XML Activity Rechartered

Volume 7, Issue 101; 20 Jun 2004; last modified 08 Oct 2010

The XML Activity has been rechartered. This is a good thing, but it’s a shame about the processing model.

The XML Activity has been rechartered. I think that’s a good thing because I participate in several working groups under that activity: I’m a member of the XML Core Working Group (WG) and the XSL WG and I participate in the XML Coordination Group by virtue of being co-chair of XML Core. In the past, I’ve also been a member of the XML Schema WG and the now defunct XML Linking WG.

So it’s a good thing, but my happiness about the rechartering is tempered by bitter disappointment. If you examine the new charter of the XML Core WG and compare it to the previous charter you will notice that all mention of the “XML Processing Model” has been expunged.

Now this is not a conspiracy theory piece. I understand how and why it happened. There were good, or at least plausible, reasons both procedural and technical for the decision:

  • Procedurally, the previous charter is vague. While it lists the processing model as in the “Scope” of XML Core, it does not list it among the “Deliverables” of the WG (not that the list of deliverables is comprehensive). It is possible to interpret the previous charter as not actually putting the task of developing a processing model specification in the domain of the XML Core WG.

    I don’t read it that way, but intelligent, well-meaning colleagues do even though the WG did produce a requirements document for the processing model. Go figure.

  • Technically, if it was ever XML Core’s task, it’s been a task for several years and the group has precious little to show for it. We just barely got the requirements document out before the new charter was approved.

    It’s also possible to argue that from a technical point of view the processing model task has never been very clearly specified. It means different things to different people.

  • Also procedurally, amplified perhaps by our lack of clarity about the task, some folks in XML Core didn’t want to see the work done there. Maybe that’s more political than procedural, I don’t know. Doesn’t matter, anyway.

  • Finally, the public doesn’t seem to be clamoring for the processing model. I think it’s important, and I’m not alone, but it isn’t something for which people are waiting anxiously. Maybe if we renamed it WS-Pipelines (Ha, ha, only serious).

Whether you find these arguments convincing or not is irrelevant; the decision has been made. My favorite B5 quote applies: “The avalanche has already started. It is too late for the pebbles to vote.” Actually, that’s not fair. The AC did vote, they’re not pebbles, they’re the representative body of the W3C, and I tried to make sure that they knew about the change.

So all that remains is my disappointment. Why am I disappointed? Well, partly because worked hard on a spec for the XML Processing Model Workshop and argued that the work should be taken up by XML Core in the first place. And partly because I want the damn thing. But mostly because now that it has been removed from XML Core, I see only four possibilities:

  1. The task will move to another Working Group. This seems unlikely: of all the groups currently chartered in the XML Activity, XML Core is the only place that seems appropriate.

  2. A new WG will be spun up to do it. This is really the only logical course left, but doing so would require the membership to agree that it was worth doing, would require additional staff resources, would require participants to sign up for yet-another-telcon, and is generally about as light-weight as an osmium brick. It may be doable, but it seems unlikely.

  3. Some other standards organization will spin up a working group to do it. Ditto what I just said.

  4. The processing model work will slip beneath the waves and vanish, at least for now. I think this is the most likely outcome, but I’d be happy to be proved wrong.

Coding in Anger

So I fought, perhaps not hard enough, to keep the processing model work in the XML Core WG and I lost. C’est la vie. In the weeks that have elapsed, frustration has turned to disappointment until finally in a fit of pique I decided: **** ’em! If I’m not going to be working on a spec, I can always write code.


Y'know, it's really not a heavyweight process to get an OASIS TC up and running: three OASIS members (and Sun is one). That would be the obvious place to take the Processing Model.

—Posted by John Cowan on 21 Jun 2004 @ 09:38 UTC #