Dictionary to translate (was:Re: Persian translations)

[Ramanan Selvaratnam <rama uklinux net>]

Cool,

I am happy Pablo arrived at something I have been struggling to raise
for some time.
(was not sure whther this will be ignored as off topic)

On Mon, 2003-12-01 at 14:38, Pablo Saratxaga wrote:

> It depends also on the final goal: is it to have a usable system, yet
> perfectible; or is it to have a showing case for academic purposes?

In the Tamil case recently there has been some potencially exciting
developments on this. While chasing up a forked attempt a developer
highlighted that a government agency was funding OOo and other l10n
efforts.  Things are not fully clear yet but I remain positive that by
explaining things clearly FS could have a huge organised input as Pablo
suggests. 

> It depends also a lot on the willingness of the official world to work on
> it or not, and how much; if the government gets seriously involved, providing
> full support providing comprehensive lists of terms, maybe also dictionaries

I will come back to the dictionary seperately and the end.

> to translators (either on paper or electronically), spell checker, program
> testing, etc. then it is perfectly legitimate for the government to
> actually demand that the translations be done following the language rules
> they dictate.

(Hesitantly I make the move on a public archived list hoping this will
never be found ever again :-)

In Tamil Nadu the govenment is even now promoting two encoding schemes
(TAB and TAM) that are outside UTF-8 and TSCII (the one mentioend by
Pablo in an earlier mail) 
So this rule is not always safe. Beware.
(Did  I mention the many keyboard layouts?)

> However, if they are slow, and not fullu involved, then it is legitimate
> I think that the translations be done separately, provided that they
> are adapted when an update on official rules is published.

Why cannot translations be done at the GUI development projects and a
common dictionary be pumped with terms (instead for the good of
mankind).
If someone drops down slow the next steps in and cooperates, shares,
helps whatever ...


> I know that officials work in a very very different way, and it is quite
> hard for them to understand how the development of free software works;

Seconded.

> but if they are commited to the developement and use of the language, 
> it should be possible to have agreements.

Given that they do not seem to do what is required for funding in the 
the health sector by a certain multinational IT Corp... I will like to
know all about any such possible agreements and how GNOME foundation or
GNU could  help here.
Is this possible? 

> The most important think imho, is to make them know that it is not just
> about translating (like you could translate a book, in a finite timespan,
> and then the work is completly finished), it is about an evolutive process,
> with the material to translate being in constant evolution, and with
> also the ability to constantly change and improve the translation itself;
> make them know that the dynamic of the process is something very important,
> and breaking it could have bad consquences, like the translations being
> always out of sync and outdated, or potential translators diverted and
> reaching far lower levels of translation due to that.

I think an electronic dictinary is the best solution.
In the case of Tamil I am not too far from a very good one maintained by
the state university.

How will the copyrights and licensing affect this?
Basically what is required in such a dictionaries terms and conditions?

Best wishes,

Ramanan