAT WG and W3C, and it’s natural for us to get together and specify the functionality for 3D canvas as well. We’re still in brainstorming stage, though, and it will probably take a while until we have a specification for 3D canvas. I expect the 3D community to contribute to the specification as well.

Aaron: How will the proposed technology compare to other web-based 3D technologies? In other words, is it expected to be a competitor to such technologies as X3D and KML; or is the tag expected to be compatible with these technologies? If they can co-exist, what would be the nature of their relationship be to each other?

Mr. Håkon Lie: These are good questions. Increasingly, JavaScript can be used to support new data formats on the web. This means that the browser only needs to provide a basic API—support for various other formats can be added on top. So, I think it’s too early to say which formats will survive and how they will relate to each other.

Aaron: How is the 3D functionality of the tag expected to add to the Opera user experience?

Mr. Håkon Lie: Our users are often technically advanced and they love playing with new toys. 3D canvas is a great toy, and I expect that—for example—many of the Opera widgets will take advantage of 3D capabilities to build better interfaces and to visualize data in new ways.

Conclusion
3D on the web is once again becoming mainstream for developers, and we may soon see support for interactive 3D content available within HTML Browsers without the need for any plug-ins. The question is: Should the Web3D community embrace the growing efforts of the HTML Browser developers? Opera expects the 3D community to contribute to the Canvas:3D spec. Will the Web3D community see the <canvas> tag be a competitor to X3D, or will a relationship of compatibility develop between these technologies?

I personally applaud the HTML Browser manufacturers and their attempts to bring 3D rendering to the web surfer. But, what about Internet Explorer? Microsoft doesn’t play with WHATWG, so is the tag dead in the water? Opera and Mozilla would probably say that the success of the <canvas> tag is not tied to IE adoption, but I guess only time will tell.

Relevant Links: - Provided by Rita Turkowski
http://wiki.mozilla.org/Canvas:3D
http://my.opera.com/WebApplications/blog/show.dml/261474
http://annevankesteren.nl/2006/05/canvas-3d

Posted by Aaron Bergstrom on 2006-10-11 | Permalink | Comments