[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Debian gsoc for i18n

Quoting Stan Ioan-Eugen (stan.ieugen@gmail.com):

> > I have mixed feelings here (which indeed explain why I didn't followup
> > immediately).
> I was beginning to think nobody cares about improving i18n.

People certainly do care. But I somehow suspect that many will also
think something like "let bubulle answer" when a mail like yours shows

> > So, assuming the Pootle is not suited is not necessarily the good
> > approach and I'm not sure that ${fancytool} is better suited..:-)
> As I mentioned, I am not very familiar with i18n internal workings. I know that
> po files can be generated at the same time as building the binary package and I
> assumed that these files can be aggregated in a central place by doing
> some processing
> on packages. It seems that I identified the wrong problem.

There are a few things in place for doing such things, already. For
instance, all PO files from debconf translations are gathered on a
regular basis and placed in an SVN. And Pootle is hooked on that SVN.

However, I never pushed this anywhere....because we need another part
to be worked on: something that would allow package maintainers to get
the l10n material back from this SVN into their packages....at build

It's probably something like a debhelper script that would do it (of
course optionnally as being online for buildshas never been a

So, a good work could be implementing that (and design the whole
process around it). Of course, it would need to be versatile enough to
not depend on a given VCS, etc.

*this* would bring one of the needed prerequisites for a useful
central repo of l10n material....and then as a side effect, for
putting in production a really useful online translation tool.

> I was thinking of replacing Pootle because it lacks a way to
> encourage competition among translators (like display a top list of
> translators).
> (I am thinking of ways to attract translators here).

I'm not sure we want to encourage competition about the amount of work
done. Things have always been more about quality than quantity.

I sometimes "play" with "emulation" between l10n teams (and thus
languages) to motivate people to raise the overall completion status
for their language...but that's just *oneù way to reach our goals.

Still, such feature could anyway be welcomed in Pootle, too (it maybe
already exists, I haven't checked).

But, anyway, the "final" tool shouldn't probably draw the overall design

> > Still, I don't want to discourage you and it would probably not be a
> > great idea to rule your proposal out..just for this.
> > However, what you also need to find a a mentor... and, here is the
> > reason for which I didn't answer immediately: I can probably be seen
> > as "the mentor to have"...however, I'm not sure sure I could be able
> > to give the needed time and attention all over the GSOC months. My
> > involvment in Debian varies from month to month, depending on my
> > professionnal activities, which are more and more demanding. So,
> > really, I would be very  careful before replyign "yay, I can mentor
> > you"...and if anyone else feels like (s)he can do that, please feel
> > free to speak and don't assume that, because this is i18n, bubulle can
> > do everything..:-)
> Thank you for your support, I still wish to improve i18n and I am open
> to ideas as to what can be done.
> Regarding mentoring, I can work on my own, without constant mentor
> attention, so please consider mentoring
> me. I usually ask for a list of things tobe done/learned and go on my
> way. If I don't manage I can go and ask
> other people.

Still, mentoring requires some important work that has to be done
(such as reporting to GSOC about the work done by students, etc.)
which I might be in trouble completing if I'm busy elsewhere. Hence my
"hesitations" because I wouldn't like that failures from my side could
be a problem for the student.

Attachment: signature.asc
Description: Digital signature

Reply to: