It seems like the best way to go is to take steps to ensure the ongoing validity of HTML 4 in combination with CSS. We call that freeze-dried spec "The Presentational Web" and leave it the hell alone except to decide over things like whether or not to canonicalize <blink>. Then the semantic thing going on can go on unimpeded, allowing the academics who inspired the Web (with whom Sam and I will likely never come in contact) their more "complete" standard so they can continue the ambitious process of cataloging and abstracting every single piece of data that ever existed.

Is this pesky and ill-informed? Likely so. But Sam has a point when he complains that the altar in the Standards Temple is slowly being sealed behind a veil by the priests who maintain it. The alternatives we face in terms of making Web publishing accessible involve two unsavory and unlikely choices: either trusting "normal joes" to master an increasingly opaque way of authoring a page, or trusting software intermediaries to do it for them, rendering the experience as word processor/dtp-like as possible. The first is a pipe dream, the second has as evidence of its underlying fantastical nature every page of HTML ever squished out by Microsoft Word or Netscape Composer.

With a Web-lite, we can get on with the part of personal publishing on the Web that's exciting to normal folks, which mainly amounts to virtualizing the water cooler and hyperlinking the family vacation slide show. That doesn't suit the people who want that "data" to "mean" more? Well, tough... I don't let the W3C tell me how to make my bed or organize my photo albums, either.

Why should we care about this? Because setting a higher and higher bar to "standards compliance" stifles creativity. It reflects the outrage of technicians over the presumption of an Idiot's Guide to Web publishing that describes anything besides passive consumption of the orderly and collected data provided by the temple priests. It's a petty vicissitude from the "Company Computer Guy" writ large, with the democratizing effects of a wonderful technology at stake

I don't mean to say that I don't like standards, or think that the W3C should be ignored. Last year I was happy to assist the Web community in dealing with the W3C's nutty insistence on patented Web standards. In private mails I argued against people who believed the W3C had lost its mandate to establish standards even as I made sure that every story that might pressure them to make the right choice got posted on LinuxToday. Eventually, they did the right thing. It just took over a year.

In the mean time, I'll confess to having given up on my own pages.

When I make a new design, over the course of a few days I load it up in Internet Explorer (6 for Windows, 5.2 for Mac), Opera (on Mac), Chimera (Mac), Safari (Mac), and Mozilla (Mac and Windows). I make sure the stylesheets can't be read by Netscape or Internet Explorer 4 or earlier, then I call it a day. I don't validate it, and I try to be polite when someone writes and tells me the design is broken for whichever browser I missed. Even "helpful" validators are opaque, and mixing Movable Type's quality markup with my own amateur stuff causes all hell to break loose. Does this mean my stuff won't work forever? Yes, it does. It will break. Browsers will begin to fail when they encounter it. I'm consigning myself to a life of fixing bugs and glitches. So it goes. I've got a small audience that can read this stuff. When they start telling me they can't anymore, I'll fix it. They're who matters.