[09:14] morning all [09:21] morning [09:24] heya happyaron, how are you doing? [09:29] dpm: fine, thanks, busy but happy, :) === jtv1 is now known as jtv [11:28] this is only work in progress... but look at the Galician, Bosnian, Slovenian and Asturian teams up there, impressive! [11:28] http://people.canonical.com/~dpm/stats/ubuntu-11.04-translation-stats.html [11:28] andrejz, good work! [11:29] thanks [11:29] and that's even before the translation maraton [11:29] i guess it will be around 103 % after it :P [11:29] you'll run out of work! [11:29] ;) [11:30] until ddtp packages exist i won't worry about it :) [11:31] also documentation still awaits us.. but it would be boring, if there were no challenges left [11:35] I think we can push some changes to ddtp project [11:35] in addition it needs to be kept in mind we were already at 99% for 10.10, so most of the work was done on other things not included in these statitics [11:35] the debian way is not so efficient, and the launchpad way will make people scared and tired... [11:36] and ddtp on launchpad are just suggestions to ddtp debian, that's a bad thing in some way... [11:36] software-center relies on ddtp heavily (IIRC), it'll be a great deal of things to be done. [11:36] back, empathy crashed on me (running 11.04) [11:36] happyaron, have you got any plans for improvement in mind? [11:37] @happyaron: what do you mean by suggestions? [11:37] do ddtp translations get used in ubuntu or do only upstream translations get used? [11:37] happyaron, nightmonkey can send ddtp translations to debian already, it's just that IIRC, the maintainer upstream was busy at the time to process them [11:38] and what do you mean that the lp way will make people scared and tired? [11:39] for debian, I can read "DDTSS is down" on debian-i18n regularly... and DDTSS needs three people to review every submission (rosetta's translations just count to one). Using the email way is making more noises. [11:40] For launchpad, it's a huge project in a single template, which make people feel tired and don't like to continue [11:40] we may split the template into more detailed categories and make each template smaller [11:41] * happyaron poor andrejz [11:41] i am a victim of empathy bug [11:42] excuse me for continious leaving/coming again [11:42] happyaron, you worte something ? [11:42] andrejz: will send to you via private msg, :) [11:42] anyway i was going to say i translated ddtp main and i believe it could easily be seperated in the strings that make sense and thnd the ones that dont [11:42] for exmaple [11:44] happyaron, ok, gotcha [11:47] as i already mentioned to happyaron [11:47] my experience with ddtp is the following: [11:47] ddtp main has 17000 strings 12:43:47 [11:47] 7000 are either python bindings, libaries for something and obscure x modules [11:48] additional 1500 are translations for gnome and kde packages (8 strings per language) [11:48] i was planning to do all the package but got frustrated (also by the length of the package - you work the whole day and make a lousy percent or so ?) and also by this advanced strings, like libbrasero, which are utterly useless in my opinion [11:49] no user is going to go and look libbrasero [11:49] User will go and install brasero and doesn't care for all the libaries it needs, system installs it for him [11:49] the same goes for ddtp main universe which i started a couple of days ago [11:49] so my suggestion is [11:49] at least get all the libaries and phython bindings into some sort of advanced package [11:50] I am translating ddtp solo, because i believe many newer membes will be intimidated by sometimes highly technical strigns they don't understand [11:51] any additional categories (besides libaries and phython bindings would be nice), but i think this is the absolute minimum [11:51] i think about 30-50 % are either a libary or phython binding [11:52] any opinions here? [11:56] +1 from me [11:59] again my technical knowledge is slightly above avereage and i have almost 0 programing / packaging knowledge, so i cannot estimate how difficult would be to implement this [11:59] it would be great if somebody from launchpad team could comment on this [12:09] andrejz, these are all good points, but we need people to be aware of them to even consider working on them. What I would suggest to do is to get a discussion started by sending an e-mail to the launchpad-users list and cc mvo (at) ubuntu (dot) com [12:10] happyaron, andrejz, in any case http://nightmonkey.ubuntu.hu/ does help [12:12] that link is rather usefull, thanks dpm === aronxu is now known as happyaron [16:54] hey translators, if you want to share your experiences, there is a session on how to run translations jams on #ubuntu-classroom in 10 min [16:55] do come along! [16:55] :-) [17:03] hi dpm can you help me to solve this bug https://bugs.launchpad.net/bugs/743153 [17:03] Launchpad bug 743153 in ubuntu-translations "Description of package 'apt-xapian-index' is badly translated to pt_BR (affects: 1) (heat: 6)" [Undecided,New] [17:05] Andre_Gondim: he's sort of busy right now ^^ [17:06] hmmmmmmm, someone may tell me where may I fix it? [17:07] askhl_: Hey, how is it, do you know where the package descriptions are stored? [17:08] hmmmmm, let me c if I found it [17:15] Andre_Gondim, yeah, sorry, running an IRC session right now [17:15] yeah I see, dpm ;) [17:16] I find it at https://translations.launchpad.net/ddtp-ubuntu/maverick/+pots/ddtp-ubuntu-main/pt_BR/+translate [17:22] TLE: hi [17:23] askhl_: ah, never mind ;) [17:23] TLE: so you don't refer to app-install-data then [17:23] which is another one [17:23] with package descriptions [17:23] askhl_: I'm unsure, it was in relation to Andre_Gondim's question [18:05] dpm, sure [18:05] hey :) [18:06] so I'm mostly happy with the output of the stats now. My goal was to keep it simple, but useful. There is now only a bug I need to tackle: [18:06] Looks realy good, the one i have done just displays a list of packages and the number of missing lines in the terminal as the statistic are generated. [18:07] posibly "Need review" is more usefull then "Total" [18:08] yeah, I should probably add it too, but I think at the point I'm generating the data files to be loaded on the page it's not easy to get the number of "Need Review" ones, so I left it for later [18:09] in any case, if you've got a branch, I can add your feature to the trunk one too [18:10] dpm, i didn't have time to make my hack preatty and the out put from yours seams to be the same+ [18:10] anyway, what I meant with the bug was that due to the way the raw data is dumped from the LP database, the templates which have got no translation at all are not shown in the output [18:10] but I think that's not critical for established teams [18:11] it affects mostly new teams with not many translations [18:11] how is the dump generated? [18:11] is to done automatically periodically or manually by some one when the time feels right? [18:13] it's done automatically with an SQL query against the staging database, which dumps the output in a couple of text files. This happens daily, but the caveat is that the staging database is not synced daily with the production database, so it happens quite often that staging is behind production, meaning that stats from the ul10n-stats tool are behind as well [18:14] unfortunately there is no easy way to solve it [18:14] ok that explains it :) [18:14] the proper solution is to finish the implementation of the lp translations reporting API, and getting rid of the raw data export altogether [18:16] Yeah, LP-trans could still improve a lot :/ [18:18] There is a real gap betwean people obliviosly making sugestions and the translation team [18:24] yeah I definitely agree, but I still think it's an awesome tool :-) [18:24] dpm, I might take a look at your code and see if i can get it to link to the actuall package on LP [18:25] Yeah we have started using it for smalle updates and package descriptions, in this area it definatly beats oure old methode. [18:26] Not to mention keeping track of what has changed and needs and update [18:27] AJenbo_, adding the link to the actual package on the table was the next thing I was aiming to sort out, but I won't have time this week, so if you look at it, that'd be really awesome!