[Mailman-i18n] i18n workflow (and encodings)?

[Clytie Siddall]

Salut Frédéric :)

On 16/11/2006, at 7:19 PM, Frederic Lehobey wrote:
>
> I am willing to provide updates to the French translation of
> mailman.

Great! :)

> I have already sent a patch¹ to Pascal George (the head of the
> French translation team:
> http://www.gnu.org/software/mailman/i18n.html) with respect to the
> version in trunk, precisely:
>
> http://mailman.svn.sourceforge.net/viewvc/mailman/trunk/mailman/ 
> messages/fr/LC_MESSAGES/mailman.po?view=log
>
> but he told me trunk was not the proper place to work on and that I
> was duplicating work. But he has not answered so far to my questions
> of what was the preferred place.
>
> So my question is: which main mailman.po should I take as an
> authoritative start for future updates? What is the proper workflow
> for localisation updates? Sorry if it is already explained somewhere
> and I have not spotted it yet².

Firstly, please read our i18n page [1] in the wiki. It links to our  
i18n howto, which I hope will answer some of your questions. If you  
ask any other questions here, we'll do our best to answer them.

However, localization projects work via language teams. It's very  
important that your language-team leader and you can work things out  
between you. Perhaps your team needs to set up closer communication,  
it's own mailing list, RSS contacts or whatever suits it best.
>
> I have an other question. In the HTML pages translated, there are many
> HTML tags like é à and so on... Is it safe to replace
> them everywhere with their 8 bit counterparts (é, à, ...)? (It makes
> the translation handling much more convenient.) If yes, what should be
> the exact encoding to use for the common mailman.po? (UTF-8, ISO
> 8859-15, ISO 8859-1?)

UTF8 is the standard encoding for use in i18n everywhere. Please use  
that. You can replace the entities with their 8-bit counterparts. My  
language is mostly accented vowels, so I know what a mess an HTML  
file is when the 8-bit characters have been converted to entities.  
There is software to convert them back. On OSX I use Unicode Checker.  
[2]

As to branches, TRUNK will be the next release. Optimally, a language  
team maintains a current translation of the latest released branch,  
but also works on TRUNK towards the next release. So both are  
relevant. However, it's up to your language-team leader, how s/he  
wants to manage the process. In smaller projects, like Mailman, some  
translators are accustomed to working alone, and may need some  
prompting to setup a shared i18n working environment. It's worth the  
effort, because each participant gained is a most valuable resource.  
So don't go away, Frédéric! We need you. :)

from Clytie (vi-VN, Vietnamese free-software translation team / nhóm  
Việt hóa phần mềm tự do)
http://groups-beta.google.com/group/vi-VN

[1] http://wiki.list.org/display/DEV/Internationalization
[2] http://earthlingsoft.net/ssp/blog/2004/03/unicode_checker_16
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : http://mail.python.org/pipermail/mailman-i18n/attachments/20061119/f669a65d/attachment.pgp