[Mailman-i18n] I wanna join in Brazilian Portuguese Translation Team.

[Barry A. Warsaw]

>>>>> "JSM" == Joao Sa Marta <samarta@ci.uc.pt> writes:

    JSM> For the moment I am the only one working on the "pt_PT"
    JSM> translation, and I have not finished the translation so far .
    JSM> But I think that there must exist two versions for the pt
    JSM> translations.

So let me see if I understand: What we currently have in cvs as the
"pt" translation should really be "pt_BR" translation, with you
providing a "pt_PT" translation some time soon?

>>>>> "GMdS" == Gleydson Mazioli da Silva
>>>>> <gleydson.mazioli@ima.sp.gov.br> writes:

    GMdS> Normally, the portuguese Brazilian language is set to "pt"
    GMdS> if it's added before the original pt_PT langauge (it's
    GMdS> mother language). As portuguese Debian translator, I suggest
    GMdS> to add Brazilian Portuguese as pt_BR for future inclusion of
    GMdS> pt_PT language, if possible :-)

    GMdS> That approach could open ways to add more support for
    GMdS> variants of mother language (documented in ISO 3166 or new,
    GMdS> if I'm not wrong)

So here's the question: will the pt_PT and pt_BR catalogs and
templates be two completely independent translations, or will one be a
partial specialization of the other?

In the latter scenario, Mailman would look for a message translation
or a template first in pt_1 and if that's missing it will look for it
in pt_2 (I don't know if pt_1==pt_BR or pt_1==pt_PR).  If this is the
case, then we may have some kludging to do.  Python 2.2's gettext.py
supports the notion of a `fallback' so you could set it up to search
pt_1 first followed by pt_2 fairly easily, although I'd have to add a
hack for the templates.

Let me know which of these scenarios we need to support.  Having pt_PT
and pt_BR be independent complete translations is the easiest thing,
so before I go hacking around I'd like to know which we need to
support.

Thanks,
-Barry