[08:47] good morning all [09:39] Good morning :) [09:41] Morning [09:44] morning RawChid [09:44] hey andrejz [09:44] hello dpm [09:44] you already know what your other projects are (besides translations) [09:46] andrejz, yes, it's going to be to grow the app developer community. Similar to what I'm doing with translations, but with an entirely different community :) [09:47] but don't worry, I'm still staying on translations, only splitting my time between them and app developer stuff [10:01] I guess you can extra stimulate to l18n their Apps :P [10:06] RawChid, absolutely, this is part of a secret plan to improve i18n support in all Ubuntu applications ;) [10:11] dpm, convince the developers that commenting where the string appear really helps A LOT [10:12] I know, I know, I keep telling people... [10:13] i have a question for you dpm [10:13] is there a list of all the pacakges which are in 11.04 by default? [10:13] (to be able to order the templates) [10:13] I found this, but this is probably ubuntu only http://releases.ubuntu.com/natty/ubuntu-11.04-desktop-i386.manifest [10:14] don't worry, just found it [10:15] i have another question , modified gnome packages should be higher in priority than nonmodified one [10:15] is there a simple way to determine if package has been modified in ubuntu? [10:16] http://patches.ubuntu.com/ [10:19] andrejz, dpm: Maybe we should be a session about that at some point, maybe at the next ubuntu app dev week: Possibilties and Pitfalls in internationalisation: Beyound scattered _( ) [10:19] *G* [10:19] thank you yurchor, but this shows difference between ubuntu and debian. in case where ubuntu uses newer gnome than debian it will be marked as patched (in regard to debian) even if it's not patched (in regard to gnome) [10:19] @TLE: good idea ;) [10:19] morning TLE :) [10:19] good morning [10:22] andrejz, re: the list of default packages - I answered this one last week: " andrejz, getting the list is doable, but not trivial. I'm doing this when I calculate the statistics with https://code.launchpad.net/~dpm/ubuntu-translations/ul10n-stats, and that's where I gave it to you from (not sure if to you or some other team) last time. [10:22] I can generate a list, but it would have to wait until after UDS" [10:22] Let me see if I can do it some time this week [10:23] off course, as long as the developer comments don't make it into LP (or I don't know whether they do now), it might be a little bit odd to do a session like that an Ubuntu forum ;) [10:23] that would be great dpm, because then it will be much easier to arrange the templates (since the order depends whether it's default or not) [10:24] andrejz: Debian does not change messages (usually). You can grab ftp with wget, then find the messages with some application (gettext messages can be readily found). [10:24] TLE, developer comments are fully supported in LP luckily [10:24] * TLE feels somewhat behind [10:25] in all cases, i.e. nomatter if the localisations comes from code being hosted on LP or from a po-file? [10:26] TLE, yeah, if the pot template has got the comments, they will appear in LP, no matter where the code is hosted [10:26] andrejz, yeah, I know it would be useful, and the ul10n-stats script should generate them, it's just that it's not trivial, and I'm lacking the time to do it. But I need to find it somehow :) [10:26] nice, then we should definitely do the session [10:26] :D [10:26] ok great, that should really help [10:27] Sometimes it is hard to find Gtk/Gettext comments without context (see FontForge). [10:28] Developers put general comment for a bunch of messages and if someone have already translated first of them it is hard to find the comments for others. [10:29] I'm not sure what you mean [10:30] You mean that developers put a comment on one string, and in that comments explains that it applies to a lot more strings like it? [10:31] TLE: Consider the following example. Developer trying to put the huge comment to the five or six option. [10:32] He write something like: //xgettext: Tru-la-la, Option 1: blah-blah, Option 2: la-la... [10:33] and then comes the code. [10:33] Developers does not want to give the comment to each option. He gives the whole comment. [10:34] And in Launchpad you can see the comment to the message on the other page. :'( [10:34] Yes, so when you say optinos, you are talking about different strings? [10:35] Yes. [10:35] yes, I know the problem, but that is not exactly LP specific, you can just as easily miss the comment in emacs or gtranslator [10:36] but the idea of group comments is also one of those bad practices that could be mentioned in such a session [10:38] chromium is an excellent example for how you can comment your code: *every* string has an explanation [10:38] TLE: Do not use emacs or Gtranslator because of their bad TM management. ;) [10:38] yurchor: same applies to all po-editors [10:39] No. Lokalize can show everything I need and have reliable TM. :p [10:39] since there is no way to detect from the a group comment how many strings it applies to, there is no way to show it for the relevant strings [10:40] trijntje: most developers will be reluctant to comment to much, as they feel it clutters up the code [10:40] But you can see it on a sidebar. [10:40] see a comment for a string that does not have one? [10:41] See the context and easily see the near string with the comment. [10:42] context.. surrounding code? [10:42] No (though it is possible too), surrounding messages. [10:43] surrounding messages based on code placement or position in po-file? [10:44] Usually it is good practice to sort messages by context, though they are regretfully alphabetically sorted by some developers. :'( [10:45] well, that is the same you would get from emacs, where you just move through the raw po-file, so there you also see surrouding strings [10:46] but whether it will work or not, depende on how many extra strings you see, how many strings are in the group and whether you make the connection [10:47] so it is a somewhat working workaround, for a problem that should not exist in the first place [10:48] Yes. But the only case with readily seen interface with translation in FOSS is Linguist/Qt. [10:49] It is a trade-off. Simplicity vs. good interface for translators. [11:20] What does "TM" above refer to? [11:20] (mentioned 30 minutes ago) [11:22] i'm guessing translation memory [11:22] ah, okay [11:23] * askhl wishes people would generally use less abbreviations [12:15] askhl: TM also confused me a bit the first time I read it, [12:16] and FYI IMHO think that gtranslators TM works quite well [12:16] ;) [12:18] however, as physicist, we are not really in a position to ask other people to dial down on the abbreviations *G* [12:26] touché [12:58] dpm, do you know if gnome-user-docs-user-guide is still being used in natty? === henninge is now known as henninge-lunch [13:04] andrejz: i don't think that it is. i think it's just gnome-user-docs [13:04] i am wondering whether it should be disabled in natty or not? [13:13] andrejz, I'm not sure about the exact template names right now, but I enabled and disabled all docs-related templates for natty as per mdke's recommendations [13:14] so they should all be up to date (i.e. the right ones enabled) in natty [13:14] ok, cool [13:54] Creating a new irc channel for a translation team, is following a special procedure or just having someone joined the channel? [14:19] thinkabout, on Freenode, you can just join a channel to create it [14:26] dpm: is there a special bot that will keep it alive [14:28] dpm: don't you need to register somewhere to keep the channel? [14:38] thinkabout, TLE: https://wiki.ubuntu.com/IRC/CreatingChannels gives detailed instructions [14:40] dpm: thank you, this was the wiki page i was looking for. :D [14:41] great :) === j1mc_ is now known as j1mc