IE5.0 does not conform to RFC2376

Rick Jelliffe ricko at
Mon Apr 5 20:43:51 BST 1999

From: Chris Lilley <chris at>
 >I was meaning autodetection in the sense of reading a whole bunch of
>text and making assorted guesses based on frequency analysis and the
>like. In other words, automatic detection based on unlabelled content.
>believe that this is a bad thing, because there is always the
>possibility (quite high) of hgetting it wrong.
>The encoding declaration, on the other hand, is not autodetection in
>that sense, it is a label. A very small amount of autodetection has to
>be done in order to be sure that the label has been read, that is all
>(ie, is this UTF-16 or is this an encoding where ASCII is represented

In academic material, this is called "codeset announcement" (e.g. in the
OS from the mid 90s). The term "autodection" does give people the idea
guessing is involved. This is important, because if developers think
that autodetection
means guessing rather than codeset announcement, they may be tempted to
guess encodings
without alerting users that something seems strange: this would not be
satisfactory for
important documents.

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