Is CSX the best solution?

[kellner at ue.ipc.hiroshima-u.ac.jp]

John Richards wrote: 
>Before I came across the CSX convention, I had created a Sanskrit TTF
>format font myself, on a quite different principle. Whereas the CSX
>convention uses the extended ASCII range, thereby interfering with the
>"normal" use of the primary ASCII characters less, I had deliberately
>chosen locations in that primary range, using the characters that are
>certainly not needed for Sanskrit text - @#$%^&*+={}[] etc. I have never
>found an instance when I have needed these in a Sanskrit text or
>quotation!

Simple question: What do you do when you write a paper where you use both
English (or any other non-Sanskrit-language) and Sanskrit? I guess you have
to change fonts. Because I don't like to change fonts, I prefer CSX-encoding. 

I agree in that it is cumbersome to type all the ASCII-numbers in
search-menues. However, I believe that there are Windows-utilities which
allow you to compose a Windows-keyboard of your choice, which lets you
insert characters in search-menues the way you please (I have heard about
such miraculous creation, alas, not found one which likes Japanese Windows
yet). 

What I find more disturbing about CSX is that word processors do not
recognize extended ASCII-characters as characters in the first place. I.e.
when you have a diacritical mark within a word, the word-processor will
treat the word as TWO words (which is especially endearing when it
automatically moves the phoney second word into a new line or stubbornly
resists your attempts at hyphenation...). This is speaking for Word for
Windows, but as far as I remember, WordPerfect behaved in the same
mysterious ways. 

Birgit Kellner
Institute for Indian Philosophy
University of Hiroshima