XHTML still works by the way, all HTML5 goodness included. I tend to write sites in XHTML myself as default HTML parsing feels broken in many subtle ways.
Heh, I wouldn't even know which URLs to put in the doctype for XHTML. I certainly won't ever memorize it. As long as HTML is `<!DOCTYPE html>` and XHTML is `<!DOCTYPE some ridiculously long string with a bunch of URLs and crap>` I don't think I'll be using XHTML.
It's also pretty risky to let any slight syntax error or semantic error cause the whole page to refuse to load. That would turn me off from targeting XHTML for any kind of dynamic page generation, not because I want to generate incorrect HTML or XHTML or whatever, but because bugs happen.
Nope. The thing that browser looks for is the Content-Type header – if it's set to application/xhtml+xml, it will parse the document as XHTML. For static sites, just setting extension to .xhtml is usually enough.
