At 09:05 AM 1/10/97 -0500, Sandeep Sibal wrote:
>Do try it out. I'm curious to see how easy is it for people to
>read/write Devanagari in this ligatureless half-consonantless script.
>
>I'll leave my own experiences for later. Let me hear from
>you first.
You have now provided crucial information that allows me to understand
your situation. You are trying to use software which assumes a one-
to-one corrrespondence with characters and glyphs. Thus you are trying
to find a way to use this software with a coding system which requires
more intelligence. Why didn't you say so in the first place?
In any case, I completely sympathize with you. I also sympathized with
users who painstakingly entered Arabic both as presentation forms and did
so backwards (on left-to-right) systems. Fortunately these days are long
past with Arabic -- it's not surprising you are unhappy with the status
quo. This reminds me of a time about seven years ago when I tore out
the guts of Motif 1.0 pertaining to character representation and display
and replaced it with a multilingual toolkit which I wrote that supported
a complex character to glyph mapping. I gave a demo of it to an OSF
developer, entering and display a multilingual string composed of Burmese,
Thai, and Japanese. Though this person was nominally trying to introduce
I18N to motif, what I had demonstrated appeared to be beyond this person's
understanding. I painstakingly explained the need in complex scripts for
a more general mapping between characters and glyphs, and that, since an
input method technology was being developed for X11 at that time, that a
similar output method technology was needed for Arabic and Indic scripts.
Unfortunately, my words seem to fall on deaf ears. And here you are today
with precisely the same state of affairs!
That being said, the Unicode Technical Committee has had a long-time
commitment to keeping presentation forms out of the standard to the extent
this is possible (given compatibility with existing standard coding
systems). I don't believe there is much chance this will change in the
near future, particularly with the Indic scripts.
The only hope for you I should imagine is to help convince prominent software
companies to support the Unicode rendering model as needed for Devanagari. I'm
sure when the Indian software market becomes interesting, that there will be
a mad rush to implement it. For my part, I have recently converted our Web
client technology (Spyglass Mosaic) to employ Unicode (UCS-2) in all aspects.
I intend to integrate a complex character to glyph rendering model some time
this spring or summer which will support Indic scripts. I'll be sure to
announce it when we have something you can download to try out.
Regards,
Glenn Adams
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:33 EDT