Yesterday I finally left the HTML working group. While it was a recent discussion that made me ask myself what was the point in staying, I have grown more disenchanted over the years.
Unlike some I don’t agree that XHTML 2.0 is an unmitigated disaster. There is much good stuff in the spec. It is also the first structural work on HTML since HTML 4.01, five years ago. In the meantime the focus has been to make HTML XML-compatible and to split HTML into modules. Whether HTML is formulated in XML or in some kind of SGML matters, but it isn’t important relative to how the language itself should evolve. HTML 4 is overdue for an update.
XML (and thus XHTML) has two obvious advantages. It separates between syntax errors (well-formedness) and expectation errors (validity), and the character set of the document is not in doubt. The XHTML modularization itself isn’t that useful for practical purposes where the conformance rules and the implicit fallback principle (ignore what you don’t understand) is what you need. A side effect of the work on XHTML 2.0 is that XHTML Modularization will be updated to actually describe the modules in the document itself, and not by reference to HTML 4.01 as has been the case up until now.
The revolutionary aspect of XHTML 2.0 was overdone. I don’t think the W3C, or browser vendors, or web editor vendors, or any other organization or group has control over the web or its formats. I don’t overly worry about Microsoft either. They tried to replace the web before and failed spectacularly. They might try again, but if they do they will fail again. The web is the world’s largest installed base. It will change, but not on command. A decade ago change by decree might have been feasible, but this is the difference between shepherding a flock of hundreds of sheep and one of several billions. That flock goes where it wants to go, not where we want it to go.