 |
From: Alex Russell <slightly...@google.com>
Date: Tue, 18 Aug 2009 22:33:43 -0700
Local: Wed, Aug 19 2009 1:33 am
Subject: Re: W3C Ain't So Bad (was: OpenAjax Browser Wish List 2009)
On Tue, Aug 18, 2009 at 9:44 PM, Doug Schepers<d...@schepers.cc> wrote: Absolutely, and I think my issues aren't with what's been done, it's > Hi, Alex- > (Sorry to hijack this thread a bit.) > Alex Russell wrote (on 8/18/09 9:10 PM): >> That's totally fair. That said, using me as your example is perhaps > Please don't get me wrong. I wasn't defending the CSS WG, per se... I > This is not to totally dis the CSS WG. I know how hard it is for WGs to what hasn't been done. That shows some good taste, if nothing else. > But... come on, can't a brother get some rounded corners up in here? Amen! >>If you How about this: I'll see what I can scrounge up on the manpower front >> think my involvement would have definitively tipped the scales in the >> positive direction regardless of anything else that has happened in >> the WG...well, then you have a lot more faith in me than I do ;-) > :) >> I have very little of the requisite patience for the process that is > Fair enough, it's not for everyone... but it does have to be someone. >> The point here is that change has been proposed from vendors for quite > Yes, and they have done a good turn by bringing those features to the >>and the CSS WG seemingly has plodded along, not bothering >> a.) web developers are losing man-years to the failings of CSS > I'll be honest, I don't track the CSS WG's progress that closely, and I (I can't make any promises...it's google after all...I'd most likely be volunteering myself, and as we've covered, you don't want that). As for the community emphasis, I'm not sure what else we should do. ISTM that a big chunk of what's going wrong is a disconnect between when new stuff shows up in browsers and when it's finalized. It'd be good for the W3C to make some sort of noise about how pre-standards work is good, useful, and above all absolutely essential to getting a better web. It'd perhaps help vendors feel like they can experiment more and the web development community take the side of progress over conformance (in the small but important cases where they're in conflict). >>> Everyone's already at the table. Allow me to dissent: XML was a doomed errand in whatever capacity it >> The same was ostensibly true of HTML before WHATWG. I don't take it > No, that was a case where W3C, following the lead of some of its was flogged onto actual humans to construct. Even strict XML dialects must be reliably parsed (i.e., support tag soup) should they get popular enough (see, RSS, XAML, etc.). No real-world, user-generated content will strictly conform, which is the beauty and terror HTML. It won by letting users be lazier than the alternatives. That's a war in which the deck has *always* been stacked against XML, and it can't win until it gets tag-soup parsing too. Well, that and a way to prevent totally insane behaviors like non-UTF8 encodings (bytes are nearly free or you wouldn't be using XML), namespace aliasing (extensibility != chaos), and the mountain of add-on specs (xlink, xpath, etc.). > I will assure you that W3C has now heard the message loud and clear, and Not to be a total downer, but I've been incredibly grateful to hear > not renewing the XHTML2 WG is just one sign of that. that, as I was when the HTML 5 WG was chartered and the WebAPIs group renewed. > Since that point, It's still a pretty long series of indirections. The W3C is > W3C has dramatically changed how it operates, particularly in the > browser-facing technologies... most of those groups are now operating as > public working groups, with greater accountability and transparency (not > that that leads necessarily to more people taking the time to look at > what the groups are doing). We have also made a concerted effort to > listen more to the browser vendors... to the implementers of many of our > chief technologies. This is crucial for making the lives of authors and > users easier, and that is the main group of people I'm concerned with. constructed as a membership organization for large organizations, most of whom implement the specs it produces. As a result, end-developer usability isn't exactly lost, but it's not always adequately represented. >>> Start by identifying what exactly you think the group should be I agree that it's not the W3C's job to pick outcomes or winners on >>> accomplishing, but isn't. >> Lets start with the easy stuff that's not at CR on the ow-my-eyes >> * standardize rounded corners, gradients, transforms, and all the > Amen on both of those. >> * get off the fence about web fonts and just make it happen already > W3C set up an informal mailing list to let people talk more openly and things other than technical merit. That said, the EOT-ish arguments should lose on those grounds too ;-) >> * get serious about variables, expressions, and inheritance. Bert Code was shipped in WebKit. It was then, AFAICT, "made clear" that >> Bos is *actively hostile* to progress on this front (a marked change >> from the normal indifference to progress, but not in a good way. See: >> http://alex.dojotoolkit.org/2008/08/css-variables-are-the-future/). > Bert is just one voice on the CSS WG, even if he's a vocal one. W3C put such a feature wouldn't be allowed in CSS3. I don't know all the details and the who-said-what's, but the data is pretty clear: CSS is a bear to write, maintain, and use at scale. >> In part, this means just lining up behind what the browser vendors are I'll follow up in separate mail then. >> already doing and working with them to make sure that the best bits >> get spec'd, instead of the insane flailing that you see around "medium >> priority" modules like Grid Positioning which have no implementations, >> despite informed and generally level-headed folks working on them. > You got it. It takes some direct action to advocate and move things in >> My pet peeves are all in DOM-land, but I take your meaning and will > It's appreciated. FWIW, I'm also the unwilling editor of the DOM3 >> I think I just have a totally different (but not opposing) Even if I tease them apart, both still persist in ways that harm the >> perspective. The W3C has become unattractive (in some sense, damaged >> goods) for several related reasons: >> * the XML and big-S "Semantic Seb" folk still hold sway, despite the > I wouldn't conflate the SemWeb and XML folks. long-term goals of building a better platform for sharing data that humans can use, build, and iterate on. The common thread in both is a for-machines perspective and (due to the inability to get it always right) a by-machines orientation. That's the common thread that holds both efforts back, punishing "dirty" markup (the real web as we have it) in the process. > But in any case, I don't Again, I dissent. The interaction specs have been held hostage long > think they hold undue sway over decisions in the Interaction (Browser) > domain... SemWeb and XML are a different "division" of W3C than the > browser stuff, though it would be nice for everyone to play along nicely > (viz. RDFa). enough. We need better semantics to describe the UI's that we're trying to build, and the lessons of the past decade suggest that when we get them, they'll *also* conflate huge value in other areas. Imagine a <datasource> tag that could be declaratively hooked up to a <datagrid>. Is it "semantic" in the big S sense? No, but if it provides a URL, it can help us discover the relationships between apps and the data they consume, produce, and manipulate in much the same way that the <a> tag helped us navigate the relationships between entities. > For the past several years, the clear shift in focus has I'm not asking for the W3C to be police, but put another way, the > been toward browser-facing technologies, so I don't think arguments like > that still apply. Mind you, I've only been on the W3C Team for a couple > years, so I can't speak to the past, only to the present and the future. >> * some group of folks, totally outside of the W3C and not > The Validator needs a lot of work, and is undergoing a big overhaul as interests of the membership are being ill-served by the Validator as it currently exists. My view on this is draconian: I'd rather see it shut down before it holds up progress any more. >> * where browser vendors have moved ahead and done new, useful I was frustrated at those events too, but for a different reason: they >> things, it has (of late) been done *around* the W3C, not with it. Why >> did WHATWG happen? > See above. FWIW, I don't think that WHATWG had a monopoly on were happening in the SVG WG. That's not the SVG WG's fault (much to the contrary, MSFT has much to answer for). >> For that matter, why is JavaScript still being developed at ECMA? I only propose it as one of the big "WTF?" things about the web since > As far as I know, nobody has proposed bringing Javascript to W3C (where it both makes it more difficult for implementers to get a broad view of everything that's expected and makes it harder for users of the platform to think of it as a platform. >>ISTM that the W3C's legitimacy is entirely It's appreciated. Luckily for the consortium, there aren't a lot of >> dependent on the web development community respecting its taste, >> timeliness, and neutrality. Where browser vendors need to go >> elsewhere, large fissures in the day-to-day experience emerged (i.e., >> DOM types are not JavaScript types...why?), pulled in part by the >> membership of the W3C toward goals which do not service the larger >> community of developers. An essential integration function has been >> left un-tended. > I believe that many of us at W3C have tried to be better gardeners in other places to play this game yet. There's *enough* legitimacy left to prevent out-and-out fracturing. >> * continued intransigence by members to implement specs that have This isn't about forcing, this is about finding a way. Maybe that's a >> been ratified points to either an inability to specify simple and >> coherent enough standards whose value is evident, or an inability of >> the W3C to hold enough sway amongst it's membership to ensure that new >> features have a chance of being available in the sense that web >> developers need them to be. > Well, some W3C specs simply aren't "browser-facing", but are intended >> Web developers are disenfranchised from the process of browser choice > Sorry, but I don't buy this part. W3C as an entity doesn't have, and discussion to have at membership renewal time. Maybe that's an advocacy function that shines a bright light on the needs of end-developers and how the specs would help accomplish them *if only they were implemented pervasively*. It's not at all reasonable for the W3C brand to stand for the universal adoption of these standards, though, and for that not to carry some sort of a responsibility to those who support the standards over those who don't....even amongst the paying membership. >> SVG is but one example, but it might be the most poignant. That's fair. > Actually, SVG (after being caught in Adobe plugin-land for far too long) >> In short, it's not clear that what the W3C does today *makes life > It's clear to me that W3C does help, but then I would think that. Then > But am I asking you to give W3C a free pass? On the contrary, I'm >>> I'm always a bit taken aback when I hear someone speak of W3C like it's Several things, then: >>> a sovereign and foreign country... >> Perhaps they talk that way because they are not as invested in -- and > I think you've nailed it... people need to be invested in the process, * create an individual member level, with available donation levels >> The consortium has a legitimacy problem which now manifests itself in Oh, don't think I'm letting implementers off the hook here ;-) >> lots of ways, > That's true, but that legitimacy problem is often exaggerated in certain >>but fundamentally, I suspect its roots are in a > I'm not as interested in the perception so much as I am in how we can My point of view is that implementers need to be driving anything that Regards You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
| |||||||||||||