Mailinglist Archive: opensuse-translation (237 mails)

< Previous Next >
Re: [opensuse-translation] openSUSE Weblate
  • From: Alexander Melentev <minton@xxxxxxxxxxxx>
  • Date: Sun, 11 Oct 2015 00:06:00 +0300
  • Message-id: <CAMe-c7XZnoMC9qkVW38+ihHr+KX=iW5brMGzxbH1ea7vOtisgw@mail.gmail.com>
2015-10-10 20:48 GMT+03:00 Tomáš Chvátal <tchvatal@xxxxxxx>:
Hello,
Hi, Tomáš. Before I start my inline ranting I'd like to let you know
how grateful I am for your extensive explanations, we all have missed
them from the beginning. Really, I mean it, thank you so much.
full and official announcement of the beta will be going at some point next
few
weeks. We just needed to "steal" some of the projects already being translated
to ensure good live testing.
And one of the "stolen" projects is RELEASE NOTES exactly a few weeks
before the release when there is still no single translator familiar
with this new and not still established process. How do you expect to
have any release notes translated in time for this release? I really
would like to see the genius eyes of the guy who decided to do exactly
that in this specific moment.

The plan is to have weblate for all the stuff, because it will by this per-
project git repository solve most of the trouble we are having wrt
longstanging SLE issues and now inherited by Leap issues.

We thought really hard about the problem and the fact that weblate actually is
open-source and can be glued to internal infrastructure was the winner.
Yes, we all know this FOSS mantra "who codes decides", but I really
don't think that trading devs headaches for translators' headaches is
a way to go.

Now for the concerns raised over the thread:

1) po translations will be allowed even externaly, ie. direct git commits
2) weblate allows mass-download and mass upload, so you can use the tool even
for some checks if you don't want to do direct commits
Having more than one way to push translations in is pure evil cause it
leads to conflicts. Code developers might think that git access for
everybody might be a good idea and is managable, but in case of
translations in terms of quality and consistency it is a road to hell:
there are substantial differences between collaborating on code and
collaborating on translations. Before making such "game-changing"
decisions it might be a good idea to consult/collect feedback from the
guys who have real experience with this "game". Lack of such
communication at the very start of significant change is really
frustrating.

So you can try the translating as of now and collaboration will be deployed in
close future. If you find any problems with the tool, like crashes where we
didn't noticed them, or something does not behave like it should please sent
sbrabec an email about it or file a ticked in bugtracker.
Please be more specific as we still have no valid info about authors
of this decision and proper ways of communication with them. For
instance, I once was denied to fulfill my NEEDINFO request to coolo,
cause the email address I picked for this was not the one coolo
expected (while still being valid coolo's address). So, what is the
correct sbrabec's address for this? What is the correct
bugtracker/component for such "early bird" bug reports? Where exactly
was the discussion about all this held and how we can take part in it
and make comments and requests? Who should we discuss the solution
architecture with, i.e. this easily foreseeable
scalability/managability issue I described earlier about a single
translator VS 150+ repos? I am really interested in all this.

PS: Did you know that some of those PO translations in SVN were not used for
like 7 years? Now they are :)
It is good that new tools can solve old issues. Now let's ensure they
will not create more new issues and I think we'd better do it together
=)


--
Regards, Minton.
--
To unsubscribe, e-mail: opensuse-translation+unsubscribe@xxxxxxxxxxxx
To contact the owner, e-mail: opensuse-translation+owner@xxxxxxxxxxxx

< Previous Next >
This Thread
Follow Ups