Re: Web based translation tool
Hi Dne Mon, 12 Mar 2012 12:02:49 +0200 Yaron Shahrabani <sh.yaron(a)gmail.com> napsal(a):
The Wine project is thinking about using Weblate in order to translate Wine, you can look at the following discussion and see for yourself.
That sounds great!
Looking good although a litle immature, not sure about authentication as wel...
For authentication you can basically use anything what can be used in Django. Any specific concerns on this topic?
I'll talk to Michal Cihar and ask him for a roadmap for this project.
Well I don't have any real roadmap right now. I urgently needed replacement for Pootle because we got hit by too many bugs there and did not manage to understand the code there to be able to fix them myself. So far I think it is in pretty good shape for day to day work (with today's 0.5 release fixing few annoying bugs). Of course there are some features missing, you can check what I plan for 1.0 release here: https://github.com/nijel/weblate/issues?state=open&milestone=1 I currently hope to have 1.0 in a month or so, but I really don't want to commit to any specific date. PS: I'm not subscribed to mailing list so please CC me on reply. -- Michal Čihař | http://cihar.com | http://blog.cihar.com
This is Michal's reply to our conversation. Please keep him in CC for any further updates. Michal, thank you so much for joining this conversation! ☺ Kind regards, Yaron Shahrabani <Hebrew translator> ---------- Forwarded message ---------- From: Michal Čihař <michal(a)cihar.com> Date: Mon, Mar 12, 2012 at 12:27 PM Subject: Re: Web based translation tool To: Yaron Shahrabani <sh.yaron(a)gmail.com>, wine-devel(a)winehq.org Hi Dne Mon, 12 Mar 2012 12:02:49 +0200 Yaron Shahrabani <sh.yaron(a)gmail.com> napsal(a):
The Wine project is thinking about using Weblate in order to translate Wine, you can look at the following discussion and see for yourself.
That sounds great!
Looking good although a litle immature, not sure about authentication as wel...
For authentication you can basically use anything what can be used in Django. Any specific concerns on this topic?
I'll talk to Michal Cihar and ask him for a roadmap for this project.
Well I don't have any real roadmap right now. I urgently needed replacement for Pootle because we got hit by too many bugs there and did not manage to understand the code there to be able to fix them myself. So far I think it is in pretty good shape for day to day work (with today's 0.5 release fixing few annoying bugs). Of course there are some features missing, you can check what I plan for 1.0 release here: https://github.com/nijel/weblate/issues?state=open&milestone=1 I currently hope to have 1.0 in a month or so, but I really don't want to commit to any specific date. PS: I'm not subscribed to mailing list so please CC me on reply. -- Michal Čihař | http://cihar.com | http://blog.cihar.com
Hey guys! Sorry for being so pushy but please tell me who do I need to talk to in order to push this forward. I can't use my Linux machine during the week and I don't have time to create patches, establishing a Web interface will help me improve and maintain the Hebrew translation of Wine. Thanks for your guidance, Kind regards, Yaron Shahrabani <Hebrew translator> On Mon, Mar 12, 2012 at 1:15 PM, Yaron Shahrabani <sh.yaron(a)gmail.com>wrote:
This is Michal's reply to our conversation.
Please keep him in CC for any further updates.
Michal, thank you so much for joining this conversation! ☺
Kind regards, Yaron Shahrabani
<Hebrew translator>
---------- Forwarded message ---------- From: Michal Čihař <michal(a)cihar.com> Date: Mon, Mar 12, 2012 at 12:27 PM Subject: Re: Web based translation tool To: Yaron Shahrabani <sh.yaron(a)gmail.com>, wine-devel(a)winehq.org
Hi
Dne Mon, 12 Mar 2012 12:02:49 +0200 Yaron Shahrabani <sh.yaron(a)gmail.com> napsal(a):
The Wine project is thinking about using Weblate in order to translate Wine, you can look at the following discussion and see for yourself.
That sounds great!
Looking good although a litle immature, not sure about authentication as wel...
For authentication you can basically use anything what can be used in Django. Any specific concerns on this topic?
I'll talk to Michal Cihar and ask him for a roadmap for this project.
Well I don't have any real roadmap right now. I urgently needed replacement for Pootle because we got hit by too many bugs there and did not manage to understand the code there to be able to fix them myself. So far I think it is in pretty good shape for day to day work (with today's 0.5 release fixing few annoying bugs).
Of course there are some features missing, you can check what I plan for 1.0 release here:
https://github.com/nijel/weblate/issues?state=open&milestone=1
I currently hope to have 1.0 in a month or so, but I really don't want to commit to any specific date.
PS: I'm not subscribed to mailing list so please CC me on reply.
-- Michal Čihař | http://cihar.com | http://blog.cihar.com
Hi Yaron, Thanks for continuing to push this forward. I have been a bit swamped recently. The thing is that there are several projects who are all JUST on the edge of being able to do what we need but none that are there yet. I got an e-mail from a pootle developer saying that they are doing work to do just this sort of thing for the Mozilla project however it is not ready yet, the wikimedia foundation is in the same boat. I do not know if Weblate will work or not. Basically we have 3 big requirements that have been sticking points for various solutions. 1) Real Names and e-mails for all contributors 2) Individual attribution for all translations. So we know exactly who changed what string. 3) Generate incremental diffs/patches/change sets divided based on translator. Generally it is #2 that has been the hardest for projects to cope with. -aric
Le 15 mars 2012 16:12, Aric Stewart <aric(a)codeweavers.com> a écrit :
Hi Yaron,
Thanks for continuing to push this forward. I have been a bit swamped recently. The thing is that there are several projects who are all JUST on the edge of being able to do what we need but none that are there yet.
I got an e-mail from a pootle developer saying that they are doing work to do just this sort of thing for the Mozilla project however it is not ready yet, the wikimedia foundation is in the same boat.
I do not know if Weblate will work or not.
Basically we have 3 big requirements that have been sticking points for various solutions.
1) Real Names and e-mails for all contributors
2) Individual attribution for all translations. So we know exactly who changed what string.
3) Generate incremental diffs/patches/change sets divided based on translator.
Generally it is #2 that has been the hardest for projects to cope with.
-aric
Hi Aric, I don't know weblate but as it is based on git (its home page says 'Tight git integration - every change is represented by Git commit') I think #2 and #3 should be handled easily. -- Nicolas Le Cam
Hi Dne Thu, 15 Mar 2012 10:12:14 -0500 Aric Stewart <aric(a)codeweavers.com> napsal(a):
Basically we have 3 big requirements that have been sticking points for various solutions.
1) Real Names and e-mails for all contributors
2) Individual attribution for all translations. So we know exactly who changed what string.
Every change made in Weblate is directly commited to Git repository with current translator as author, so if you count authorship of Git commit as attribution you get it. This was actually my reason for starting with Weblate and Git integration is thing which makes it different from other available tools.
3) Generate incremental diffs/patches/change sets divided based on translator.
Weblate currently does not try to merge subsequent commits from same author, so you get single commit for each change. -- Michal Čihař | http://cihar.com | http://blog.cihar.com
2012/3/16 Michal Čihař <michal(a)cihar.com>
Hi
Dne Thu, 15 Mar 2012 10:12:14 -0500 Aric Stewart <aric(a)codeweavers.com> napsal(a):
Basically we have 3 big requirements that have been sticking points for various solutions.
1) Real Names and e-mails for all contributors
2) Individual attribution for all translations. So we know exactly who changed what string.
Every change made in Weblate is directly commited to Git repository with current translator as author, so if you count authorship of Git commit as attribution you get it. This was actually my reason for starting with Weblate and Git integration is thing which makes it different from other available tools.
3) Generate incremental diffs/patches/change sets divided based on translator.
Weblate currently does not try to merge subsequent commits from same author, so you get single commit for each change.
--
Michal Čihař | http://cihar.com | http://blog.cihar.com
If this is an issue, an intermediary repository can be set up to merge translations and submit them. In fact, this can probably be done automatically every 24 hours. J. Leclanche
Hi Dne Sat, 17 Mar 2012 22:49:31 +0000 Jerome Leclanche <adys.wh(a)gmail.com> napsal(a):
2012/3/16 Michal Čihař <michal(a)cihar.com>
Weblate currently does not try to merge subsequent commits from same author, so you get single commit for each change.
If this is an issue, an intermediary repository can be set up to merge translations and submit them. In fact, this can probably be done automatically every 24 hours.
For the projects where I use Weblate I don't consider this as an issue as merging commits would probably cause more problems than it would solve. With current approach, you simply merge weblate branch to master and everything works nicely. With some sort of merging of commits things get more complex. For example for merging in intermediary repository you need to rebase weblate's working repository and keep possible changes made meanwhile. I'm not saying it can not be solved, but changing Git history (what merging of commits actually is) can be tricky. -- Michal Čihař | http://cihar.com | http://blog.cihar.com
I can't tell if it's a good idea but having a seperate repository should lead to some crucial changes, the only way it will be possible is this: This repository will have only the translations with a POT/PO file from the original project (will be updated every couple of hours). This repository will include only the English source and all the translations, once in a while (Every month or week) the changes will be merged back as one commit, the problem is I'm not sure git will enable that... It's like having a seperate module outside Wine tree... Kind regards, Yaron Shahrabani <Hebrew translator> On Mon, Mar 19, 2012 at 11:52 AM, Michal Čihař <michal(a)cihar.com> wrote:
jects where I use Weblate I don't consider this as an issue as merging commits would probably cause more problems than it would solve. With current approach, you simply merge weblate branch to master and everything works nicely. With some sort of merging of commits things get more complex.
For example for merging in intermediary repository you need to rebase weblate's working repository and keep possible changes made meanwhile. I'm not saying it can not be solved, but changing Git history (what merging of commits actually is) can be tricky.
Hi Dne Mon, 19 Mar 2012 10:52:58 +0100 Michal Čihař <michal(a)cihar.com> napsal(a):
For the projects where I use Weblate I don't consider this as an issue as merging commits would probably cause more problems than it would solve. With current approach, you simply merge weblate branch to master and everything works nicely. With some sort of merging of commits things get more complex.
It turned out such feature seems to be important for more projects, so I will end up implementing some way of merging of commits for 1.0. Details can be found at https://github.com/nijel/weblate/issues/16 -- Michal Čihař | http://cihar.com | http://blog.cihar.com
On Mon, Mar 26, 2012 at 16:23, Michal Čihař <michal(a)cihar.com> wrote:
With some sort of merging of commits things get more complex.
It turned out such feature seems to be important for more projects, so I will end up implementing some way of merging of commits for 1.0. Details can be found at https://github.com/nijel/weblate/issues/16
I don't know how your tool/system exactly works, so my comment may be a bit naive/wrong, but instead of merging commits, couldn't there be instead (or as a complement) some kind of session/transaction system where a user translates a number of msgids, then sort of "commits" its changes to close the session/transaction, which would trigger the git commit creation? That would (help) limiting the # of git commits created IMO, at the cost of some additional complexity Frédéric
Hi Dne Tue, 27 Mar 2012 00:00:04 +0200 Frédéric Delanoy <frederic.delanoy(a)gmail.com> napsal(a):
On Mon, Mar 26, 2012 at 16:23, Michal Čihař <michal(a)cihar.com> wrote:
It turned out such feature seems to be important for more projects, so I will end up implementing some way of merging of commits for 1.0. Details can be found at https://github.com/nijel/weblate/issues/16
I don't know how your tool/system exactly works, so my comment may be a bit naive/wrong, but instead of merging commits, couldn't there be instead (or as a complement) some kind of session/transaction system where a user translates a number of msgids, then sort of "commits" its changes to close the session/transaction, which would trigger the git commit creation?
You've basically described what I've outlined in issue referenced above and implemented just now in Git, so yes I think this is better approach than merging :-). -- Michal Čihař | http://cihar.com | http://blog.cihar.com
Hi Dne Thu, 15 Mar 2012 10:12:14 -0500 Aric Stewart <aric(a)codeweavers.com> napsal(a):
I do not know if Weblate will work or not.
Basically we have 3 big requirements that have been sticking points for various solutions.
1) Real Names and e-mails for all contributors
2) Individual attribution for all translations. So we know exactly who changed what string.
3) Generate incremental diffs/patches/change sets divided based on translator.
Generally it is #2 that has been the hardest for projects to cope with.
I think Weblate now does all you request here. The software still has some rough edges, but I want to polish them on way to 1.0 release in upcoming weeks. To help you with deciding, I've set up test installation with Wine sources. For now I've only imported master branch: http://wine.weblate.org/projects/wine/master/ Changes in Git repository are exported as: git://git.weblate.org/wine-master.git Please note that due to enabled lazy commits, the repository will not get changes made in web interface until it is needed or explicitly requested. -- Michal Čihař | http://cihar.com | http://blog.cihar.com
On Wed, Apr 4, 2012 at 5:05 PM, Michal Čihař <michal(a)cihar.com> wrote:
To help you with deciding, I've set up test installation with Wine sources. For now I've only imported master branch:
http://wine.weblate.org/projects/wine/master/ Hi Michal! This is so wonderful! I'm glad you did that! I can't register, can you please check why?
Kind regards, Yaron Shahrabani.
Hi Dne Thu, 5 Apr 2012 11:20:12 +0300 Yaron Shahrabani <sh.yaron(a)gmail.com> napsal(a):
On Wed, Apr 4, 2012 at 5:05 PM, Michal Čihař <michal(a)cihar.com> wrote:
To help you with deciding, I've set up test installation with Wine sources. For now I've only imported master branch:
http://wine.weblate.org/projects/wine/master/ Hi Michal! This is so wonderful! I'm glad you did that! I can't register, can you please check why?
Err, sorry. I somehow messed up email setup after testing out registration. Should be fixed now. -- Michal Čihař | http://cihar.com | http://blog.cihar.com
On 04.04.2012 16:05, Michal Čihař wrote:
Hi
Dne Thu, 15 Mar 2012 10:12:14 -0500 Aric Stewart <aric(a)codeweavers.com> napsal(a):
I do not know if Weblate will work or not.
Basically we have 3 big requirements that have been sticking points for various solutions.
1) Real Names and e-mails for all contributors
2) Individual attribution for all translations. So we know exactly who changed what string.
3) Generate incremental diffs/patches/change sets divided based on translator.
Generally it is #2 that has been the hardest for projects to cope with.
I think Weblate now does all you request here. The software still has some rough edges, but I want to polish them on way to 1.0 release in upcoming weeks.
Hi all, meanwhile there's Weblate 1.5, it seems to have what we want. Is there still a reason not to use it for Wine? Otherwise, could someone on the codeweavers side please have look at integrating it?
Hi, I just wanted to say that I have not totally abandoned this. A few months ago I again contacted my contacts at the various web translations sites and go no new good news. Still nothing gave us everything we needed. I have been super busy with other work (hence the smaller number of patches as well) so I have not had time to really dig into weblate. I have it bookmarked and want to dig. Very first pass seemed to have very little information about what is required and how to set it up. I have some more time this week so I will give it a much more serious look and try to see what I can find out. -aric On 6/17/13 4:30 PM, André Hentschel wrote:
On 04.04.2012 16:05, Michal Čihař wrote:
Hi
Dne Thu, 15 Mar 2012 10:12:14 -0500 Aric Stewart <aric(a)codeweavers.com> napsal(a):
I do not know if Weblate will work or not.
Basically we have 3 big requirements that have been sticking points for various solutions.
1) Real Names and e-mails for all contributors
2) Individual attribution for all translations. So we know exactly who changed what string.
3) Generate incremental diffs/patches/change sets divided based on translator.
Generally it is #2 that has been the hardest for projects to cope with.
I think Weblate now does all you request here. The software still has some rough edges, but I want to polish them on way to 1.0 release in upcoming weeks.
Hi all, meanwhile there's Weblate 1.5, it seems to have what we want. Is there still a reason not to use it for Wine? Otherwise, could someone on the codeweavers side please have look at integrating it?
participants (7)
-
André Hentschel -
Aric Stewart -
Frédéric Delanoy -
Jerome Leclanche -
Michal Čihař -
Nicolas Le Cam -
Yaron Shahrabani