On Sun, 01 Apr 2007 03:53:16 +0200, Asbjørn Ulsberg <[EMAIL PROTECTED]>
wrote:
On Wed, 21 Mar 2007 12:29:09 +0100, Anne van Kesteren <[EMAIL PROTECTED]>
wrote:
Personally I would like to have things for XMLHttpRequest work
similarly to other content (not fetched through XMLHttpRequest), but it
seems this might be tricky to get right.
It is indeed tricky, especially considering RFC-3023 and the 'text/*'
media types. A lot of 'text/*' content on the web breaks RFC-3023 and
RFC-2046, and discussions on the Atom mailing list concluded (at least
for me) that those specifications needs to be disregarded, at least for
all XML media types.
This issue is not about the imo bogus part of RFC3023 that says that
text/xml without a charset parameter must be treated as if US-ASCII was
specified. I'll just assume that in due course we'll have a more sensible
specification that says to treat text/xml identically to application/xml.
Especially given how widely deployed text/xml is versus application/xml.
[...]
Defaulting to anything but iso-8859-1 or UTF-8 on the web today doesn't
make much sense, especially not US-ASCII. It made sense in the 80's, but
doesn't any more and I strongly believe disregarding RFC-2046 in this
regard is the right thing to do, since the compatibility of existing
MIME parsers isn't a concern the XHR specification needs to take.
This is not the problem though. I'm not sure which part of my original
e-mail led you to think it is.
It's an interesting discussion, though, and I would love to see a global
and final conclusion on the matter, not just applying to XHR or Atom.
--
Anne van Kesteren
<http://annevankesteren.nl/>
<http://www.opera.com/>
