Re: Revised language on: ISO/IEC 10646 as Document Character Set

Glenn Adams (
Tue, 9 May 95 16:20:34 EDT

Date: Tue, 9 May 95 15:23:37 EDT
From: (Dan Connolly)

> Glenn seems to agree that the charset does not have to be a subset
> of 10646.

Er... well... no. Glen says you can have characters in your document
that are not in the document character set. I don't really agree, and
it seems that neither does James Clark nor Charles Goldfarb.

I think you misread my discussion with James and Charles. What we were
talking about there was *character numbers* not *data characters*.

What became clear from our discussion was that every *character number*
must be in the described in the document character set; otherwise, it is
not a legal character number.

That is different from saying that every *data character* must be in
the document character set. For example, you can use SDATA entities
to represent data characters which aren't in the document character
set (and which may have no character number); and you can also use code
extension techniques in the same manner.