15:47 <@TabAtkins_> alex: I wrote an incendiary blog post a while ago, with the suggestions that the W3C divest itself from all the technologies that aren't specifically web.
15:47 <@TabAtkins_> alex: I was hoping we'd have more people from the SemWeb side of things to discuss this with.
15:47 <@TabAtkins_> Henry: I can play devil's advocate.
15:48 <@TabAtkins_> Henry: I skimmed the blog post and some of the followup.  At that point you weren't as clear as you are now, but I hope to get you to be even *clearer* here.
15:48 <@TabAtkins_> Henry: Who's gonna be pushed out?
15:49 <@TabAtkins_> alex: Anything SOAP or XML web services.
15:49 <@TabAtkins_> Mike: They're gone.  Nobody's interested in those.
15:49 <@TabAtkins_> Mike: I voted that XML activities should be wound down.
15:49 <@TabAtkins_> Mike: The rest of the team responded politely, but...
15:49 <@TabAtkins_> Mike: I think XML had its time, and web services too.
15:50 <@TabAtkins_> Mike: They're not absorbing much resources any more.
15:50 <@TabAtkins_> Monica: What about the social web charter?
15:50 <@TabAtkins_> Alex: My hope is not a discussion about my specific prefs, but rather what it woul dmean to servic the web better.
15:50 <@TabAtkins_> Daniel: In the list of groups you probably don't want, you'll likely include e-government and similar things.
15:51 <@TabAtkins_> Daniel: So e-government isn't really aligned.  BUT, it bring sin a lot of money, and attention from the gov.
15:51 <@TabAtkins_> Mike: 5 years ago web services brought it a lot of companies like Hitachi; it was a revenue generator.
15:52 <@TabAtkins_> Alex: Is the set of things which are functionally unaligned with the web - do we accept them as revenue generators only?  Should someone communicate that they're being played?
15:52 <@TabAtkins_> Daniel: It's not solely revenue.  Making orgs outside of IT adopt web techs requires communication.
15:52 <@TabAtkins_> Daniel: Spreading the word to IT companies is trivial.  But tallking to the french military police is very complex.
15:53 <@TabAtkins_> Alex: Does their buy-in demonstrably change how things evolve?
15:53 <@TabAtkins_> Alex: If web tech is successful otherwise, is it useful to laise with them
15:53 <@TabAtkins_> henry: If they're revenue-neutral...  We can stipulate they bring in at least as much as they cost.
15:53 <@TabAtkins_> Henry: Is there a net benefit to having them?
15:54 <@TabAtkins_> Alex: I think you could also argue that they drag technology away from the natural use-cases.
15:54 <@TabAtkins_> Daniel: Elaborate?
15:54 <@TabAtkins_> Alex: It's special pleading in the general case, but...
15:54 <@TabAtkins_> Alex: There is tech which won't be adopted by anyone without standardization.
15:54 <@TabAtkins_> Alex: And there is tech that will be used no matter what.
15:55 <@TabAtkins_> Henry: You could dramatize the original HTML move from IETF to W3C as the IETF wasn't handling the players properly, so the W3C got the business.
15:55 <@TabAtkins_> Henry: You also have, the membership says "we need agreement in X, nobody's there yet, but we're headed there".
15:56 <@TabAtkins_> Henry: There is a direction in which the browser vendors wanted to move, and they said we should point in that direction.
15:56 <@TabAtkins_> Henry: Those both happen here.
15:56 <@TabAtkins_> Daniel: Back to the roots of the problem.  I hear your request to downsize.  Why?
15:57 <@TabAtkins_> Tab: We have an existence proof - when HTML went to the WHATWG because the W3C's companies didn't like HTML, but the browsers did.
15:57 <@TabAtkins_> Alex: Similarly, RDF and its integration.  It's not widely used publicly, but it receives great attention.
15:58 <@TabAtkins_> Alex: Is there a policy the W3C could follow to adjudicate this better?  Or should the WGs or interests to be on equal footing?
15:58 <@TabAtkins_> Henry: That's the right basis to have a discussion.  How do we broker discussions between customers and the future?
15:59 <@TabAtkins_> Henry: HTML left for the WHATWG because W3C made a wrong predication about where the web went.
16:00 <@TabAtkins_> Mike: I don't think so.  It was 2008 or so.  There was a vision handed down from on high.  It wasn't a predication, it was a pronouncement.
16:00 <@TabAtkins_> Henry: Let's not get too much into the details of that, but it was *specifically* that the W3C made the wrong decision.
16:00 <@TabAtkins_> Cameron: I think pehaps the W3C should then not be trying to predict the future, but rather to just facilitate.
16:00 <@TabAtkins_> Henry: So how do you do that?
16:01 <@TabAtkins_> Daniel: They see the web from a different point of view, a different owner.
16:02 <@TabAtkins_> Tab: To come back to the subject of the talk, a leaner organization wouldn't *have* to make these sorts of choices.  They could bet on more horses.
16:02 <@TabAtkins_> Alex: So I'd like more discussion from people around when CSS started.
16:03 <@TabAtkins_> Alex: I understand that CSs at the time didn't come from people *knowing* that it was good, but rather from a prediction.
16:03 <@TabAtkins_> Alex: But other technologies today, like SemWeb, seem to be in a similar boat, but should they be treated the same?
16:03 <@TabAtkins_> Daniel: CSS started at a different time.  There was *nothing* at the time.
16:03 <@TabAtkins_> Daniel: Most of the stylesheet languages were batch processors only.  Nobody thought of dynamic environments.
16:04 <@TabAtkins_> Daniel: It was in the first browser war.
16:04 <@TabAtkins_> Daniel: You're right there were at least two directions we could have gone.  There was JSSS in Netsuite, and Microsoft had nothing.
16:04 <@TabAtkins_> Daniel: CSS came from the consortium, or rather individual contributions.
16:04 <@TabAtkins_> Daniel: And Chris Wilson made it enter IE, which made it a success.
16:05 <@TabAtkins_> Daniel: The web was so young that it was still possible to propose things out of the box and make them succeed.
16:05 <@TabAtkins_> Daniel: At that time in France we probably only had about 150 websites.
16:05 <@TabAtkins_> Henry: I thought it was a useful example of where the W3C came out in front, and got it right.
16:06 <@TabAtkins_> Henry: It's easy to pick winners with hindsight.
16:06 <@TabAtkins_> Alex: My point is, should the W3C be in the business of predicting, or only using hindsight?
16:07 <@TabAtkins_> Henry: Really hard standards *need* prediction.  You can't let powerplugs be experimented with; there's too much coupling and coordination.
16:07 <@TabAtkins_> Henry: Is there a space of innovation outside the W3C where the W3C can act as a referree, patent guarantor, etc.
16:08 <@TabAtkins_> Henry: But not try to predict things?
16:08 <@TabAtkins_> Mike: That seems like what CGs are for now - we can let a thousand flowers bloom there.
16:09 <@TabAtkins_> Tab: So let predictive approaches only  happen in CG, offering the special W3C services (basic referring, tech, patents) which it's good at and can provide cheaply.
16:09 <@TabAtkins_> Tab: And only make a WG when it proves itself.
16:10 <@TabAtkins_> Henry: What the W3C puts energy into stuff that the membership tells it to put energy into.
16:10 <@TabAtkins_> Sylvain: How does canning RDFa or whatever let Tab ship CSS faster?
16:11 <@TabAtkins_> Monica: Better separation between specs?
16:11 <@TabAtkins_> Daniel: We miss a strong feedback loop to *stop* activity.
16:12 <@TabAtkins_> Mike: Look at all the effort that HTML and the TAG went through to talk about RDFa/Microdata.
16:12 <@TabAtkins_> Sylvain: That's a sunk cost, though.
16:12 <@TabAtkins_> Tab: And can we avoid those sunk costs in the future?
16:13 <@TabAtkins_> Tab: The W3C can offer useful services cheaply that only it can eaisly do (tech, patents, etc), and only get heavier (/more expensive) when things have proven themselves.
16:13 <@TabAtkins_> Alex: We can point to examples where many people suffered a lot of pain for the sak eof tech that was optimistic.
16:13 <@TabAtkins_> Alex: That draw a large mindshare externally *on the basis of W3C sponsorship*.
16:14 <@TabAtkins_> Alex: Let's be honest.  The W3C has legitimacy, I would argue, because it defines tech relevant ot them.
16:14 <@TabAtkins_> Alex: But if they make things not relevant to the devs, it hurts our flexibility and reputation.
16:14 <@TabAtkins_> Alex: XHTML2 hurt us a lot.  HTML5, if it doesn't settle down well, will be the same.  CSS seems to be on a good keel.
16:15 <@TabAtkins_> Alex: The membership of the W3C depends critically on the members believing in the W3C.
16:15 <@TabAtkins_> AlexMog: Any examples of things that we are doing wrong right now?
16:16 <@TabAtkins_> Mike: Recently, RDFa/Microdata.
16:17 <@TabAtkins_> Tab: The question itself is the problem.  We don't need to predict winners or losers.
16:17 <@TabAtkins_> Henry: By that argument, should we remove both Microdata and RDFa from HTML5, and let them grow individually?
16:18 <@TabAtkins_> Alex: Intersting.  I think the W3C has done exactly what you ask.  And we still have pain.
16:18 <@TabAtkins_> Jacques: I think we should move to considering better processes rather than being extreme with downsizing.
16:18 <@TabAtkins_> Jacques: CGs are something we should look at.
16:19 <@TabAtkins_> Alex: Can we move some techs to CGs?  Mike suggested there are some things on their way out.
16:19 <@TabAtkins_> Alex: So could the activity continue to have a published spec, but operate as a CG?
16:20 <@TabAtkins_> Mike: I'd at least say that *future* proposals should incubate in CGs before gaining a WG.
16:20 <@TabAtkins_> AlexMog: There are problems with creating experimental technology.  When something starts on the web, it often just remains there.
16:20 <@TabAtkins_> AlexMog: If we made a tech in IE and released it before it was a proposed standard...
16:21 <@TabAtkins_> AlexMog: First, it's a proprietary tech we'll be bitten for.  Second, it's hard to get others to follow.
16:21 <@TabAtkins_> AlexMog: And we could screw up in ways we wouldn't have done if we talked to more people.
16:21 <@TabAtkins_> Mike: Client-side storage is a place this has succeeded.  There were lots of solutions and prototypes, but they weren't *proprietary*.
16:21 <@TabAtkins_> Mike:  They were pre-standard.
16:22 <@TabAtkins_> Alex: And that' simportant.  The W3C lending its weight to a proposal before it's chosen/proven can lead to people believing that the W3C is *the* way that tech is developed.
16:22 <@TabAtkins_> Alex: But we can make the distinction between proprietary and pre-standard.
16:22 <@TabAtkins_> Henry: A large practical thing here is that the W3C team are who they are because they have a degree of scientific engagement.
16:23 <@TabAtkins_> Henry: We would have to change the W3C team if there wasn't the forward-moving part of it.
16:24 <@TabAtkins_> Tab: In the standard hierarchy, user > author > impl > spec author.  They're all > than staff.
16:24 <@TabAtkins_> Kevin: And if we're seeing that that leadership position isn't healthy, we should change it.
16:24 <@TabAtkins_> Alex: The end users give the W3C legitimacy.
16:25 <@TabAtkins_> Henry: If user demand is your litmus, you have to accept that the XML community has large and substantial interests that the W3C needs to serve.
16:25 <@TabAtkins_> Alex: There is a scientific question to answer here, yes.
16:26 <@TabAtkins_> Alex: There's also affinity.
16:26 <@TabAtkins_> [session ended]