Re: Transliteration scheme(s)

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

Dear Vasu/Sujatha/Muthu:
I am happy that we are picking up some momentum in
discussing various avenues open for possible standards.
Even at the risk of repeating myself, I would like to state that
choosing a standard scheme for transliterated tamil is URGENT
and CAN BE DECIDED EASILY without too much controversy. 
I am happy that you all think this way.

(For those of who are not familiar with various transliteration
schemes that are being discussed, I have a webpge - part of
my singapore talk presentation- that shows some of the most
popular schemes including the Library of Congress scheme (LC)
with diacritical markers used by, of course LC, UC Berkeley, 
Roja Muthaiah Library of UC Chicago, Inst. of Asian Studies
 (incl. comprehensive Tamil Lexicon series), ....). Without a
specific font it is not possible to reproduce them here. 
The URL  for the gif image is  
http://www.geocities.com/Athens/5180/trnslt.gif

I prefer that this topic be addressed concurrently with the
choices for the tamil character glyphs that can constitute the
font encoding scheme. As I said in my earlier posting, we 
need to have hardly 4 marker signs (!) and these four can be
readily accommodated in the font encoding scheme I gave.
One integrated font that allows typing tamil using tamil scripts
and transliterated text using standard scheme with diacritical
markers can go a long way in standardising tamil computing.
We do not have to have two different fonts, one for each task.

Of course, any one can use a simple bilingual (tamil/roman)
 font to have transliterated tamil text according to a plain 
ASCII scheme. So we do not have to do anything special
other than deciding on if we want to adopt a standard for this.

So the specific questions we need to answer are:
i) do we go for two transliteration schemes - one with diacritical
markers and one based on plain ASCII that can co-exist  (like
keyboard layouts) or decide on one and only one standard? 
In my opinion, plain ASCII scheme would be inevitable 
evil to live with for the following reasons: 
a) if transliterated form of keyboard input has already been 
accepted as one of the standard options for input (yes yes,  it
has been decided!!) , it would be better if we can agree on 
one translit. scheme to be used in all tamil software packages. 
Obviously this can only be a plain ASC II scheme.
b)  I cannot see any other way tamil texts/words quoted in newsgroups
that have been part of the Internet, one that will go more important
as a means of communication/info. exchange in the future. 
It may be too much to ask for each and everyone to have the 
TN 8-bit bilingual, integrated font to post anything in tamil.
(I guess I have already expressed my stand/cast my vote but
I am open to discussions/debate on this point)

ii) for a scheme with diacritical markers, the clear and obvious
choice would be the LC scheme used widely by indologists.
It may be a wise decision to specify adopt this scheme. I am
not if any other candidate would make a serious claim. If there
are others, please bring them up here so that we can have give
a fair deal to all prospective candidates. 

iii) for plain ASCII scheme (if we are going to accept one
standard for this to co-exist!), there are many candidates.
Here the relative merits of 'linguistic' (Vasu) and 'human'
(Muthu) needs are to be considered before the choices are
narrowed down. In an earlier posting, I have indicated my
preferences in this context - those based on Adami/Madurai
schemes. It would be better if Vasu, Muthu and others
specify their preferences so that we can discuss this issue
in some detail.

With best regards,
Kalyan

--
*******************************************************************
Dr. K. Kalyanasundaram,            |
Institute of Physical Chemistry,   | Tel: 41-21-693 3622 (off)
Swiss Federal Inst. of Technology  | Fax: 41-21-693 4111
CH-1015 Lausanne, Switzerland      | Email:kalyan@igcsun3.epfl.ch
*******************************************************************