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

Gavin Nicol (gtn@ebt.com)
Thu, 11 May 95 19:48:33 EDT

>Unicode for HTML is being put forward as a practical engineering
>solution for most cases, and we shouldn't worry too much about
>marginal ones. Over in MIMESGML these are exactly the issues
>we have to take on, because there's no way 10646 is going to
>be declared as the document charset for SGML over the Net; it will
>be one possibility, but only one among many.

This is exactly right. I proposed using ISO 10646 as a 95% solution,
and to provide a simple model that enables multilingual processing.
MIMESGML is a whole different ballgame.

Given that other (perhaps many other) document formats will be
accepted in next-generation browsers, I think worrying about marginal
cases overly much, is a waste of time, and will just lead to
complications we don't really need.