Thinking Clearly .

Some have asked, “Why not do OWL 1.1 under the auspicies of the W3C in an “incubator group” (XG)?” Many of the things I wrote about the OWL 1.1 effort (lightweight, agile, etc.) are relatively true about XGs (at least, by their design; there are none, so no experience).

So why not do an XG? Well, historically, it was because when we tried to engage the W3C in the workshop/1.1 effort, we were rebuffed. The strong impression we came away with was that the W3C was not interested in evolving OWL at this time, and, in fact, was quite wary about anyone even mentioning that change or additions might be a good thing. It was akin to what Danny Ayers wrote in comments on my original post: Don’t hurt adoption by making it seem like OWL is unfinished. I don’t think this was universal from W3C folks, but it was a strong enough impression that we got the sense that to “play nice” we should do what we ended up doing.

Also, we were unfamiliar with the XG process (I think most of the documentation only came out right before the workshop; it, obviously, can’t have happened before the incubator activity started). The Incubator Activity is fairly recent and we had no idea what it would take to put together an XG. At the workshop, we didn’t want to lose the great momentum we had. (In fact, I had the the mailing list set up during discussion so everyone there could get the address right away.)

We still want to work through and with the W3C (hence our goal of a submission). My hope with the workshop and OWL 1.1 effort was to make it more likely that we will get new Recommendations.

Two other points in this ramble (ramble ramble):

1. There’s not a lot of agreement left to be forged, so we probably don’t need anything fancy. The implementors are going to implement this stuff (and we’ll throw out whatever they won’t). It’s all considered high value (i.e., there are actual users) and low pain (i.e., not that bad to implement). If we all implement this stuff (the same way) then that’s all that needs to be done.
2. The workshop is its own thing and doesn’t really fit into the W3C (e.g., it’s an annual event—we are hoping). I think it’s perfectly reasonable for the W3C to show up at next years workshop and try to see if there’s interest in an XG from that community.

Or if they’d really like to see it happen sooner, come on the mailing list and woo us. All we had was Sandro Hawke sitting in on one session mentioning the possibility of an XG (with no details really; it wasn’t a hard or even medium hard sell) and now Dan Connolly in a comment. AFAIK, there was no W3C participation in the workshop, so what do they expect? We must do extra due diligence about every possible way of doing something at the W3C?

(Once the documents are done, we’ll circulate them far and wide for comment. We’re not exclusionary. After all, one reason I wrote that post was to raise awareness. On the other hand, a lot of people have rather unrealistic expectations about the weight of their voice and comments. I weigh these three things most highly in debate: 1) technical merit, 2) user need, especially where I think that satisfying such users will give me good talking points, and 3) implementor resource allocation. Everything else weighs pretty close to nothing for me. The stuff in OWL 1.1 is supposed to score highly in all of these, or very highly in either 1 and 2 and we don’t have outright refusal from the implementors. So, take QCRs and datatypes: we understand the technical points, they are really big with the bioinformatics folks (who are showcase OWL users) among others, and aren’t that bad to implement (Pellet will have some pain with QCRs, but we’ll suck it up). Most of the SROIQ features are easy and there was reasonable strong interest. Punning has strong interest, and I feel should be easy enough to do, but I still have some technical qualms. We’ll work these out.

I could be convinced by marketing points. If some big company said that they would kill OWL if we went ahead, obviously that would count. But I want specific, grounded evidence, not vague worries.)

Innocent people illegally incarcerated for four years at Guantanamo and the courts have no power to free them. The administration could do so in a heartbeat. Words fail me.

This entry was posted on Saturday, December 24th, 2005 at 2:23 am and is filed under OWL, W3C. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.