W3C wants Atom
Filed in: Web, Thu, May 13 2004 01:22 PT

It’s official. Today, the W3C stated publicly that we want to start an Atom Working Group. (Later, we said it twice more when the first one hadn’t gone through.) The Atom community has proposed working under the IETF, and we have commented to the IESG that we think it makes more sense for us to work with them. We now have to hope for two things: that the IESG sees our point, and that the Atom community agrees we’d make a good home. (Here is where I mention I’m not the official W3C spokesblogger, but I do know what’s going on.)
Here are the answers to the FAQ: No, it doesn’t have to be RDF. No, it won’t have to take ten years to become a standard. No, you don’t have to pony up $5,750 a year to become a Member to participate. Yes, it will be royalty-free. (All our new stuff is.) Yes, you can do everything in the public eye. No, some W3C Member company didn’t put us up to this. Yes, we can make Atom feeds of our RSS feeds.
Why do we want Atom work to happen here? A number of us on the W3C Team know that Atom is an important missing portion of the Web’s architecture. We see something that is relevant to more than blogging apps. Since Atom also contains a protocol for extracting and manipulating content, I can see loads of potential for content management systems in general. I can also see new applications of Atom such as a system for maintaining threaded discussions across sites, while keeping the overall referential integrity of the Web intact. RSS can’t do that. Atom can do it without breaking a sweat, or worse, suffering from scope creep.
And that, in a nutshell, is why we want to standardize it. At the W3C, Atom would go into the big format stew with HTML, XML, RDF — and who could forget WebCGM? Actually, WebCGM and PNG are good analogues to Atom: it was a mature group who came to us with a good spec, built it out inside the W3C’s process, and walked away about a year later with a Recommendation.) The profile of Atom would be higher here than I think it would be at IETF: we have an active communications team that promotes our work all over hell’s half acre. We do our press releases in English, French and Japanese. We’ve even got an office in Morocco. What you want to have happen once you release a standard is to get the word out. That’s something we do very well.
And we care about the stuff that we’re doing. We know our work needs to be complete, clear, well-defined. It also needs to work well with others. We have an Internationalization Activity to help working groups avoid potholes when they get localized. We have a device-independence group to make sure things like event models and interaction modes work on a wider range of devices. I work in the Web Accessibility Initiative, and we review the specs that come along to clear away barriers for users with disabilities. We also have a Candidate Recommendation period specifically to ensure that specs are actually implementable and interoperable.
Does all that review take time? Yeah, a little. But it makes for better specs. Even with these groups looking in, based on the work already done on Atom, and the number of implementations already in the can, It could be one of the shortest Candidate Rec periods ever. Atom could be a full Recommendation by mid-2005.
So, all of you in the Atom community, think it over.
We like you.