[Indic-computing-users] patiala seminar summary

[LinuxLingam <lin...@bh...>]

dear all,

just back from the seminar/talk i conducted at the Thapar Institute of 
Engineering and Technology (TIET), in patiala, punjab.

the talk was held on friday, 18th april 2003.
i think about 70 participants attended.

initially scheduled for 4 hours, during the discussions in november, it came 
down to 2 hours, and then finally, due to some timing confusions and 
overlaps, i only got about 1 hour and 15 minutes.

which was sufficient.

topics:
+ the state of indian languages, and how they are rapidly dying.
+ the 'indian language mess' in encodings, standards, technology.
+ a brief overview of unicode.
+ a brief overview of opentype.
+ how/why both these are important and suitable for indic languages.
+ some limitations of them.
+ a hands-on demo of indic support under winXP using raghu font and hindi.
+ a first-look at pfaEdit under gnuLinux and how it can be used to design otf.
+ beyond indic font design: a punjabilinux (punjalinux?) gnuLinux distro, 
indic language technologies such as OCR, text-to-speech, speech-to-text, 
search engines, translation engines, and other similar projects.

+ a call for volunteers, supporters, enthusiasts, to jump in and take it into 
their hands to solve the 'indic language mess'.

+ for the linuxnewbies: a brief talk on gnulinux, knoppix, and what are the 
freedoms that the gnu GPL delivers. and the importance of gnuLinux-based 
solutions for indic languages.

results:
+ several students and faculty volunteered to devote time, resources to indic 
language development, and towards development of other projects that demand 
software skills.

+ the faculty members promised to create and nurture more academic projects 
for students, focussed on indic computing. [for example, undergrad, grad, 
postgrad, and ph.d projects]

beyond the seminar:

+ Dr. Gurpreet Singh Lehal, an assistant professor at TIET, and chief 
Investigator, resource centre for indian language technology solutions 
{project funded by Min of IT} is an amazing resource, as follows:
he has done a Ph.D in his work on a Gurmukhi OCR. he demoed it to me, and i 
was quite impressed with its rather fast processing speed, and the way it 
handles tough images with noise and rotation issues. the OCR however, cannot 
handle really challenging stuff like gurmukhi scriptures on old-stock paper, 
but is getting there.
he heads a fantastic laboratory at TIET on indian languages, funded by the 
Min of IT, he confirmed the TDIL. under the aegis of this lab, apart from the 
gurmukhi OCR, alongwith his team, research students, he has also spearheaded 
the creation of an impressive website that teaches newbies how to read and 
write gurmukhi. check it out at:
punjabirc.tiet.ac.in

the lab has also developed gurmukhi fonts, using fontographer. these are ttf 
fonts, *not* based on unicode. Dr Lehal pointed out several critical issues 
with unicode that are major stumbling blocks, and which have forced him to 
follow his own, non-standard, encoding, based loosely around ascii. i 
explained to him how unicode could greatly benefit the project, and how 
issues of gurmukhi could be easily tackled by communicating with the unicode 
consortium. he is open to the idea and wishes to explore this further.

they are also working on a text to speech engine, a speech to text engine, a 
gurmukhi dictionary project, and several other exciting projects.

he mentioned by May 31st, his work at the lab is coming to an end. thus he 
finds an urgent need to have an indic-language regional workshop on opentype, 
unicode, [much like the bangalore workshop i feel] organised, asap. i 
promised him we are working towards that. perhaps we must start working on 
this in earnest, probably holding it in delhi, at sarai, or at TIET, or 
wherever. a proper, 3-day workshop, make it 4 if we can throw in hinting.

as usual, the font quality developed by the lab, though functional, is not 
professional quality, lacks hinting, and has various issues.

i spent some time post-lunch at the lab, addressing queries from Dr Lehal's 
team, on various issues on font design, the placement of BCPs on the curve, 
hinting, the use of ligatures, metrics tables, and several other technical 
issues. this further heightened the need for otf with its much superior 
approach to the issues.

okay, all the stuff done by Dr Lehal and his team is entirely based on win. 
all the licenses to their work are proprietory, nonfreedombased, and they 
wish to commercialize their work. Dr Lehal categorically mentioned that if 
given free (as in free lassi or free buttermilk) their work would not be 
appreciated.

they are also working on a gurmukhi wordprocessor /editor, at which point i 
mentioned to them OpenOffice.org. they were not aware of the software suite, 
so i requested the network admin who is gnulinux-savvy, to download and 
install a copy of the software. briefed them also on the work of bharteeyaOO 
and how they are localizing Openoffice for indic. shared how the opensource 
community works, shares, grows, and how invaluable the contributions of the 
opensource, freesoftware world, is, to the growth of such things. Dr Lehal 
was quite impressed with this, and agreed to install gnuLinux, and start 
testing and using it, since he is already familiar with unix.

am happy to note that Dr Lehal is always open to new and fresh ideas, and 
wishes to learn from the community and contribute to it. i request others to 
show and share.

Dr Leha's email: gs...@ma...


+ Maninder Singh, Network Admin, TIET,
is a veteran gnuLinux user and adept at both win and gnulinux. incidentally, 
TIET has dual-boot machines strewn across its campus, has even a high-end 
parallal computing lab, and loads of other impressive resources.
i was impressed by maninder's high skill-levels with gnulinux, and yet rather 
disappointed by the lack of deeper awareness of gnulinux among the students 
and faculty members. discussed this with maninder.
he immediately decided to launch a GnuLinux User Group in patiala, at the 
TIET. that's great news. i request all to support him whichever way you can.

i briefed him about the commendable work done at IndLinux, by Bengalinux, and 
other localization efforts. He also promised he'd love to head and work a 
punjabi linux, and we joked we could call it PunjaLinux or something. :-)
meanwhile, students who attended the lecture also volunteered to work on a 
Punjabi distro.

Maninder also discussed with me possible topics for a Ph.D he wished to work 
on. suggested some ideas, related to indic, which he found quite enthralling. 
he also dicscovered how he could further delve into the research. however, i 
would rather leave it to him to announce his work, once he has finalized and 
formally enrolled for his Ph.D.

+ Seema Bawa,
Head of Department, Computer Science 7 Engineering Dept, and Network Manager, 
who is also pursuing her Ph.D
was quite motivated by the talk. as the H.O.D, she promised more projects 
will be initiated among the student community towards Indic computing, 
personally volunteered for any project she may find interesting, and also 
mentioned she would like me to conduct a similar talk/seminar at Gwalior. She 
is to get back to me with details. Am also going to forward this email to 
TIET, and hope at this point, she responds with the contact details.

Misc:
I noticed that through the day, students tend to converge to the newspaper 
stands near the entrance and reception of the computer lab, on the ground 
floor. I would keenly observe, of the english, hindi, and punjabi, 
newspapers, do students read the regional newspapers? and found an equal 
number of students did read the regional papers. so that's a healthy sign.

as usual, i have taken some photos (but not enough) and once the roll is 
developed, will courier a cd as usual. perhaps vijay could then publish the 
photos to the indic website.

acutely felt the need for a printed brochure, or a pdf, of indic computing 
consortium. it is difficult ot explain what ICC is to people. we also need a 
similar website, and must upload pictures and summaries and developments as 
they happen. so that people can distill how ICC is the umbrella to all indic 
computing developments, from fonts, localizations, software projects, etc.

please get back to me immediately with your responses, so may forward them to 
patiala.

[ and no, i did not try any patiala pegs, since i don't drink! ]

:-)
LL