Annoucement: New translation status page is installed

Bernd Groh bgroh at
Fri Jun 25 02:46:13 UTC 2004


>Just to summarize my position: As long as this system allows
>unconsistent translations, I'll think it is flawed, and thus I'll oppose
>to it (in a nice way). That I oppose to it doesn't mean I will not
>respect it.

Point taken. But I believe every system allows inconsistent 
translations, as such, every system is flawed, no? ;)
Of course, *iff* you have a really dedicated maintainer who QAs all 
translations before a commit and makes sure everything is consistent, 
this system actually works, I completely agree! :)

>And yes, I think the translation of Fedora has been flawed since day
>one, it's up to us to get it right. I was really surprised when I got
>the CVS account at Red Hat for translations, without questioning or
>hesitation from your side, I appreciate the trust, but it can also bring
>undesired lateral effects, fortunately I'm a nice guy :)
>Finally, I believe that keeping the distribution consistent and coherent
>is more important than getting a new translator started.

I'd even agree, to why I disagree with your position, I will come to in 
a minute.

>Colophon: Who can tell if a translation is "right" or not? Or if it is
>"consistent" or not? that will be hard... maybe looking at "previous
>art", merit, or experience... Tough question, really, that's the weak
>part of my argument. Probably a team of translators for the same
>language would be able to decide what's "right", like in a democracy, as
>long as anyone can join a translation team, of course :)

Agree. :)

>>Believe me, this point is very well taken, but at the same time, you can 
>>watch who is translating what, and if you recognize a new translator, 
>>invite her/him to your team. Why does it always have to be the new 
>>translator who gets in contact with the group? I personally believe in 
>>making it as easy for people to get started as possible. Wouldn't it be 
>>easier for the group to send an invite?
>Right, if there's the case I (or someone else) will ask a new translator
>to join us, of course. It is just that s/he might not want to do it, or
>might not want to use the same terminology we've used for the rest of
>the translations... then what?

Then what? Well, then we can talk to this person, and if s/he unwilling 
to use the "right" terminology, then we can think of disabling her/his 
account. Btw, I do translate 'Forward' with "Next" if 'Next' is meant, 
and not with "Forward", consistent or not! IMO, it's just plain wrong! 
;) Just kidding. :)

> One application says "Quit" the other
>says  "Exit"... This is the only concern I have, and as a solution I
>only see enforcing (doh!) translation teams. But maybe that's not the
>goal of Fedora Translation Project (FTP?, ouch), maybe the goal is just
>to have the distro translated not consistent as well.

No, both is the goal. But, really, I much prefer to have one application 
say "Quit", and another one "Exit", than have either use the english 
term. Why should a user who always only uses program ("Quit") and never 
uses program ("Exit") care? And anyway, both are understandable, both 
are right in some respect and both have meaning to a user. And no, I am 
*not* arguing that I don't care whether translations are inconsistent or 
not, I do care a great deal, but I am arguing that I rather have an 
inconsistent, yet understandable, translation than none at all. *Iff* 
you have a dedicated maintainer who is willing to coordinate with such a 
great effort, and can manage that all translators agree on a certain 
(and not another) terminology, I completely agree with you, then even I 
believe it's a better way of doing things. But, I doubt it's the case 
for most of the languages. Some languages don't even have a translator 
yet, do we not give them access, because we don't speak xyz and nobody 
is there who could invite them? Some languages might have a lot of 
translators, but all of them are reluctant to really put that much 
effort in and be such a maintainer. I, personally, don't like the idea 
of having a system in place where it is required to have a great 
maintainer who does all these things, cos if a language doesn't have 
one, it may just die. No, really, I rather have people contribute 
unmaintained than not at all. And if there is such great maintainer, 
well, that is great, you have the option of being a maintainer, and if 
not for the entire language, for a module, or a group of modules. A 
maintainer has the additional right to release modules of translators, 
so they can excercise some level of control. And if somebody 'causes 
problems after that, well, then even I am happy to kick that person.

>  At least forcing the translator to commit himself to make a
>translation that will be consistent with the rest of the translations
>for a particular language.

I disagree. Asking her/him to do so, yes. And if s/he is unwilling to do 
so, and doesn't have a good reason of why, then we can indeed disable 
this persons access.

>  Just another example to illustrate what I mean: "log viewer" (the
>application) appears in three or four different files: menus, dist,
>anaconda and redhat-logviewer (iirc). All translations should use the
>same translation of "log viewer" or we'll confuse the user, but four
>different translators could use four different translations (yes it
>could be possible!).

Sure it could be possible, but if you don't have a maintainer who makes 
the effort to really check *all* of these cases, then there's no 
difference really, is there? And it could also be possible, that all 
translators translate it the same way, maybe by discussing it, without 
having to go through a maintainer. Yes, it could be possible!

>  I don't know if Red Hat has a policy on how the interface to the user
>should be, and in what way should it address the user, for instance if
>the policy (if any) encourage to ask the user: "Please, press the
>button", or to order: "Press the button", or at developer's will (which
>I think would be inconsistent)... Thinks like this make a difference.
>Our policy, to give yet another example, tells us to address the user as
>"you" [polite you] ("ni" in Swedish), and orders from the user to the
>computer are in imperative form. Now, being consistent with this is key.
>If there's an application that doesn't follow the same style (or policy)
>it will disturb the user, and will harm the distro (imo). I'm not saying
>that ours is the best policy, I'm saying that it has to be consistent,
>and I know you agree, but the new method doesn't guarantee that (or does

As you've said yourself, you know I agree. :)

>I'll apply for maintainer of ca (Catalan), if no one objects. The web
>page doesn't really tell what the role of a maintainer is, though, I
>think it should be stated somewhere. So I wait until I know what I'm
>messing with :)
>Policy on how to become maintainer and how to depose one should be
>stated too, probably.

Noted. A maintainer is, as opposed to the translator, a permanent (well, 
in theory) entity. The maintainer can commit at any time and release the 
assignment of a translator at any time, hence, keep someone from 
comitting. A maintainer is a person who is willing to take 
responsibility for the quality and consistency of a module, a group of 
modules or for an entire language. And while it is a good thing to have 
a maintainer, it is not a requirement. I know you think it should be, 
but that's exactly where we disagree. I believe a translation community 
can function well, even without having a designated maintainer, and 
that's what the system really is made for, to better manage the efforts 
of self-maintained translators.

>I just took dist, and it looks kind of unfair to have my name as a
>Translator, as we're, so far, 4 people working [slowly] on that file,
>and I am probably the one that has contributed the less. But well, it's
>something we can live with, I suppose :)

As said, the Translator field is not about getting credit, it's about 
knowing whom to contact regarding translation issues of a certain 
module, or finding out that a module is available and one can take the 
module and translate or fix things herself/himself.

>Now, I know I seem to take this a bit religiously, and maybe I do, I
>just want you to see the light and become a convert, then I shall give
>you a place in the heaven of translationland :)

That's ok, and if you show me the scriptures in which I shall find the 
law that for every language such great maintainer is instantly found and 
either always holds this position, or if required at any time, another 
such great maintainer as replacement is instantly found, then I will 
invite you to detail these scriptures for me. :)

Best Regards,

>P.S.: I promised myself it would not be a long mail, but... here it

Isn't it always like that! ;)

More information about the trans mailing list