IE5.0 does not conform to RFC2376

Rick Jelliffe ricko at
Sun Apr 11 12:40:15 BST 1999

From: John Cowan <cowan at>

 >Remember that I spoke of a client that *cannot* understand EBCDIC
>(because it has no conversion table, say), but perhaps has the 4-byte
>heuristic from Appendix F.

Just a side comment: In Appendix F of the XML spec, it is not encoding
declarations which are non-normative, it is the particular (details of
the) algorithm for interpreting them.

It is non-normative, because (apart from anything else) it does not
attempt to completely cover the full range of possible IANA character
sets which could be used with XML: it demonstrates how a processor uses
the encoding declaration for some common character encodings.

In just the same way as the term "auto-detection" fails to indicate that
it is not guesswork, "heuristic" may also suggest to some to some that
the Appendix F algorithm involves guesswork. So I certainly prefer
"algorithm" to "heuristic".

No matter what kind of  system for representing differing character
encodings (internal, external, defaulted, BOMed, etc) , there is the
possibility that encodings will go astray. I think this is a fault of
operating systems: without a way to reliably indicate charset and so on
for documents, webservers do not have anything to query in order to
source the information: I guess that UNIX is the worst player here--Macs
have the resource fork and the business operating systems have
registries or databases (AS400 can register text as 8-, 16, or 32 bit;
however, I don't think it has a way to specify character set beyond

Rick Jelliffe

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