Re: [WMASTERS] Re: Old Orthography, response to Kalyan

["Dr.K. Kalyanasundaram" <kalyan@igcsun3.epfl.ch>]

Dear Bala:
> How cast in stone is UNICODE 2.0 as it applies to Tamil script? Is it not
> possible that our discussions would affect a revised version of UNICODE 2.X
> ? My limited knowledge of UNICODE Tamil deliberations suggest that it did
> not get much (not anybody's fault mind you) airing.

Let me first summarise the current state of affairs before I give my
2 cents worth opinion.

If I am not totally wrong, there is some consensus that the proposed
encoding scheme should be such as to provide compatibility with the
current version of UNICODE 2.0. This was one of the main conclusions
reached at the last TamilNet'97. The reasons for this are i) to keep
going
now, ii) ensure that it can happily co-exist with current unicode and
iii) make the transition to unicode at the latter date smoother.

 It is my understanding that ISCII of Govt. of India in reality is a 
subset of UNICODE  (UNICODE takes the ISCII proposals as the
keyinput). Philosophy and implementation approaches are the same.

As far as the implementation of UNICODE in real world, Microsoft does
distribute free unicode fonts carrying ca. 800+ glyphs covering about
six or eight european languages but is not yet comprehensive to 
include any indian language glyphs (incl. Tamil).
But Apple recently introduced an indian language kit in India for Mac
based on ISCII and associated INSCRIPT keyboard layout. 
This package includes only Hindi and Gurmukhi (no Tamil).

It is very likely that soon (within a year or two ?) we will have
unicode
fonts containing indian language glyphs available. 
It is also likely the present UNICODE 2.0 as applicable to Indian 
languages (particularly Tamil) may undergo revision, given the stated 
resentment of TNC with the existing ISCII standard for Tamil.
The revision of these standards are slow processes and they do occur
once in five years or so. ISCII was revised in 1988, 1991 and 
now in 1997.

Thus the UNICODE is an emerging scheme for multi-lingual 
word processing but not fully established, particularly for Tamil. 
Trying to have compatibility with a scheme which itself is not firmly
established is a tricky business. If we can define our 8-bit standard
not 
as a rigid box but with some flexibility, there is no reason why we
cannot
launch our standard right away. If some of the unicode characters are
redefined, it should be possible to rewrite the mapping table to conform
to the revision. If we are comprehensive in our approch in dealing with
basic tamil characters/glyphs, I do not see any problem. 
We have to keep going with what we have.

Instead of waiting for Microsoft or Apple to take the lead, why cannot
we launch ourselves a unicode conformant font, software containing
English, Tamil and few others ) and have some field testing
concurrently?
Can we not adopt the same approach as that of Microsoft and Apple but
tailored to Tamil? What are the technical hurdles if any?

Kalyan

PS: I thought we all agreed to go first for a scheme with compatibility 
with the existing proposal (unicode 2.0) and take up the unicode scheme
itself for grinding later and make our suggestions based on what filters
out of these two exercises.