| dpm | good morning all | 07:58 |
|---|---|---|
| andrejz | morning | 07:58 |
| dpm | hey andrejz :) | 07:58 |
| artnay | hi dpm | 07:58 |
| artnay | dpm: chrome upstream importing overwrited (once again) chromium translations, should I file a bug against LP itself? | 07:59 |
| artnay | I did /q to fta but he hasn't replied yet | 07:59 |
| artnay | I've fixed many oopsies made by chrome translator(s) (hired by google, I guess) but chrome imports have overwrited those fixes | 08:00 |
| artnay | hence those translations are now wrong in both chrome and chromium | 08:00 |
| artnay | grr, chrome translators should review the changes made in rosetta and take those to upstream when suitable | 08:01 |
| artnay | I'm pretty sure this problem exists in all languages that chromium is translated into | 08:02 |
| artnay | henninge: | 08:03 |
| andrejz | @artnay: probably languages without official google translations are better off | 08:03 |
| dpm | hi artnay, just tell fta on #ubuntu-mozillateam, with the exact details. afaik, the setup he's got is explicitly to overcome this problem | 08:05 |
| artnay | I wonder if this affects other projects as well | 08:12 |
| andrejz | hm, artnay i am looking at it right now and it seems some strange strings arrived on 2011-01-10 | 08:13 |
| artnay | andrejz: chromium strings update approximately once a week | 08:14 |
| andrejz | but i am not sure whether these were untranslated strings, which were automatically translated from upstream or if our strings were overwritten | 08:14 |
| artnay | a lot of new strings appear and unused ones gets tossed away | 08:15 |
| andrejz | yeah, we noticed that. maintaining chromium translations is rather difficult due to high rate of changes | 08:15 |
| andrejz | i can see that a lot of strings we have changed in launchpad haven't been overwritten | 08:16 |
| henninge | artnay: reading | 08:30 |
| henninge | artnay: sorry, distracted | 08:57 |
| henninge | artnay: this is bug 740225, I am sorry. | 08:57 |
| ubot4 | Launchpad bug 740225 in launchpad "The differences between New and Translated for upstreams was removed (affects: 2) (heat: 27)" [Undecided,Won't fix] https://launchpad.net/bugs/740225 | 08:57 |
| artnay | henninge: I don't know if that's the case... and even if it is, it's plain wrong | 08:58 |
| henninge | artnay: what do you mean | 08:58 |
| artnay | Rosetta states that differences between rosetta and upstream will be held in Rosetta | 08:58 |
| artnay | but, umh, which one is upstream? chrome or chromium? | 08:59 |
| henninge | I don't understand | 08:59 |
| artnay | henninge: this way we can't fix the errors made by Google's translators | 08:59 |
| artnay | sure we can file bugs against chromium issues but they don't tend to get fixed | 08:59 |
| henninge | artnay: that will only work between Ubuntu packages and upstream projects which are hosted in Launchpad. | 09:00 |
| artnay | for example in Chrome it says my connection to LP is secured by 128 byte encryption | 09:00 |
| artnay | so I fixed it in chromium to be 128 bits | 09:00 |
| artnay | and during the last import the fix got overwritten | 09:00 |
| henninge | artnay: not between projects hosted in Launchpad and upstream projects hosted elsewhere. | 09:01 |
| henninge | artnay: yes, that is bug 740225 | 09:01 |
| ubot4 | Launchpad bug 740225 in launchpad "The differences between New and Translated for upstreams was removed (affects: 2) (heat: 27)" [Undecided,Won't fix] https://launchpad.net/bugs/740225 | 09:01 |
| artnay | sorry to hear that. it basically makes fixing a pita/impossible | 09:02 |
| artnay | unless the Google translators review the changes and fix it in Chrome | 09:02 |
| henninge | artnay: yup, I am sorry but that is correct. | 09:02 |
| henninge | artnay: the only way to get that would be to do the translations in Ubuntu: | 09:04 |
| henninge | https://translations.launchpad.net/ubuntu/natty/+source/chromium-browser | 09:04 |
| henninge | why is there an extra "translations" series on chromium-browser??? | 09:05 |
| artnay | henninge: fta2 said it shouldn't be that way | 09:07 |
| henninge | what way? | 09:07 |
| henninge | the series or doing it in Ubuntu? | 09:07 |
| henninge | artnay: ^ | 09:09 |
| artnay | henninge: series | 09:09 |
| artnay | I'll file a bug later today about this | 09:12 |
| artnay | if there are chromium translators here who are also affected by this, it would be great if you could confirm | 09:12 |
| henninge | artnay: I agree about the series. Is that what you want to file a bug about? | 09:14 |
| artnay | henninge: basically the bug would be "in chromium, don't overwrite changes made in LP when doing upstream import" | 09:16 |
| artnay | it makes fixing impossible without Google's help | 09:17 |
| artnay | don't get me wrong, I'd like to see those fixes in Chrome as well | 09:18 |
| henninge | artnay: I understand taht | 09:19 |
| henninge | that | 09:19 |
| henninge | artnay: in the bug you should make clear that you are aware of bug 740225 and that you seek a different solution (i.e. translating on the source package although chromium-browser is not in main) | 09:21 |
| ubot4 | Launchpad bug 740225 in launchpad "The differences between New and Translated for upstreams was removed (affects: 2) (heat: 27)" [Undecided,Won't fix] https://launchpad.net/bugs/740225 | 09:21 |
| artnay | henninge: I'll try to be precise with my bug report (will include screenshots, that bug # etc.) | 09:21 |
| henninge | artnay: I don't think screenshots are necessary. | 09:22 |
| artnay | thanks for looking up at that bug (again, you already did it before ;-)) | 09:22 |
| henninge | artnay: you don't need to explain much about what is going on because that bug already does that. | 09:22 |
| artnay | henninge: with screenshots I'm able to show a) original translator (fta) and time b) string c) my fix (name and date) | 09:22 |
| henninge | artnay: yes, but we are already aware that strings get overwritten. You do not need to proof that. | 09:23 |
| henninge | artnay: you can kick off the finding of a different solution, though. | 09:23 |
| artnay | henninge: ok, no screenshots. fta2 was asking for those wrongly translated strings | 09:24 |
| artnay | with dates one could track back to upstream imports | 09:24 |
| henninge | artnay: that would include keeping the upstream translation in the project and the LP translations in the source package. | 09:24 |
| henninge | artnay: that is something the Ubuntu people like dpm will have to comment on. | 09:25 |
| henninge | because chromium-browser is not in main. | 09:25 |
| henninge | artnay: oh, I forgot | 09:26 |
| henninge | artnay: fta has his machinery outside of LP doing stuff. | 09:26 |
| dpm | artnay, henninge, there will be no translatable ubuntu source package for chromium, as long as it remains in universe | 09:26 |
| dpm | fta explicitly created all his machinery to overcome that issue | 09:26 |
| henninge | dpm: yes, I just rememberd | 09:27 |
| artnay | I've seen fta's flow charts, gee... :-) | 09:27 |
| artnay | no wonder if stuff like this happens | 09:27 |
| henninge | artnay: ok but then this is not a new bug for LP to report | 09:27 |
| artnay | henninge: andrejz said their fixes aren't being overwritten so something is quite fishy here | 09:29 |
| artnay | I'm seeing all my fixes (being the only active chromium translator) being overwritten | 09:29 |
| artnay | in Finnish, of course... *g* | 09:29 |
| andrejz | Well, i checked for a couple of strings which we have changed and those weren't overwritten | 09:30 |
| andrejz | it's also possible those strings have been accepted in upstream chromium and are not intact | 09:31 |
| andrejz | now* | 09:31 |
| artnay | andrejz: when were those strings originally fixed? | 09:31 |
| andrejz | sometime in december 2010 | 09:31 |
| artnay | andrejz: ok, I have to check if any fixes from 2010 are still as-is. all fixes from 2011 have been overwritten. | 09:32 |
| andrejz | new launchpad rollout was around January 20th | 09:33 |
| artnay | yep | 09:33 |
| andrejz | so maybe that's when strings started to be overwritten | 09:33 |
| artnay | andrejz: well #740225 was probably released in 2011 | 09:34 |
| artnay | it saddens me to see regressions like this in both LP and transifex | 09:34 |
| andrejz | me too. it takes a lot of work to fix this and can seriously demotivate contributors | 09:36 |
| henninge | artnay, andrejz: Can either of you point me to fta's blog where he posted his flow chart? | 10:00 |
| andrejz | http://ftagada.wordpress.com/2011/02/10/chromium-translations-explained-part-2b/ | 10:01 |
| andrejz | http://ftagada.wordpress.com/2011/01/23/chromium-translations-explained-part-2/ | 10:01 |
| andrejz | http://ftagada.wordpress.com/2011/01/19/chromium-release-management-explained/ | 10:01 |
| andrejz | http://ftagada.wordpress.com/2011/01/08/chromium-translations-explained-part-1/ | 10:02 |
| henninge | andrejz: thanks | 10:06 |
| artnay | In Basque, some time ago I made a typo and translated the word .Offer. (in .generated resources.) with the word .Eskaiki.. It should be .Eskaini.. Now, each time a correct the mistake in Launchpad (4 times by now), the next day I got the mistaken word again in Launchpad. | 10:07 |
| artnay | It.s a restriction i had to add because Launchpad now overwrites the LP contributed strings when there is an upstream update, even unrelated, in the template, which happens almost every day in Chromium. | 10:08 |
| artnay | As I said, if the problem is ever solved in Launchpad, I should be able to remove this restriction. | 10:08 |
| artnay | it seems that LP/something now thinks upstream strings as "updates" to those strings which have been corrected in LP | 10:09 |
| artnay | that's a major drawback | 10:09 |
| artnay | basically makes correcting impossible unless Google's translator(s) fix their own oopsies | 10:11 |
| artnay | we can file those translations oopsies at chromium issues but as has been said, they don't tend to fix their own oopsies (or if they do, it takes a lot of time) | 10:12 |
| artnay | More seriously, I had to make some choices. The most important one is that for the .updated in Launchpad. strings, I always prefer those over the upstream translations. The drawback is that it is no longer possible to return to an upstream translation once it diverged. I consider it a small price to pay at the moment, but maybe it will have to be revisited one day. | 10:12 |
| artnay | http://code.google.com/p/chromium/issues/detail?id=1528 this is one of the best examples I could find. totally wrong translation, reported 2,5 years ago, not fixed in Chrome, fixed in Chromium, gets overwritten. | 10:30 |
| artnay | I'll file a bug at Chromium issues, something like "co-operate with Chromium translators, review their changes and commit into Chrome where appropriate" | 10:32 |
| === henninge is now known as henninge-lunch | ||
| andrejz | hello! i am wondering if anyone is experincing a problem in indicator date-time | 12:54 |
| andrejz | "Add event ..." appears untranslated | 12:54 |
| andrejz | does any other team experience this | 12:54 |
| andrejz | ? | 12:54 |
| === claudinux is now known as Claudinux | ||
| === henninge-lunch is now known as henninge | ||
| dpm | andrejz, I've noticed that too, but I thought it'd be because translations have not been shipped yet. | 14:17 |
| dpm | If that is not the case, would you mind filing a bug? | 14:17 |
| andrejz | well that was my question actually :) do i need to fill a bug or will it resolve by it self (when translations ship)? | 14:17 |
| dpm | andrejz, is the string translatable in LP? | 14:18 |
| dpm | if it is, you can test this: | 14:18 |
| dpm | 1. Export the translation as mo file from LP | 14:18 |
| dpm | 2. Rename the exported mo file to indicator-datetime.mo if necessary | 14:19 |
| dpm | 3. sudo cp indicator-datetime.mo /usr/share/locale-langpack/sl/LC_MESSAGES | 14:20 |
| dpm | 4. Log out of the session | 14:20 |
| dpm | 5. Log back in again and check if "Add event..." is now translated | 14:20 |
| dpm | I need to step out for a while, but I'll check the log when I come back if there is anything else | 14:21 |
| dpm | Oh, and 6. If it is not translated, please consider filing a bug | 14:21 |
| dpm | :-) | 14:22 |
| andrejz | @dpm: it works, so it's not a bug | 14:46 |
| === Claudinux_ is now known as Claudinux | ||
| dpm | andrejz, great, thanks for testing :) | 15:09 |
| andrejz | glad to help | 15:10 |
| === henninge_ is now known as henninge | ||
| === dpm_ is now known as dpm | ||
| artnay | do you have "Launcher & Menus" untranslated? | 19:13 |
| artnay | is it even marked for translations? | 19:13 |
| artnay | https://bugs.launchpad.net/ubuntu-translations/+bug/757884 | 19:43 |
| ubot4 | Launchpad bug 757884 in ubuntu-translations "No translation template for "Launcher & Menus" (affects: 1) (heat: 6)" [Undecided,New] | 19:43 |
| artnay | to which package does it belong? | 19:43 |
| === Claudinux_ is now known as Claudinux | ||
| === Claudinux_ is now known as Claudinux | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!