Following on some off-blog correspondence, Hyatt and the Safarians look like they’re willing to try a sensible semi-pseudo-namespaced approach. My earlier piece on this provoked a flurry of conversation. Herewith some technical notes, plus words on the culture and politics of adding new tags to our browsers’ diets.
Semi-Pseudo-Namespaces ·
What Dave Hyatt is proposing to do is namespaces.
Arguably, they’re
pseudo-namespaces because namespaces are an XML thing and they’ll be
supporting <apple:canvas> or some such even in (perhaps
non-well-formed) HTML.
In this context I guess it’s OK to insist, as Dave proposes, that the
xmlns:apple="whatever" be on the <html> element.
However, they’re only semi-pseudo because in the case that someone actually
uses real live XHTML and serves it as such, they’re going to be able to put
the namespace declaration wherever they want and use whatever prefix they
want and it’ll have to work.
It’s worth noting there’s some prior art for this in Microsoft’s proto-SVG VML vector graphics language; they do exactly this trick of using XML namespace syntax even in the bowels of any old Web page. Here’s some documentation. It’s a perfectly legitimate approach.
Another thing that Apple is doing that’s not worth getting mad
about is
the IE bugwards-compatibility hacks, things as basic as
innerHTML and working upward from there.
This is a fact of life in the Web, and particularly forgivable
given Safari’s generally superb CSS implementation.
Anyhow, a tip o’ the hat to Br’er Dave and the Appleonians on this one.
Party Like It’s 1996, Redux · A very small number of people, in email and online, spoke up to share my concerns about polluting the HTML waters (notably Eric Meyer), but most didn’t. The people who thought things were more or less OK offered these reasons:
Don’t be a standards policeman, innovation is good, go with the flow.
It’s OK because this is just for Dashboard and Safari/RSS.
It’s OK because the W3C has dropped the ball and all the cool stuff is happening over in the WHAT Working Group.
On the first argument, I need only respond: “What if Microsoft were doing this?”
The second point is better, and in fact if they ship
<apple:canvas> as opposed to looks-like-HTML-but-isn’t
<canvas>, I agree 100%, that’s a good solution to the
problem.
What TF? · There are definitely some strong emotions swirling around here; check out Joe Gregorio’s little explosion and follow the links from it.
I’m fairly unexcited about both sides in this shouting match. I am cynical about the W3C HTML Working Group, having been on the receiving end of one or two too many little jewels like this. And I’m about equally impressed by the WHAT gang’s sweeping denunciations of XForms, SVG, and friends as by the HTML WG’s flamage against anything they didn’t invent.
Having said all that, WHAT has some smart people and if they can hammer out consensus on how to move the state of the browser art along and actually get stuff implemented, more power to ’em. Somebody pointed out they should have constituted themselves as the WHAT Task Force, as opposed to Working Group, for a more memorable acronym; anyhow, I just joined the mailing list.
HTML What? · I’m sitting here typing this and idly listening to the Gillmor Gang talk to Brendan Eich, who’s a smart and articulate guy, and I’m not hearing much to disagree with. Out of the blue, he just said “Over at WHAT-WG we’re thinking about HTML 5.” Oh, joy.
