""The real problem is of course that the W3C is still copying our work even after we asked them to stop doing that," [Anne] van Kesteren said. It's legal, but "oftentimes it comes pretty close [to] or is actual plagiarism."
It's one of many instances of copying, Hickson said. "For reasons that defy my understanding, the W3C staff refuse to treat the WHATWG as a peer organization" that relies on WHATWG's work, he said. Instead, it creates its own copies of some standards. "They'll eventually say they have a 'final' version, and then they'll stop fixing bugs. It's very sad."
It's amusing to me to watch the WHATWG people complain about the W3C having the temerity to put out revisions to their own specification, and then complaining about how the W3C is copying their work. Would they rather W3C put out an ENTIRELY DIFFERENT HTML specification from theirs? Would that be better for anyone at all?
Yes, actually. We'd encourage the HTML working group to do original work. Many W3C working groups do that, e.g. the webapps working group worked on web components and various other specs, and it helps the web platform.
So they should reinvent the wheel when there are good solutions out there?
Isn't the whole point of standards to stop that sort of thing happening?
I get that you may feel that someone is just appropriating your work rather than doing their own but if your aim is to build a standard then reusing things is probably going to be part of doing that.
The specs were written from scratch, because the W3C holds the copyright on them (the old specs are also of relatively little use). The spec now called HTML5 was officially called Web Applications 1.0 till after the W3C HTML WG was (re-)chartered.
I dunno, perhaps if the W3C and WHATWG produced radically different specifications, one of them would be explicitly rejected by everyone and thus die a clean death. That might be better than having two very similar, equally authoritative, but subtly different specifications, one of which is sort of but not quite a snapshot of the other.
The other huge accomplishment of HTML5 is completely standardizing many fundamental parts of the web that previously were a mess of browser incompatibilities. 6 years ago, if you wanted to parse HTML, you might reach for BeautifulSoup, or libxml, or Hpricot, or Nokogiri...and they would all be subtly different in the parse tree they produced. And they couldn't do any better, because if you viewed the page in IE, or Firefox, or Chrome, or Safari, you might get a different parse tree.
Now, IE9+, Firefox, Chrome, and Safari are all basically guaranteed to look at the same page in the same way, and the "toolsmith" parsers like Gumbo or html5lib are all rapidly converging on the standard. So it's finally possible to see a page the way a browser sees it.
> Now, IE9+, Firefox, Chrome, and Safari are all basically guaranteed to look at the same page in the same way, and the "toolsmith" parsers like Gumbo or html5lib are all rapidly converging on the standard. So it's finally possible to see a page the way a browser sees it.
The web projects I have to take part on, are a distant reality from that description.
Right, this. There's still a lot of experimental features that are implemented differently or not at all in one or more of the big three. But I can't remember the last time that I've had different browsers produce fundamentally different DOM trees from the same document, even when I'm using invalid attributes or made-up tags (coughangularcough).
If you stick to the basics of CSS that have been around for a decade or so, both of them are basically solved problems. Browsers are remarkably uniform in how they handle the stuff that was a PITA in 2008.
The problem is that expectations adjust too, and now we want all these new HTML5 features that were just pioneered in a single browser a couple years ago. Those have a lot of cross-browser issues.
In many environments, expectations don't have to adjust. With corporate clients, expectations were never reasonable in the first place, and many common requirements that have been around for decades still aren't reasonable.
Clients often demand idiocy like "pixel perfection," specific fonts, custom scaling behavior for size (including application-controlled zooming), and other things that simply aren't part of the paradigm for accessing websites through web browsers.
Between that and the fact that going "beyond the basics of CSS" is a pretty common thing, browser interfailure is still a real problem. It's not as bad as it used to be, but the problem is still there, and many of us are still dealing with it regularly.
Is there any particular reason? Requiring support for older or legacy browsers?
I just recently wrapped up a project utilizing customized Twitter Bootstrap / CSS3 / HTML5 and I was pleasantly surprised at how smooth everything went between IE/Chrome/Safari/FF. This is the first time, in a long time, that everything just "worked."
All projects tend to require IE 8+, specific versions of FF, Safari and Chrome.
Oh and the fun that is working with native iOS, Android and WP browsers.
Then put on top, whatever web framework might be required by the customer.
The fun starts when the UI isn't working pixel perfect to those Photoshop mockups or the cool HTML 5 effect that can only be partially implemented in all browser versions mentioned in a "Request For Proposal".
I always try to explain that pixel perfect just isn't feasible when we're going across platforms browsers and devices sizes / types.
If they would like to create 20+ different mockups and supply them to me I'll gladly do it for an exorbitant sum money, but otherwise they need to realize that the relative fluid nature of the web and it's many form factors is going to preclude them from getting that pixel perfect design.
Then all of that flies out the window because we have that one asshole manager who promises the world and ensures it will be delivered yesterday :)
Yes and no, it depends, If you use the entire suite of bootstrap, yes. We did not, we used it for the grid system (getting those nice big perfectly sized columns). Everything else was hand written CSS and HTML.
Indeed, the W3C's work on HTML5 is irrelevant. It exists only for the purpose of giving the W3C-qua-organization the appearance of being involved in the further development of HTML.
They have different purposes. In practical terms, the "HTML5 standard" is really defined by whatever the most popular browsers implement - we saw this in vivid detail when IE6 was the most popular browser and Microsoft made a mockery of the standard. The WHATWG standard exists to track the evolving consensus, so that the major browsers don't diverge too much and we don't go back to the web c. 2005. The W3C spec is so that other interested parties, ones who need to have a finalized doc to shoot for, have something to shoot for that everybody has agreed upon.
I think that's disingenuous — even as someone who has been around the WHATWG for longer than the W3C has been working on HTML5. One clear reason that publication as a REC is important because it implies a royalty-free patent grant between all members of the WG.
I haven't been following the situation super-closely, but I understand that the threat of W3C not ratifying the HTML5 spec as it is (or was) has curtailed the behaviour of certain individuals on a number of occasions during the HTML5 process. If so then the W3C's involvement with HTML5 has not been at all inconsequential, even if you view the formal announcement today as unimportant.
In the past, only Mozilla cared about what the W3C said, and they built Firefox on the idea that interoperability by following the standards is the way to go. Not that long after Apple initiated Safari, and led Google to their own browser initiative.
All this because Mozilla, the insignificant actor of the all IE time, decided to follow W3C.
You're not totally wrong telling that W3C doesn't mater that much, but browsers are what they are today because of W3C for a good part, and it still have a very important place in the browser game.
This is pretty much the opposite of how things went down.
If browsers followed the W3C, we'd be living in XML utopia (XHTML2, XForms, XLink, XEvents, etc.). In 2004 a group of implementers (specifically Mozilla, Opera, and Apple) proposed to the W3C to focus on web applications by evolving HTML and the DOM, but were flat-out turned down, and had to go off and form their own standards organization---the WHATWG. Today, that is where much of the foundational work of the web platform is still done: https://spec.whatwg.org/
That gap is constantly closing. The sky is falling on Flash. Sure, if you legitimately have a project that won't work in HTML5, maybe Flash is a good choice. However, most of the time I see these statements used by Flash developers that don't want to learn javascript or admit their time is limited as Flash devs.
I'm hoping for a scene graph standard for HTML6 and a spec that allows multiple documents per window (which themselves could be nodes on a scene graph) instead of the one document per window.
The spec published by WHATWG is a living standard, and is just called "HTML". The expectation is that every so often the W3C will pick a revision of the WHATWG's HTML spec and stamp it with a version number.
> The expectation is that every so often the W3C will pick a revision of the WHATWG's HTML spec and stamp it with a version number.
W3C HTML5 is not just a snapshot of a particular revision of the WHATWG Living Standard. In a reasonable world, W3C HTML5 (etc.) would be a subset of the WHATWG Living Standard on the date the former was published, but I'm not sure that even that is strictly the case.
Well, this announcement seems to indicate that from W3C's point of view, it's not living anymore. They've finalized the standard, and are starting work on HTML 5.1. WHATWG apparently no longer numbers their standard. [1]
Does anyone know how is Doctype versioning going to work with W3Cs snapshoting of the living standard? Or are browsers just going to ignore W3C and stick to implementing WHATWGs spec?
No browser ever paid any attention to the version of the HTML specified. The only use for DOCTYPE—and the reason it remains in HTML5—is "doctype switching", i.e. different rendering modes are triggered in the browsers. You can use "html", "html5", "html6.2" or "foobarbaz" as your doctype, the effect will be the same—they will all trigger standards compliant mode.
>'You can use "html", "html5", "html6.2" or "foobarbaz" as your doctype, the effect will be the same—they will all trigger standards compliant mode.' //
You're wrong if your asserting that any doctype produces the same results globally but it probably doesn't matter unless you're targeting legacy UA.
Maybe but if you use real DOCTYPE such as <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd"> then you will see huge difference in rendering.
I was hoping they would add file api support for downloading to non sand boxed environment with mandatory user interaction prompts. The api would prompt for the file download location (via file dialog) but after that, how the file is filled up is up to the web client (and happens in the background). This would allow for parallel download workers via the existing get range option.
http://html5doctor.com/the-ride-to-5/ is an interesting overview of perspectives of various people who've been around HTML5 for years about its publication as a REC.
""The real problem is of course that the W3C is still copying our work even after we asked them to stop doing that," [Anne] van Kesteren said. It's legal, but "oftentimes it comes pretty close [to] or is actual plagiarism."
It's one of many instances of copying, Hickson said. "For reasons that defy my understanding, the W3C staff refuse to treat the WHATWG as a peer organization" that relies on WHATWG's work, he said. Instead, it creates its own copies of some standards. "They'll eventually say they have a 'final' version, and then they'll stop fixing bugs. It's very sad."