lynx-dev Multiple character set documents (was: Lynx bookmark display problems)

[David Woolley]

> > > Also, does HTML 4.0 support multiple document encoding within a single
> > > HTML document?  For example:
.....
> > 
> > The correct way of doing this is to use a character set that can cope with
> > all the languages used.  This could be one with escape codes to switch
> > encodings, but the real world of browsers means that it would almost 
> > certainly have to be UTF-8.
> 
> And more probably, some programs/drivers that translate codepages of fonts
> from/to UTF-8 on the fly.  Otherwise you will need to install a lot of
> fonts.  :(


I think you may have highlighted a problem in cascading style sheets here,
at least for CSS1 (I haven't had time to properly read CSS2).  The assumption
of the font support is that you are working in a multiple font, single
language environmnet, typical of marketing type material.  You'd be
OK in a multiple language, information only environment, but, if you
mix European, Indic, CJK, etc., etc., in the same document and you also
want to be artistic with fonts, the fact that fonts are tied to scripts
(I doubt that there is a Times Romon Devanagari, and almost certain there
isn't a Times Roman Kanji) means that you have to switch fonts, when you
switch languages.  This seems an unecessary coupling of physical aspects
with content.

If you are intested in this area, could I suggest that you look into
CSS2, and, if not too late, and you agree that this is a problem, raise
the issue with W3C.

Note though that neither CSS nor HTML cares about problems arising because
of limitations in the electronic forms of fonts installed on a machine;
if some combination of implementations of Times Roman can represent all
the characters, but not any one single one, it should be the browser's
responsibility to mix and match, not that of the page author.