Re: Numeric Char Ents in 2.0 draft

Terry Allen (
Fri, 31 Mar 95 10:10:20 EST

Roy responds to me:
> So why isn't the language "character octet references are not
> dependent on the character set encoding of the document" bogus
> and to be deleted?

Well, we can't just delete it -- a suitable (and conformant) alternative
should be proposed.

My suitable and conformant alternative is to say nothing about a matter
it should not be necessary to discuss. If we sanction people writing
docs in ISO-8859-[not-1] with numeric character entities referring
to 8859-1, we will be encouraging the production of broken SGML
(and hence broken HTML), which we will later be called on to
support, even though we can't do that in SGML.

As for Larry's and Gavin's suggestion that the HTML be preprocessed
to change the character set encoding to something not indicated in
the SGML declaration associated with the document (and how to construct
that SGML decl for non-ISO-8859-1 charset encodings is work still
to be done; I will welcome Gavin's paper), let's let well enough
alone. The ISO-8859-1 charset repertoire is available anyway,
why create nonconformance?


Terry Allen  (   O'Reilly & Associates, Inc.
Editor, Digital Media Group    101 Morris St.
			       Sebastopol, Calif., 95472
occasional column at:

A Davenport Group sponsor. For information on the Davenport Group see or