We don’t like invalid HTML. Yet we ourselves commit the transgression of invalid HTML from time to time, mostly under duress: We keep running into Web servers that insist on using ampersands in URLs yet are unable to accommodate encoded ampersands (&, let alone &). And we’re pretty easy when it comes to XHTML peccadilloes: <img> really is the same as <img />, no matter what the W3C validator (or HTMLHelp’s superior alternative) might claim.
But, please, megacorps, could we try a bit harder?
HTML is, in broad terms, simple, as were early authors. Simpleton Web developers forced browser makers to tolerate nonsensical HTML. If browsers had not been folded, spindled, and mutilated to display folded, spindled, and mutilated HTML in a half-arsedly correct manner, there would be no Web as we know it today, because the vaunted BARRIERS TO ENTRY would have been too high.
To enable average people to get their homepages online in the early years, we had to let them put out any kind of Katzenjammer Kids HTML they wanted. Content took priority over code. We know perfectly well that crapola authoring programs like FrontPage and ClarisWorks were the real culprits. But, to draw an analogy from print, if the choice is between typewriting or no literature at all rather than between typesetting or typewriting, we will put up with typewriting.
(What we won’t put up with is handwriting. Too many lost-kitty and garage-sale flyers stapled to utility poles have soured us on that one.)
Can someone please explain to us, though, why the world has gone Bizarro and now it is only individual people who can manage to produce valid HTML?
We would expect nothing but top-of-the-line results from big companies with serious Web infrastructures. But that’s not what we get.
It’s the one-year anniversary of a bright, shining Microsoft lie – that its Web sites comply with W3C standards – and the software juggernaut celebrates by redesigning. (Now you see how it helps that we at NUblog have been at this for a while: We can look up the history.) The Microsoft redesign is a now-notorious HTML and accessibility atrocity.
It should be amusing that a staff of hundreds cannot achieve valid HTML while some guy futzing around at his computer can – yet again. Dylan Foley, come on down!
Frankly, this is getting embarrassing. It’s the kids showing up the adults. It’s an after-school special transposed to the Web, a Mormon teaching video interpreted in HTML. It’s daddy-o peering through the bifocals at the end of his nose for so interminably long that his kids put aside their book reports and set up the damned Ethernet network for him.
But are we done yet?
We are quite happy to be of two minds when it comes to validation.
But what we want is exactly the inverse of what we actually have. Valid sites are rare, but they almost all sit on one side of the fence. The only reliable source of valid HTML is the individual Web developer, and even then only the experts; commercial sites are almost invariably invalid.
Amazon, for example.
We’ve got not one but two reformulations of Amazon.com in valid HTML. Anil Dash did it in HTML 4.0 with tables, while Steve Clay made it work using stylesheets for layout.
As a subset of this phenomenon, we would mention the case of inaccessible accessibility sites. Jukka K. Korpela rightly complains about this inconsistency, but we caution against getting on one’s high horse and leveling accusations of hypocrisy. The issue is not cut-and-dried; human beings must interpret and assess the accessibility of a site. Still, if you go so far as to announce that your new site is “screen-reader accessible,” it had better validate (it doesn’t: W3C; WDG) and at least include alt texts (some 10 are missing).
Aside: You are of course aware that the allegedly-accessible Amazon analogue, itself nearly the worst possible idea, will not even return the same search results as the real site? Try it yourself: Look up Lars von Trier’s film Zentropa just by title. You’ll find a hit at Amazon.com and even at the text-only Amazon, but not at Amazon.com/access/. Searching by catalogue number, 6302722551, works in all cases, but how accessible is a catalogue number?
We don’t want to take this too far. We don’t want to pull a Lars Von Trier (him again!) and publish a Proclamation of Dogma W4, where W4 signifies the malapropist neologism “Worth World Wide Web.”
We don’t want to go that far because we have seen just how well Dogme turned out for cinema. The Celebration, sure. Dancer in the Dark, by all means. But not Mifune. None of that, please.
As a gimmick, absolute purity vies for top honours with Queen Amidala’s headgear. But it is perhaps too strict a life<hyphen>style for the living Web.
We will, in the interim, settle for valid HTML from firms that can handily afford to produce it, and continue to cheer on the little people for showing up the megacorps for the chumps and amateurs they truly are.
Posted on 2002-10-10