[07:22] morning all [07:30] morning [07:33] dpm i have a question about import queue [07:48] hey andrejz, good morning, sure, go ahead with the question [07:52] in the import queue the is a bfd.pot from package "crash" but in natty there is bfd from binutils [07:54] this confuses me a bit [07:55] andrejz, this is not a problem in Launchpad, as we can either give the template a different name, or leave it the same. Both can coexist with the same name, as they come from different source packages. However, if their translation domain is the same, then this is a bug in the package, as there cannot be two programs using the same translation domain in the system [08:11] automatically suggested value for translation domain is the same (bfd) but i don't know what it ought to be [08:11] where can i check that [10:13] Hey, sometimes there are empty translated strings in LP. Anyone familiar with this problem?\ [10:33] RawChid: yes, this can indeed happen [10:34] You know why? [10:34] if you accidentally click in the translation field, but write nothing in it and you save it, this will appear as translated - to the empty string [10:34] The strings are added via web interface of Launchpad. Don't know if that has to do with it... [10:35] Hmm [10:36] I'm thinking of a RFC for Launchpad now.. Maybe don't accept empty translations? Or better, if the source is not empty, then translation may not be empty. What are you thinking? [10:39] kelemengabor^ [10:39] hm, perhaps this is already fixed? I just tried it and it looks like I'm wrong [10:40] yeah, I'm still here :) [10:40] Hm yeah. I can't imagine that I made this "mistake" so many times [10:41] And I've seen it from several translators [10:41] I'm sure that for long time, when you clicked into the new translation field, the radio button was automatically selected, but now it looks like this is not the case anymore [10:43] Yeah, and that causes another usability problem; that my translations is not submitted :P [10:44] Sometimes you must click the radio button explicitly [10:46] I think I experienced that not so long ago, but how is that possible? [10:47] I'm not sure. I can't reproduce that at the moment. [10:47] I have typed something in the field, saved it, and next time I looked at it, it was not there [10:48] Ah, I see. When you have a text field (small one, NOT a textarea). Then klik on the button to expand it (tooltip: makes field larger). And then type your translation, the radio button stays on "no translation" [10:50] I use this functionality oftentimes because the small field doesn't have spell check in my broswer [10:51] So my work disappeared :( [10:57] argh, with my disappeared translations I used the same, they were short strings which I wanted to break in two lines, so I expanded them and boom, they gone [10:57] congrats RawChid, you just caught an LP Translations bug :) [10:59] do you want the honor of filing it? :) [10:59] https://bugs.launchpad.net/launchpad/+filebug [11:01] Yeah sure, I fill it later this day [11:01] Lunch time :) [11:44] kelemengabor, I filed the bug: https://bugs.launchpad.net/launchpad/+bug/811014 [11:44] Launchpad bug 811014 in launchpad "Translated strings don't get saved (affects: 1) (heat: 6)" [Undecided,New] [11:44] cool, thanks! [11:45] But coming back on my original problem (maybe I need to file another bug) [11:46] 12:13:37 < RawChid> Hey, sometimes there are empty translated strings in LP. Anyone familiar with this problem?\ [11:46] That problem is a bigger deal to me [11:46] Not sure if it's a bug though [11:51] well, I'd say this is not really a big problem, unless you change an existing translation to an empty one [11:51] which is an unfortunate case, and perhaps should not be allowed at all [11:51] dpm: danilos: what dou you guys think? [11:52] hi kelemengabor, reading the scrollback... [11:53] I asked the person who said this to me to give me examples (string ID's) for further investigation... Can't give an example at the moment [11:53] kelemengabor, setting a translation to empty one is the only way we can turn something into a suggestion if it's already approved; also, sometimes you don't know what the correct translation is, but you do know that the current one is incorrect [11:55] danilos: these look like valid use cases, thanks. [11:56] danilos: so there is a distinguish between "no translation" and "empty translation". I'd prefer to use "no translation" for your examples [11:56] as for bug 811014 RawChid filed, I am not sure if this ever used to work for textareas (it still seems to work for regular input fields you get originally) [11:56] Launchpad bug 811014 in launchpad "Translated strings don't get saved (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/811014 [11:56] RawChid, if you are certain that this used to work for textareas, please add a comment like that so it can be marked as a regression and critical [11:57] I'm not sure what you mean danilos. If it has worked or not. It is a usebility issue. [11:57] kelemengabor, right, the problem with "no translation" is that Launchpad uses current translation review date as a cut-off date for reviewed suggestions, so you might get already reviewed suggestions listed as "new suggestions" if we just remove the translation [11:58] RawChid, I agree, but we are currently stomped with many critical issues, and if this has never worked, it's unlikely to be fixed in the next 6 months or so :/ [11:58] danilos I GUESS it hasn't worked at all. [11:59] Should I mention this? [11:59] RawChid, right, thanks [11:59] RawChid, nope, no need to do that then [11:59] This is more a RFC than a bug then right? [12:00] RawChid, it is a bug indeed, but I was just wondering if we can make it critical with the rest of our critical bugs (what we do for regressions) [12:00] Oke, thanks for the heads up [12:07] And the "empty translation" problem. You said it can be user to indicate that a translation is not correct. So when I see empty translations it is because someone else has made it empty explicitly? [12:07] danilos^ [12:21] RawChid, yes, most likely (we used to have a bug where we set translations to empty if there were non-existent in the import, but long time ago and for a short while) [12:27] I can't imagine that someone has done that to my translations. We should communicate such things to each other. But I'll keep an eye on it [12:28] Thanks anyhow [13:16] kelemengabor, just a quick heads up that I haven't forgotten about the call notes, it's just that I'm having a busy day and I'll probably send them on Monday [13:16] no problem :) [15:31] kelemengabor, can I ask you something about: name of translator present but no translation yet (RawChid's question too) [15:32] sure [15:32] hi dpm [15:32] kelemengabor, I'll try to explain: [15:33] In ddtp-ubuntu-universe I regularly see this: Translated by Hannie on xxx, but (no translation yet) [15:34] My name is just an example [15:34] Why is there a name of an translator, whilst there is no accepted translation yet? [15:35] hannie: if somebody removed the old translations [15:35] translation, I mean [15:35] it will look like that. [15:35] I do not think any of my fellow translators removes accepted translations [15:37] I think this can happen when someone uploads a translation file which has untranslated strings [15:38] at least some time ago I saw this to happen, altough recently I don't see it anymore [15:38] kelemengabor, that's what I thought, but my fellow translator hasn't done that [15:39] It is no big deal, but I am only curious why this happens [15:40] I have no idea, why - Launchpad is such a beast, and it is known to have strange habits :( [15:41] You tell me ;) [15:41] But thanks for your answer, and henninge too [16:01] dpm_: is there any way to see a debdiff between two versions of a package, like https://launchpad.net/ubuntu/maverick/+source/ubuntu-docs ? [16:01] I'm asking because of this: https://bugs.launchpad.net/ubuntu-translations/+bug/690248/comments/16 [16:01] Launchpad bug 690248 in ubuntu-docs (Ubuntu Maverick) (and 2 other projects) "In Maverick 'About Ubuntu' displays Natty info (affects: 64) (dups: 23) (heat: 258)" [Medium,Confirmed] [16:02] looks like this went out, but I'd like to double check [16:03] and, if this is everything that was applied, then it's pretty obvious what happened: only the original file was updated, its translations in the same package weren't [16:03] but, this is only my theory [16:07] kelemengabor, I need to go, but I'll try to have a look at it on Monday. Please ping me again about it if I forget [16:07] cheers!