IE5.0 does not conform to RFC2376

Chris Lilley chris at
Sun Apr 11 03:27:53 BST 1999

John Cowan wrote:
> Chris Lilley wrote:
> > MIME actually need not have those constraints; *email* has those
> > constraints (although increasingly it does not, in practice). HTTP is
> > always 8-bit clean. I agree that the MIME RFCs have steadfastly tried to
> > pretend that MIME is an email-only thing.
> The requirement for a charset with text/* documents (whether minimized
> or not) has nothing to do with email, although the fact that a
> minimized charset parameter means "US-ASCII" surely does.

We agree here

> A charset must be specified for interoperable text (as opposed to
> application-specific formats) because text means nothing if you
> do not know the charset,

We also agree here

>  and no text-using application can fail
> to either decode, guess, or simply presume some charset.

And here. Wonderful. So, given some random text page, with no other
labelling, then yes, you would need to guess. And there are ways to do
so. No way to tell if you guessed right.

With XML, having made a determination of what the encoding is, it can
instantly be verified, because the encoding declaration is required. So,
the label is there all along, is absolutely required to be there
otherwise it is a fatal error, and all that is needed is a little
bootstrapping. This is clearly not the same thing as the previous case,
at all.


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at
Archived as: and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at

More information about the Xml-dev mailing list