Re: Validator charset

olivier Thereaux wrote:

> That code may be a mistake... I don't recall being around when
> it was added, so it may be coming from a zealous interpretation
> of RFC 3023...

Hard to judge, "it annoyed me" doesn't count.  But you wrote that
the validator ignores the Latin-1 "default" in RFC 2616 for other
text/* types.  If that is okay the validator can also ignore the
text/xml US-ASCII "default" in RFC 3023, simply for consistency.

Ideally the validator could treat "default" as a default, using
it only if there is no better indication within the document, but
that is a separate design decision.

 Frank