MozillaZine

Full Article Attached Open XUL Alliance Site Goes Live

Sunday May 25th, 2003

Gerald Bauer writes in with news that the Open XUL Alliance site has launched. The site aims promote XUL and encourage interoperability with a collection of XUL news articles, mailing lists and links.


#19 XUL, the W3C, and Peer Review

by choess <choess@stwing.upenn.edu>

Monday May 26th, 2003 10:48 PM

You are replying to this message

Gerald,

I'm not trying to argue that XUL should fall under the aegis of the W3C, although in retrospect I can see how my posts suggested that. Indeed, I'd be happier if the W3C divested itself of some of the more tangential recommendations it develops, such as the Web Services-related activities.

What I am arguing is that there needs to be some sort of forum for establishing a universally agreed-upon definition of "XUL". The attitude I'm seeing is "We'll all throw something together, and eventually we'll write down all the good stuff and make it a standard." If there's more of a process that I've missed, I'd be happy to hear about it. I don't think this is a viable model.

To talk some specifics: XForms: I'll withold comment other than to say that denouncing XForms in vague and less-than-constructive terms, about 3.5 years after public discussion first began about the recommendation and its direction, may have been less than politic.

XML Schema: This is a good example of the W3C process gone awry. Because of pressure from the Web Services side (esp. including Microsoft), the final product was bloated, overcomplicated, and is disdained by those seeking to create XML languages such as XUL in favor of RELAX. Unfortunately, this seems like the sort of problem that would only be exaggerated by a "implement and sort out later" approach. Each implementor will probably have a different vision of what they want the language to do, leading to overspecialization in different areas in each implementation, and making them very difficult to reconcile later.

That said, here's what I think is a reasonable approach to guiding XUL development:

Form some sort of "XUL Consortium" or whatever you want to call it. I don't think it need be as formal or fee-based an organization as the W3C; open source implementations will be important here. Essentially, membership should be open to anyone actively working on a XUL implementation, or major consumers of XUL (the two will probably be the same initially, although I can certainly see major software products implemented on such a toolkit as yours). Other than that, I see XUL evolution progressing roughly in the manner I described in my last post: members (all of whom have some "hands-on" experience with XUL development and implementation) create a working draft based on what they *in consensus* view as useful additions (or deletions!) to XUL. Once they've hammered out a document that everyone can at least agree on, the new features can be provisionally implemented, and feedback from real experience provided to the group (good, bad, hard/easy to implement, hard/easy to use, etc.). Again, when a consensus is reached and two interoperable implemenations exist, it can be considered a standard/recommendation.

What I think is most important is the element of peer review. You may dislike committees/consortia as incapable theoreticians, but your own perspective as an implementor is likely to be colored by your own knowledge, and the needs of your applications and its users. But any standard/recommendation is more than you and your own application, and having other users of the standard, with their own knowledge and needs, weigh in on the direction of the standard is an important check to keep the standard relevant to everyone.