[Ilugc] Fwd: Non-community-based approaches to localisation
Sri Ramadoss M
shriramadhas at gmail.com
Wed Sep 3 15:28:21 IST 2008
If there people from CDAC-chennai, please make note of this and reply if
---------- Forwarded message ----------
From: Gora Mohanty <gora at sarai.net>
Date: Wed, Sep 3, 2008 at 1:10 PM
Subject: Non-community-based approaches to localisation
To: "gnome-i18n @ gnome. org" <gnome-i18n at gnome.org>
Cc: "Indian Linux group ," <indlinux-group at lists.sourceforge.net>
The GNOME localisation community in India is faced
with a very peculiar situation, and it would be good
to arrive at a consensus on how to deal with this.
The BossLinux (http://bosslinux.in/) folk based at
CDAC-Chennai have gone ahead, and translated large
parts of GNOME (I believe version 2.18) into 18
Indian languages. These are available at
I applaud the scale of this effort, but unfortunately
there are some serious drawbacks here that make it
difficult, if not impossible for this work to be
integrated into GNOME:
1. I know of no attempt to contact existing language
teams prior to starting on this work. This is true
at least of Bengali, Hindi, Malayalam, and Oriya.
Worse yet, the language team line in the .po file
header has been changed to some CDAC address, which
can only lead to myriad problems down the road.
2. As CDAC made no attempt to talk to people about
consistency, the translation terms used are out
of sync with accepted ones that were used earlier.
At least for Hindi, and now increasingly for other
language, the terms that the FOSS community uses
are reviewed by outsiders.
3. The translation quality is low, at least in the
Oriya .po files that I saw. For example, "parent"
as in "parent process" has been translated into
the equivalent of "biological parent".
4. CDAC has offered these files up for the community
to submit upstream, but has apparently no intentions
of being involved in the process.
>From what I can see, and after discussions on #indlinux,
here is what I see as a possible approach:
(a) For languages that are, say more than 60% complete,
I see little benefit in trying to integrate these
files, because of points 2, and 3 above. For Oriya,
I will ask the Redhat person who now does the bulk
of the work to make a judgement call.
(b) For languages that have not been started, or are at
a very low level, it might make more sense for
people to integrate these files. However, even here
there are issues, such as unsolved Unicode problems
for some languages like Kashmiri. I am not sure how
CDAC has done translations in spite of these. I
strongly feel that good-quality translations are more
important rather than just ticking off a box for
having added another language, and would be against
the lazy way out of just integrating these files
without a review.
(c) The list of CDAC language translations with existing
teams: Assamese, Bengali, Hindi, Gujarati, Kannada,
Kashmiri, Maithili, Malayalam, Marathi, Oriya,
Punjabi, Tamil, Telugu, Urdu. There is an incipient
team for Sanskrit, and no teams yet for the CDAC
translations into Bodo, Konkani, and Manipuri. I suggest
that existing teams take a call on trying to integrate
these translations, and someone with at least a working
knowledge of Bodo, Konkani, and Manipuri step forward
to start teams.
Would like to hear your views.
gnome-i18n mailing list
gnome-i18n at gnome.org
Sri Ramadoss M
More information about the ilugc