William F Hammond wrote:
Perhaps you should clearly state your definitions of "bad" and "good"
in this case? I'd also like to know, given those definitions, why
it's bad for the "bad" documents to drive out the "good", and how you
think your proposal will prevent that from happening.
"Good" and "bad" here apply to document instances. "Good" means
compliant xhtml+(mathml|svg)*; "bad", as I casually used it, means
other.
OK.
My only point is that a user agent should parse as xml a
document whose preamble indicates xhtml even when the mimetype is
text/html.
That would break a large fraction of popular websites out there. In
addition detecting "the preamble" requires assumption of a parsing
model. I'm pretty sure one can construct documents that have different
"preambles" when treated as HTML and XML.
Or, if that is too hard or too politically difficult,
going forward the WG should provide a formula for the front of a
document that asks for an xhtml parse.
What is the benefit over using a MIME type as now, though?
-Boris
