Re: character sets

Mike Piff (
5 Jul 94 11:57:32

Nathan Torkington writes:
%%>Could someone please refresh my memory about the ways and means of
%%>inserting special characters (eg s-cedilla) into HTML documents. As
%%>this character isn't part of the ISO Latin-1 set, how am I to
%%>efficiently and sensibly code my Romanian pages?

Umm! You *could* try doing it in (La)TeX, then translating it into HTML,
which is what we mathematicians are supposed to do(?) to get all our
special symbols onto the web:-)

The solution is quite easy! Just convert each character that doesn't work
into a bitmap, then stick the bitmap in at the appropriate place;->

Seriously, though, such whimsical solutions apart, I guess there are real
problems here that need to be addressed. The TeX world is facing the same
problems in adapting to non-English languages; accents are not so much of
a problem, as TeX markup allows any accent on any character. Specialized
mathematical fonts and arabic/chinese, etc, fonts, mixing left-right with
right-left/up-down, etc, present real difficulties. Plus the fact that
many of the symbols mathematicians use just aren't available in the
standard fonts as used by the likes of Word and WordPerfect. (I believe
there is but one commercial Postscript font package that includes them all,
apart from the ones we may manufacture in the future---you see,
the alphabet itself is extensible---and that is LucidaBright+LucidaNewMath.
Thus, we mostly use the Computer Modern fonts and TeX for our work.)

TeX, on the other hand, is completely paragraph-breaking and page-breaking
orientated, and totally unsuited to displaying stuff in a continuously
scrolling window. Also, who wants to see footnotes, floats, marginal notes,
etc, in a hypertext environment?

So where does this leave the involvement of mathematicians in WWW?
Are any mathematicians involved? Do the WWW people appreciate the complexity
of mathematical typesetting? Is the TeX markup going to be re-invented in
HTML in its entirety, or are the needs of mathematicians to be ignored, as
they generally are in commercial word processing packages?

%% Dr M J Piff, School of Mathematics and Statistics, University of %%
%% Sheffield, UK. e-mail: %%