Brutally missing the point

Statements like "real people don't care about the W3C specs, they care about the website they just visited and the content there. Are you interested abstract purity, or making a sale?" are a bit dated given the work those guys (including Pilgrim) have been doing over the past couple of years on HTML5, which is largely an attempt at writing down how most browsers act in the wild today, not some crazy excercise in spec-writing for the sake of spec-writing. Joel's post, too, might have made more sense a couple of years ago when the W3C was stuck in the upper atmosphere and when Microsoft wasn't chairing the next-gen HTML working group. The whole discussion seems to be full of lots of old assumptions and ignorance in general.

I'm not sure anyone's made a point worth getting yet.

If the new spec has even one 'behavior is undefined' in it, then user agents will behave differently. This is one of the major problems with the original HTML and CSS specs, you can try and conform but they leave much undefined. Not to mention the obscure wording whereby two intelligent readers interpret the intent quite differently.

"If the new spec has even one 'behavior is undefined' in it, then user agents will behave differently."

Absolutely, and this is precisely what the HTML5 effort is working on, by having the browser vendors get together and discuss these incompatibilities, and come up with standard error-handling so that every browser treats these "undefined" cases the same.

The problem is that in the years that Firefox/Mozilla, Opera, KHTML (and later Webkit/Safari) were all cooperating to increase interoperability, Microsoft were ignoring IE development, ignoring web standards, and hoping the mass of their platform would create sufficient gravity to pull everyone into their way of dealing with the web. Tying web standards to proprietary IE features was microsoft policy.

Microsoft found themselves in the unfamiliar position where their platform strategy had failed, Firefox was gaining ground slowly, and the only thing to do was try to play standards catch-up. Because Microsoft ignored standards for so long, it's going to be a painful process for them, and for the users they've been screwing for most of this decade with a deliberately inferior product.

The real solution (as described here) is to emulate what Apple did when transitioning to Safari from the (discontinued) IE for Mac:

• Ensure IE8's user-agent string and document.all behaviour does not trigger all the "make it work in IE7" workarounds out on the web, so it has the best chance of getting standard-like content.
• Put in the hard yards that every single other browser vendor has already had to do in order to maintain compatibility with the real web, including site-specific workarounds and evangelising compatibility to existing sites. 
• Keep IE7 around on systems for a couple of years for backwards compatibility

Pilgrim also has a great summary of Microsoft's involvement in the HTML5 process.

Jim, this might be true if sites weren't going to have to be updated anyway. As you know, if your site isn't continually updated with new information and occasionally with a new layout, no one will come back. Right now, when you develop a site, you have to test against different versions of the browsers that your site's visitors are likely to be using, including Opera and Safari/Konqueror. Then you have to come up with work-arounds for the quirks of different browsers (or if a product's users are not important to you, put up a "please use product X" sign when those users visit the site).

The original IE8 proposal would have required that all browser-makers and all sites change their code so that sites could decide when to use standards-mode and when to use quirks mode (or "tag soup" mode). In corporate and government environments, IE7 hasn't yet had much uptake in the corporate world yet. (That's where the bulk of browsing happens, and the reason why IE6 is still such a big chunk of the market.) So there isn't much of a downside if Microsoft pushes out IE8 later this year with better standards compliance as the default mode of operation.

I'd think you'd look forward to not having to do as much sniffing and compensating for their failure to keep up over the years. Or as my friends put it, create the site and then add the IE-specific hacks to make it work right for IE users. (In my work environment, IE6 is all we have, so it isn't so bad.)

Mark Pilgrim's embrace of HTML5, is a step backwards, I think. WHATWG writing competing standards with W3C looks like it could make it even harder to get the standards compliance that makes it easier for browser makers, site developers, and end users. But I admire the pragmatism that led to the change. I've been using his Dive Into Accessibility at work to try and upend some of the obstacles on our intranet.

As it is, Joel Spolsky is a fascinating read, but a better Microsoft apologist than Doug Mahugh ever will be. If I adored a former employer that much, I'd go back and work there.