[Indic-computing-users] Appeal to Translate GNU/Linux in Indian Languages

[jnshah <jn...@bo...>]

Appeal to Translate GNU/Linux in Indian Languages ( Marathi , Hindi, 
Gujarati to start with)
( Come , it is time for everyone to join the party)

By Jitendra Shah

(Following is a rolling document to which all can add. . This is meant 
to be an introduction and appeal  for newbies (typically computer 
literate , students of computer professionals working in ther than 
language technologies)

Who are we , Why you ?
We are a volantary group and intent upon making GNU/Linux available to 
 all in indian languages. We are a mix of language enthusiasts, 
technical zombies, down to earth pragmatists, high-in-the-sky idealists  
and researchers. We seek support from all those who may want to see 
their mother-tongue  powering the computer revolution that is shaping 
our knowledge world.
 

What is new?:System Level  Support
Till recently the world of indian language writing sysytems was 
restricted . We had no basic support from the Operating sysytem as it 
was based on ASCII.  There were two options; Somehow accomodate within 
the limit of 256 encodings or simulate wider width . It was (and is) 
possible to devise ingenious  methods to write and show indian language 
on these (otherwise unfriendly operating)  systems by emulating broader 
width to accomodate more characters(aksharas) . In fact many 
applications have been used extensively on that basis.. However they are 
not truely multilingual but support only one language or two say 
Devanagari and Devanagari+English Some go to the extent of a few indian 
languages. Like Akruti/APS fonts/C-DAC's ISM etc.
Unicode (www.unicode.org)
With the advent of unicode (under pressure from China/Japan/Korea? They 
use ISO standard which is currently co-incident bwith unicode. Unicode 
is better supported and more dynamic.) , now the operating systems have 
a native support for multilingual writing systems. Naturally, GNU/Linux 
was quick to change over to unicode support.
.Internationalisation (i18n) (www.i18n.org <http://www.i18n.org>) and 
Localisation(l10n) (www.l10n.org <http://www.l10n.org> )
 
Our object is to make them available in local indian language on 
computers using Free software. One way of doing this ( followed so far 
by ingenious vendors) is write applications in indian languages. Problem 
is these would remain lagging behind in features. They may also remain 
incompatible with the global documents unless compatibility is 
constantly maintained. Thus it would be better if one could make 
available programs that are globally popular , quickly available in 
indian languages. Internationalisation and globalisation are the GNU ( 
who else?) initiatives to standardise the process and simplify the same. 
Instead of rediscovering the wheels, the logic of programs is maintained 
as it is and interface part is so modularised as can be swapped ( that 
is interbnationlaisation) and only the interface is changed to 
appropriate language/script (that is localisation).. Thus  instead of 
making each application indic language enabled seperately what we can do 
is provided is a system to quickly localise applications and manage 
changes for updating in a centralised manner through a purely voluntary 
(non-profit) effort. .
Translation
Translation is not a purely technical job. However background of 
computer world is an important factor in good translation. While we dont 
want a hardcore  computer programmer do spend her time and skill on 
translation, we do feel that a combination of those with computer 
programming/application background with those with cultural/language 
background will be able to reach a better translation.
It must be noted that translation may remain an iterative process for 
some time and even a few dialects may emerge. As an example for 
programs, people have come up with "Anuprayog", simply program written 
in devnagari script, and "Karyakram" as alteratives.  a mechanism for 
ariving at a consensus or for managing diversity is required. As a 
principle , I personally believe that words which have a parallel in 
common culture will be more easily accepted. However when there is no 
similarity in cultural practice , a new word may be coined . There is 
bound to be a difference in sanskritised or anglicised or 
"Khadi-boli/maraaThMoLi" inclinations..There is a need to give these 
variations time to come to consensus . That is the challenge to our 
project. It is utmost important that we have a few motivated and yet 
accomodating language stalwarts to carry out this task as experts.  
These will have to work closely with those who can spend time and give 
out translations/ We can get some junior level (rough-cut) translation 
from young enthusiasts who are ready to accept leadership of editors and 
experts and would also be able to use tools and aides for translations
Thus we have a hierarchy  
1.    Translators
2.    Editors(Language mid level Experts)
3.    Expert

The translator will not be required to do the job of typing but only 
required to translate the text.(typing support will be provided, at 
least till translator is adept herself) The Language Expert will than 
verify the translated texts.The expert will look over all the process 
and input the translated strings.
Modes of communication
We can do the communication through emails or on paper( hard copies).For 
all those not connected to internet we can distribute the strings to be 
translated through the paper medium
Input Methods
For the input methods we can use the following editors as standard (or 
any other which gich gives UTF-8 format)
1.    gedit:- the default editor of gnome in linux has a very good 
support for writing  indian languages
2.    yudit:- it is one of the editor which works very well in both 
M$windows and linux platform. (www.yudit.org <http://www.yudit.org> ), 
about 3 MB. Add raghu font to fonts directory.

Translation Aid:

kbabel:- it is application of kde which is used to input the translated 
string and provides a method to build up a database for the translated 
part. The problem is that display is not good. This is a management tool 
for managing authority, fuzzy translations, to know how much work is 
done, to provide dictionaries etc  However not yet (june 2003) support 
for hindi input.You have to cut paste from gedit or yudit. You can see 
hindi but not write.

gtranslator: Similar to kbabel.

Dictioneries : Openoffice, IIIT(Hyd), IndLinux
Anuvadak: A program on the net (www.parixa.com) that help translate on 
line , keep uptodate translations, provide dictionerries etc.( 
www.parixa.com/anu )

 Programmer's Aid
    gettext : xgettext to extract strings for translation and making 
files translation work that can be used by kbabel , msgfmt a tool that 
converts translated file into a binary fle that can then go in 
appropriate file. If found, the program will appear in this language 
merely by changing the "LOCALE" . (LOCALE is a parameter that spcifies 
for a given script info such as date format,currency label, etc.)
 Tasks:
GUIs: Gnome and KDE : why you should bother ? :The major tasks are to 
translate sentences (strings) from menu/labels and messages in popular 
and basic programs.  As you know , gnu/linux uses some popular GUI 
programs which are associated with their libraries of programs. This 
GNOME is based on GTK libraries and KDE is based on QT libraries. (The 
formaer is written in C and latter is in C++.) While most programs work 
with both GUI, the re are some which work only in one or the other. 
While GNOME is a free software in full sense, QT is free only for 
non-commercial use.
 What to translate ? It may be noted that what one sees on the screen as 
labels and messages in any language come from a number of programs. For 
example the file/edit/help etc one sees in main menu bar come from a 
file while working in GNOME and from another file when working with KDE. 
In order that one sees one's language while running a program, one needs 
to translate sentences from many programs.  Thee sentences are separated 
and put up on the websites as dot_po"(or  '.po' ) files.
We propose to distribute this translation work to different teams and 
organise the work so that efficiency is maintained.
We shall also keep many resources on the website ( address to be announced)
A status report and contributors' contact will also be pot up.
This will facilitate translation as all current translations already 
done will be available for next translation.
A paid assistant is proposed to be employed for the assistance and 
organisation of translation work. Thre will also be a team of volunteers 
with some paid assistance for maintaining the website.
Please contact the undersigned or named persons below if you are 
interested in helping the translation work.

Languages
    While 50% of work of  major programs is complete in Hindi,we propose 
to begin with Maraathi.  But we propose to encorage other languages like 
hindi and gujarati, where almost no work has even started.

Marathi translation incubation site
Following site and group have been created as a starting point  for 
initial interaction , The space is created by the active support from 
IIT Mumbai students and Faculty. On maturing the same will be uploaded 
to sourceforge or similar site.
Swapnil Zarekar, a comp science student , has volunteered to master the 
site and monitor the group..

http://www.cse.iitb.ac.in/~swapnil/marathi

mar...@ya...

For further contacts

jitendra shah  ( jit...@vs... )

Other contacts : Venkatesh Hariharan ( ve...@vs... ) , representing 
IndLinux
Dr Nagarjun  of FSF ( nag...@hb... )