Terry Allen <[log in to unmask]>:
>One problem might be that this is an SGML error, if IDEO-SP isn't
>in the syntax charset, or is but isn't defined as a SEPCHAR. If so,
>the instance is invalid and an error checker (or, what a thought,
>SGML parsing) will catch the error, and the author will correct it.
>Another problem is that authors won't bother to check for errors,
Which, of course, is what will happen most of the time. Validating
documents is a Good Thing, but is not generally done in the real
world. Especially since people use Netscape extensions all the time,
and there is not even a real, complete DTD for that.
>and browser developers will feel they have to tolerate it.
Why, then, not simply allow them in the spec? We have a (rather
small, admittedly) problem, with an extremely simple and easy
solution, both conceptually and implementation-wise (remember, no new
name characters). Standards-conforming, moreover. Why decline?
Gavin Nicol <[log in to unmask]>:
>There is an alternative: we would map the wide versions of spaces and
>characters to the ASCII form.
I don't think SGML has provisions for that.
>That said, I think putting in SEPCHARS is good, but not zenkaku
Exactly my opinion, and exactly what is currently in the draft.
François Yergeau <[log in to unmask]>
Alis Technologies inc., Montréal
Tél: +1 (514) 738-9171
Fax: +1 (514) 342-0318