Re: Transifex instance for GNOME

[Kenneth Nielsen <k nielsen81 gmail com>]

2010/10/15 Khaled Hosny <khaledhosny eglug org>:
> On Fri, Oct 15, 2010 at 11:26:40AM +0200, Kenneth Nielsen wrote:
>>
>> WOW easy does it. It sounds a bit like we have already made the
>> decision. This should be discussed thoroughly before we decide
>> anything or ask people to do work on it.
>>
>> While I can certainly support the idea of of-loading work from Claude
>> and others, we need to carefully consider the functionality, because
>> right now we have something that works for almost all that we want to
>> do, and we do not want to go backwards on that account.
>>
>> Claude writes that while functionality may differ, that it might be ok
>> as long as the total functionality is high. I only partly agree with
>> that. I think there is some functionality that is _essential_, and if
>> that can not be implementer then we should not move.
>>
>> The functionality that I am talking about is simple though and pertain
>> to the fact that we work a lot more with sets that with individual
>> packages. We have to be able to have overviews of sets of modules[1],
>> (and it would also be nice to have overviews of the overview[2]),
>> including translations status in the same screen, the same way we have
>> today. This is so essential to the way we work that we, in my opinion,
>> cannot do without.
>
> I was about to say some thing along these lines, but you summarised it
> better than what I would have done. I just want to added that every time
> I use Transifex I find its UI very confusing and things that can be
> simply achieved with damned-lies are either impossible or done in a very
> convoluted way. I fail to say what features Transifex offers that we
> need; we neither need the complex multi-project multi-team
> multi-workflow setup nor the near useless online translation tool (if we
> ever need one, then IMHO only Pootle is worthy consideration currently).

Well I should say, that I am not necessarily against. I just want us
to consider our options carefully.

Regarding advantages, it would be to be part of a larger community
centered around the tool. The way it is now, if we want a new feature
then we should either write it ourselves or ask Claude, who probably
already have enough to do.

Regards Kenneth