Localization Newsletter Issue 7, Vol 1 ( October 15, 2004 )

***********************************************************
Localization Newsletter Issue 7, Vol 1 ( October 15, 2004 )
***********************************************************

Editor Speak
------------
On Sept 18-19th, the Indic Computing Coordination meet was held where
members of all localization groups in India met to share their
experiences and chart out a road-map for future of Indic computing.
Participants were all language team coordinators, team members ,
contributors and organizations working on related lines. A report is
coming up at http://www.indlinux.org/wiki/index.php/IndicMeetReport.
For a glimpse of all the participants see
http://www.indlinux.org/doc/indicteam.html .

In this issue we have Oriya group speaking about their localization
efforts. Oriya localization effort called Rebati project, hosted at
http://oriya.sarovar.org. Also announcing next release of
IndLinux-Hindi 0.9 (Unnati) with updates for KDE Hindi.

As usual, we are looking for volunteers who will provide us with news
about Language Workshops, Usability Tests and general L10n news.

Feedback on content, presentation, suggestions for newsletter welcome
at <feedback at indlinux dot org>


News Bits
*********

IndLinux Hindi 0.9 (Unnati) released
------------------------------------
IndLinux Hindi 0.9 (unnati) is released. It contains updated KDE
translations Hindi translation update for KDE is released. This
includes updated translations for KDE, including the work done at KDE
Hindi translation review workshop -
http://www.indlinux.org/wiki/index.php/ReviewWorkshop


GNU/Linux to power voter list search in Mumbai and Thane this assembly
elections More than 1.2 crore voters of Mumbai (main and suburbs) and
Thane districts of Maharashtra will be taking benefits of GNU/Linux
powered search engine developed by Indictrans team. The Chief
Electoral Officer (CEO) of Maharashtra have put up a voter list search
engine for these 3 districts for the upcoming assembly elections on
their website.

The main features of software used for this search engine include:
    * Full Text Search can be made by name and/or address OR Photo
Identity Card No.
    * Search results include Polling booth name,address and Other info
such as serial No. and Part No.
    * completely based on GNU/linux
(Knoppix3.4+Apache2+PHP4+PostgreSQL7.4.3)
    * original ISCII data in DBF format was converted to unicode and
imported into PostgreSQL
    * complete source code of the search engine is released under GNU
GPL on savannah.nongnu.org (see vl2-libs) copyrights belong to Swapnil
Hajare & Sreepathi Pai
    * powered by intuitive similar sounding search technique called
"souindics" developed @ Indictrans
    * Although the data itself is in Marathi search can be made in
Marathi as well as English
    * Search results can be in Unicode Marathi/modified ITRANS English
OR in legacy fonts like Shusha,DVBWTTYogesh.
    * Results are also available in Dynamic font
AkrutiHnbDynamicYogini (courtsey: Cyberscape Ltd.)
    * Search time less than 15sec on entire data on an average
    * Search results can also be obtained in the form of voterslip as
PDF with embedded Shusha font which can be viewed in any PDF viewer
without installing font


Critical bug in OpenOffice.org Calc affecting Indian language
localised builds
---------------------------------------------------------------------
---------
Hindi and Kannada builds of OpenOffice.org Calc crashes whenever the
user tries to insert Functions. Gregor Hartmann, on the dev at l10n
mailing List points out the problem is because the translations in the
calc module are too large and exceed a 64K limit imposed by OO.o's
current resource system. Possible solutions are,
    * Split the large resource
    * shorten the translations

Discussions on Issue 33228 conclude the changing the resource file
format to 32 bit will solve the problem. This won't happen for OO.o
1.1.x, since it will break the builds, but has been fixed for OO.o
2.0. New Kannada builds of OpenOffice.org has been frozen because of
this bug.


Team Watch: Rebati project
**************************

Do gnomes like crane-meat snacks?
---------------------------------

* Rebati project homepage - http://oriya.sarovar.org/
* Rebati mailing list subscription -
http://lists.sarovar.org/cgi-bin/mailman/listinfo/oriya-group
* Rebati mailing list archives (public) -
http://lists.sarovar.org/pipermail/oriya-group/

The Rebati project grew out of the need of the institution of
Srujanika for an Oriya desktop publishing solution. Nikhil Pattnaik,
and other Srujanika members, had long since become tired of being tied
to various mutually-incompatible, proprietary solutions. The move to
open-source was helped by Gora Mohanty, who happened to be on
sabbatical in Bhubaneswar at the time. Looking up Google, and asking
around in various Oriya groups netted us more members, including
Hariram Pansari who has long been interested in local-language
computing issues and has already made contributions to Hindi-computing
development; Ajit Nayak who provides Linux expertise while marshaling
his band of enthusiastic, if sometimes rowdy, band of student
volunteers; Rajesh Pradhan, our resident fontographer and calm voice
of reason; Parasar Nayak, S. Phatak and D. Pradhan who have been keen
participants from the beginning; and D. Pattnaik who has et to deliver
on his promise of taking team members out for Chinese food. Several
renowned linguists, including Debi Pattnaik, J. Mahalik, and L. Singh,
have kindly volunteered their time to help out with language problems.
The list would not be complete without acknowledging Grizzly Kukur and
other assorted animals from Srujanika who do their level best to
disrupt all group meetings. Srujanika will continue to be a focal
point for the work, and we have unoffical groups at various other
places, including Silicon Institute of Technology, Institute of
Physics, NALCO, Sarang Engineering College, etc.

Why the name, Rebati, you ask? Ah, therein hangs a tale. Rebati is the
name of the first-ever published Oriya short story (and one of the
first Indian short stories published), by Fakir Mohan Senapati in
1898. The story is titled after the main character, a little village
girl called Rebati, who seeks an education, contrary to the mores of
the day, and against vehement opposition from her own grandmother. It
is a starkly realistic picture of village life, and in our minds a sad
commentary on the state of affairs in India that even now, over a
century later, one would not be hard-pressed to find modern-day
Rebatis in Orissa. While it might be beyond our powers to immediately
change things for her, we can at least hope to remove some of the
obstacles in her path.

We plan to provide a usable computing platform to the silent majority
of Orissa, who neither speak English nor can be expected to do so in
the near future. Our hope is to shrink the yawning gap between digital
haves and have-nots, which, if not attended to now will result in
ever-worsening economic backwardness and social alienation of the
disadvantaged. Further, we recognize that for Oriya culture and
literature to continue to flourish, we have to adapt to the
possibilities offered by new computing technology. Our initial effort
is going into localizing GNOME, so that at least one Oriya interface
is available to the user. After GNOME is largely translated, we plan
to start on KDE, OpenOffice, and the system software under the Free
Translation Project. At some point of time, we will also get working
on the documentation, HOW-TOs, etc., and start making Oriya literature
freely available online. Simultaneously, we would like to develop an
easy-to-use, open-source platform built around Linux, that provides
the basic Oriya software that would be needed by the average user:
editors, spell-checkers, word processors, optical character
recognition, desktop publishing, etc. Today, we assimilate just one
GNOME program, tomorrow the world.

To come back to earth, progress has been somewhat slow, if steady. We
have an OpenType, Unicode Oriya font that largely works, apart from
some issues currently being addressed, and that is certainly
sufficient for the translation work. Oriya rendering is still a
problem, with yudit being the one place where it seems to work
perfectly. We hope to work on Oriya rendering in Pango, and under
OpenOffice (ICU), Scribus, and KDE. For now, yudit serves as our
workhorse, with improved Inscript, Srilipi typewriter and ITRANS
keymaps. xkb keymaps are also available, and work on IIIMF is in
progress. A glossary of computer-related terms, originally built
around the GNOME glossary is available, and is being continuously
updated. We are trying to locate a freely available electronic Oriya
dictionary, and are also trying to develop one of our own. We have
completed less than 2% of GNOME, though quite a bit more of the core
programs. While we have a sizable team of translators, more are always
welcome, and we have a pressing need for computer literate people to
help out on a wide variety of technical and maintenance issues. This
is an entirely voluntary project, with our volunteers paying for
incidental expenses like Internet access, etc. We have no immediate
plans for seeking external funds, though that might change when it
comes to the implementation phase. The enthusiastic help from the
IndLinux people, and other open-source volunteers has made this
worthwhile.

I will end this long screed on a light note, explaining the sub-title.
The first localization job that we tackled from GNOME was bug-buddy,
as we were looking for instant gratification with minimal translation
effort. A native Oriya speaker, on hearing the term bug-buddy, would
transliterate it into her own language, where it ends up meaning
roughly "crane-snack". So, we figured that crane-meat snacks must be
what gnomes feast on.


Help
****
Contributions for user documentation for Indic Live CD required
---------------------------------------------------------------
All languages Indic Live CD - Rangoli will be coming up soon. At
present it has boot options for languages - (GNOME) Bengali, Gujarati,
Marathi, Malayalam, Punjabi. KDE - Hindi , Tamil. Xfce - Kannada. We
need many howto's/ guides for users to be written, preferably they be
in indian languages, and using Unicode. For more on what contributions
are needed, checkout
http://www.indlinux.org/wiki/index.php/RangoliDocs