Dysmey Post > Essays > Computing > What Are Not Web Standards

What Are Not Web Standards

Apple ≠ Web Standards

On 6 May 2010, Apple put up a Web page on its site, extolling the virtues of HTML5 as the new Web standard. Apple did this in preparation for the release of Safari 5.0 a couple of days later. The problem is that the items Apple offers as examples (including a trailer for the new TRON movie) do not work on any browser but Safari — and one item works only on Safari running on Mac OS X Snow Leopard!

This is the same company whose agents in the World Wide Web Consortium (W3C), back in late 2007, helped to scuttle the Ogg codecs as the official foundation technology for the HTML5 <audio> and <video> tags. No doubt Apple intends to use its own technology, based on the proprietary H.246 codec, in order to insure that only Apple — and Disney, with whom Apple is in personal union — can present audio and video on the Web.

(BTW, Apple would like to keep it that way; but the Xiph Foundation, which developed the Ogg codecs, has replied to the Steve's insinuation of litigation with the polite equivalent to See you in court, ◊◊◊◊head!)

What makes the Apple page even more deceptive is that HTML5 is not a Web standard. In fact, HTML5 is still in working draft status. HTML5 has been in working draft status for years. HTML5 will likely continue to be in working draft status for many years more. HTML5 will not be a Web standard for a long, long time yet! Got that?

The reason for this long wait is the nature of the Web standards organization itself. The World Wide Web Consortium (W3C) is slower than molasses left outside on a winter morning. The W3C also has a tradition of corporate shenanigans dating back to the early 1990's, to the days of a nascant Web, when Netscape ran roughshod over the W3C and its own plug-in developers.

These are the reasons HTML5 originally began life with an independent Web group called WHATWC. Then, in 2007, the W3C threw away its own failed (X)HTML working groups and absorbed the WHATWC to keep HTML standards within its purview. But, as a result, progress on HTML5 slowed down to the speed of cold roofing tar and became ensnared in the corporate politics of the W3C membership — hence the Ogg fiasco.

And this is why Apple (and Google and Mozilla and Opera) talk about HTML5 as if it were a Web standard: Apple kaj aliaj do not want to wait until the W3C gets around to putting its recommendation stamp on it. But in doing so, Apple (whatever its intent) makes the Internet development community think that it is highjacking HTML5 — again!

This chutzpah, that is so typical of Apple and its chief suit-who-does-not-wear-a-suit, brought about a nasty debate on Slashdot, and an even nastier commentary on the blog of Mozilla's chief evangelist, Christopher Blizzard. Blizzard is furious because the propaganda blitz of Apple (and Google) is making even the semi-talented apelings think that only Safari and Chrome support HTML5:

This is best described by a recent experience that we had from a candidate that came in to interview. He asked:

Hey, are you guys ever going to support html5?

Like, what? Are you [◊◊◊◊ing] kidding me? Truth that marketing works. The perception-to-reality gap is giant.

Blizzard is right to be angry: He's been busting a gut trying to get the word out on Firefox/HTML5, and all he gets is ignorant crap like that above. Evidently, nobody seems to know, that in fact the only browser that does not support HTML5 is Internet Explorer.

The True Divide

This sort of folly looks like the seed of a future bifurcated Web. There would be a simple, low-grade, free-as-in-speech as well as free-as-in-beer Web, based on old XHTML/CSS/PNG standards, and used by those who cannot or will not use iDevices. And there will be the Web for the rest of them: A gaudy, tawdry, very heavily multimedia proprietary Web accessible via subscription on iDevices and similar machines.

This will be the true digital divide, not that false divide prattled about by Esther Dyson and her kind. It will be a divide almost impossible to bridge.

— Apple and other guardians of the proprietary Web — and that includes the paywalled newspapers and magazines — will not open their domain while there is profit, and lots of it, to be made. This is because the rest of them, that great general public exemplified by this cartoon, is too well-off (money coming out of every notional orifice) and too ignorant to know what they are missing from the original Internet.

— The denizens of the free Web will not work with the technologies that make the proprietary Web possible. Trying to compel the free Web into proprietary compliance by litigation will merely shatter it like a glass pane into thousands of local-net fragments, or else/also would force the denizens of the free Web to abandon the Internet entirely.

This makes the Hirschorn article in the Atlantic (Closing The Digital Frontier, July/August 2010) absurd, basing it has he does on the quotes of some leftist popinjay who had nothing to do with the hackerdom that helped to build both the Internet and the Web, and mixing in some cowboy mythology for good measure. Manifest Destiny, my ass! My claim on the true divide is not changed by this article.

If that were to happen, say hello to the new Bell System — and then assume the position!

HTML5 ≠ Web Standards

Browser-specific CSS properties (those with prefixes like mozilla- or safari-) are currently the only way to implement HTML5 in its draft state. This ensures that (outside of a bloated stylesheet) HTML5 works on some browsers and not on others. It also ensures that HTML5 is not widely adopted, since the big browser among the others is Internet Explorer, whose maker, Microsoft, has not been heavily involved in HTML5 development until recently.

This development is well known, and is a source of annoyance.

css3-transform is not proprietary [w3.org]. Nor is css3-images [w3.org], which describes gradient properties. The reason that these properties are implemented using the -webkit- prefix is because these standards have not reached candidate recommendation status and are still subject to change. A vendor prefix doesn't mean proprietary — it means experimental. Once the standard reaches final recommendation status, which can only occur once two independent implementations have been created, then the vendor prefixes will be dropped.

For what it's worth, there are a good number of people within the development community that are not happy [vcarrer.com] with vendor prefixes [quirksmode.org], but it is the best option that currently exists to ensure that incompatible implementations do not use the same property name.

This got the following response:

Likewise, it is not a standard, then. If people code their pages to fit what Apple are currently touting as a standard, they will find in many cases that once the standards are solidified, they will have to recode to ensure cross-browser support.

To make this clear, the only real Web standards are:

Almost everything else (XML-based protocols like SOAP, XLink, XQuery, SPARQL, kaj simile) is there to make the business community and their IT staffs happy, but is far too complex (and therefore useless) for everyday use.

Living ≠ Web Standards

27–28 January 2011

If this article in InfoWorld (as reported by Slashdot) is any indicator, then there are now no Web standards — or at least no hypertext markup standards — anymore.

The end of numbering for HTML versions beyond HTML5 hides two painful realities, argues Neil McAllister. One is that the HTML standards process has failed, becoming a seemingly never-ending bureaucratic maze that has encouraged the proliferation of draft implementations. That's not great, but as all the wireless draft standards have shown, it can be managed. But the bigger problem is that HTML has effectively been abandoned to four companies: Apple, Google, Opera, and Mozilla. They are deciding the actual fate of HTML, not a truly independent standards process.

The WHATWC and W3C have thrown away any HTML standards when they announced that they will stop releasing version numbers for HTML because it's, oh, so tiresome to wait until everything is finished and tested and working properly before it is released.

This abandonment is an admission: Not that HTML5 will not being a Web standard for a long time, but that HTML5 will not be a Web standard at all! It is an admission that we are going to return to the Web-standard status quo of the mid-1990's, when one company — the then-dominant, now-defunct Netscape — rode roughshod over the W3C and its HTML working group. And that was before Microsoft became its undoing.

The abandonment of Web standards implied by the falling away of version numbers signifies that today, in the 2010's, it's the turn of Apple and Google/Mozilla to ride the HTML horses, dragging the W3C, their bloated and half-dead victim, behind them. And, this time, Apple/Google are more than a match for Microsoft. Microsoft cannot gather to its aid a distrustful developer community, which refuses even to believe its claims of commitment to open source. Nor can Microsoft compel the business world to upgrade their Web browsers.

Now, with Web standards an unstable, shifting goal due to the failure of W3C leadership, it will become almost impossible for Microsoft or anyone else to convince businesses to upgrade their Web browsers. The commercial Web will become a stable but static thing, forever using <embed> and <object> tags, forever using Flash (making Adobe happy), forever eschewing the new and better standards because there is no guarantee of their existence as standards.

Thanks to the wienie-ness of the W3C, the Web is just no good for business — unless that business is Apple or Google.


Written by Andy West on 28 January 2011.