Bug#855223: debian-i18n: DDTSS rejects valid alias

[Thibault Polge]

Package: debian-i18n
Severity: normal

When trying to create an account on ddtss, my usual alias "thblt" (lowercase 
letters T, H, B, L and T) is rejected with the message:

> Aliases are numbers, letters and underscore and at least 4 letters long, 
> thanks.

Where it obviously matches these restrictions, being five letters.  After a 
(very) quick testing, it seems the actual minimum is six letters.

Thanks,
Thibault

-- System Information:
Debian Release: 9.0
  APT prefers testing
  APT policy: (500, 'testing')
Architecture: amd64 (x86_64)

Kernel: Linux 4.9.0-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=fr_FR.UTF-8, LC_CTYPE=fr_FR.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Re: Bug#855179: debian-reference-it: Disable Weblate for currently .po and git maintaned Italian translation

[Holger Wansing]

Hi,

Beatrice Torracca  wrote:
> I saw that a recent change was made via Weblate (changing the
> translator assignement, as I said).
> 
> I always assumed that Weblate was enabled only for those languages who
> did not have an official translator in the Debian project.
> 
> In the past I sent my translations for review in the
> debian-l10n-italian mailing list the first time and for any major
> change. And I did commit my translations.
> 
> I understand that translation, like everything else in Debian, is a
> community work. But as a wishlist I would like to ask to enable
> Weblate only for those languages that don't have an official/active
> translator; or at least to disable it for those languages that have an
> active translator and request it.
> 
> I feel that translation needs consistency and review, when possible
> (think of a scenario where everyone was able to do NMU on every
> package...I think every developer would be very wary of such a situation).
> 
> In my case, I feel like I would have to either recheck every change
> made or abandon a translation that I worked a lot on (the Debian
> Reference is quite a big document) but that I can not guarantee the
> quality of (even if possibly the quality of the Weblate one would be
> better than mine).
> 
> Moreover being noted as the Last-Translator in the .po header, I
> always replied to any call for update or any other request of
> information/correction. So I think that having someone in the
> Last-Translator field who is officially responsible of a translation
> is a plus. The same is true for the Language Team header that was
> changed from "Italian " to
> "Italian
> ". I
> have no idea if/where the Weblate Italian team is reachable (hopefully
> mails would ge redirected to the correct debian-l10n- team.
> 
> This is just a general wishlist based on my past experience with
> Weblate; it has nothing to do specifically with the person or the
> quality of the recent change (1 string, correct as far as I can see)
> made via Weblate to the Italian translation.
> 
> I put on CC the debian-i18n mailing list because I see that
> translation are moving more and more towards Weblate and I would like
> to point out this side of having an uncontrolled/unreviewed source for
> Debian translations. I apologize for the noise if the CC was out of place.

I totally support this request.

I have already noticed one of "my" translations being hijacked by Weblate
service, where one string (!) (an URL, to be strict) was translated.
That did not gave much value to the translation.
And it looks to me like working without any concept.

Moreover, translations in Debian are community work, which means working
together, as a team. Hijacking translations is not fine style.
At least I have learned it like this here in Debian.

Also, anonymous translations have been disapproved in several other Debian
sub-projects, since you cannot communicate with the translator, to inform
about errors or to ensure consistency with terms. I feel, this Weblate
translations are also some sort of this "anonymous" translations, especially
because they happen without coordination, as it seems.


I am unsure, when this Weblate thing has entered Debian ... ?

How does this work?

Did I miss an announcement about this on debian-i18n or something like that?


Regards
Holger


-- 

Created with Sylpheed 3.5.0 under
D E B I A N   L I N U X   8 . 0   " J E S S I E " .

Registered Linux User #311290 - https://linuxcounter.net/

Re: Bug#855179: debian-reference-it: Disable Weblate for currently .po and git maintaned Italian translation

[Helge Kreutzmann]

Hello Holger,
On Wed, Feb 15, 2017 at 07:46:15PM +0100, Holger Wansing wrote:
> I totally support this request.

I totally agree to you both as well, especially Beatrice, given 
that I do the same for those translations where I'm listed as 
last translator.

> I am unsure, when this Weblate thing has entered Debian ... ?
> 
> How does this work?
> 
> Did I miss an announcement about this on debian-i18n or something like that?

This is the first time I'm informed about weblate, so I would like to
get some more information as well, especially how to configure
(disable) it for those translations where I'm responsible for.

Greetings

   Helge
-- 
  Dr. Helge Kreutzmann deb...@helgefjell.de
   Dipl.-Phys.   http://www.helgefjell.de/debian.php
64bit GNU powered gpg signed mail preferred
   Help keep free software "libre": http://www.ffii.de/


signature.asc
Description: Digital signature

Re: Bug#855179: debian-reference-it: Disable Weblate for currently .po and git maintaned Italian translation

[Laura Arjona Reina]

Hi

El 15/02/17 a las 21:23, Helge Kreutzmann escribió:
> Hello Holger,
> On Wed, Feb 15, 2017 at 07:46:15PM +0100, Holger Wansing wrote:
>> I totally support this request.
> 
> I totally agree to you both as well, especially Beatrice, given 
> that I do the same for those translations where I'm listed as 
> last translator.
> 
>> I am unsure, when this Weblate thing has entered Debian ... ?
>>
>> How does this work?
>>
>> Did I miss an announcement about this on debian-i18n or something like that?
> 
> This is the first time I'm informed about weblate, so I would like to
> get some more information as well, especially how to configure
> (disable) it for those translations where I'm responsible for.
> 

I guess this is a decision about how to handle translations just for
the package debian-reference in particular.

I know that other Debian packages are handled using weblate too, at
least for several languages. The debian-handbook comes to mind.

I am a weblate user (translator) in hosted.weblate.org for other
project (F-Droid) and when I login, I can lock a translation so it
only allows me work on it. Maybe that's a workaround for all the
languages that have a translation-maintainer that does not work with
Weblate.

Weblate also allows to members of a project to receive notifications
about many different events that happen (updated strings, updated
translations, new people joining the project, etc). Maybe that helps, too.

CC'ing Osamu Aouki, the maintainer of debian-reference package,
(probably knowing how is the current workflow or having better ideas
than me).

Maybe a mention of the proposed workflow for translator should be made
in the README file, too. I suggest to use
https://anonscm.debian.org/cgit/debian-handbook/debian-handbook.git/tree/README.translators
as reference, since that project also uses weblate for some languages,
and the usual git commits for others.

Cheers
-- 
Laura Arjona Reina
https://wiki.debian.org/LauraArjona

Re: Bug#855179: debian-reference-it: Disable Weblate for currently .po and git maintaned Italian translation

[Beatrice Torracca]

Hi,

thanks to all for the feedback.

On mercoledì 15 febbraio 2017, at 22:08 +0100, Laura Arjona Reina wrote:
> I am a weblate user (translator) in hosted.weblate.org for other
> project (F-Droid) and when I login, I can lock a translation so it
> only allows me work on it. Maybe that's a workaround for all the
> languages that have a translation-maintainer that does not work with
> Weblate.
> 
> Weblate also allows to members of a project to receive notifications
> about many different events that happen (updated strings, updated
> translations, new people joining the project, etc). Maybe that helps, too.

I am not sure that notification will be of help. It's one thing to have
notification on a 100 messages translation of a program and another
thing to get notifications for every change on something like the
debian-reference that has more than 7500 messages.

I don't think a translator should "police" their own translation if
they are keeping it reasonably up to date. 

Translations particularly gain a lot from consinstency. The same
term translated in two ways (both correct) across a document can lead
to a lot of confusion.

And another main thing I worry about the Last-Translator and
Language-Team assignments which are the addresses used in case of call
for updates or bug reports.

On the other hand I don't know the inner workings of Weblate and maybe
it has some ways to guarantee quality assurance.

> Maybe a mention of the proposed workflow for translator should be made
> in the README file, too. I suggest to use
> https://anonscm.debian.org/cgit/debian-handbook/debian-handbook.git/tree/README.translators
> as reference, since that project also uses weblate for some languages,
> and the usual git commits for others.

My main experience with Weblate comes from debian-handbook; as a
coordinator, not translator. It was problematic. Translors did not
like when their work was changed and they perceived the change in
Last-Translator as "stealing" their work. Some translation were
published before review and with errors not corrected.

I understand that Weblate can make translations easier and thus get
more things translated and that is good. But I think some work-flow
should be put in place such as to try to guarantee the same quality
standards that Debian enforces for any other aspect of the Project.

Allowing anyone to make whatever changes to everything and *also*
becoming the person officially responsible for that translation no
matter how big/litte/correct the change was sounds not reasonable to
me. Maybe to understand my worries think of the same situation in
terms of packetization instead of translation (yes, packetization
surely is more critical than translation as far as errors go! just
trying to get my thoughts through :)).

I just wanted to express my thoughts as a translator and coordinator,
so that developers and package maintainers could maybe take into
account this side of things when deciding how to handle translations.

Thaks,

beatrice

signature.asc
Description: PGP signature