[00:13] nixternal: You there? [00:13] what's up? [00:13] nite. [00:13] nite fdoving [00:13] Some guy is cussing up a storm in #kubuntu ... "[18:12] fuck diz all life, fuck that, my mom hiding cigaretes from me and i wanna to smoke now, omg i just will blow her head" ... I dunno if he's calmed down now, or not, but may be something to keep an eye on... [00:14] yes, I saw [00:14] he's mine stdin, back off :p [00:14] stdin: Ah, I'm sorry! I didn't see your name on the access list when I first looked. I do now. [00:14] he's calming down now, but I was *just* about to remove... [00:15] hehe [00:15] warn, then quiet, then remove :) [00:15] I warned, but he swore 3 more times as I was typing the warning :p [00:16] hahaha [00:16] Anyway, thanks, both of you. :D [00:17] nivek: if you think there's no one around in #kubuntu and you need help, poke someone in #ubuntu-ops [00:17] stdin: Will do. I knew that nixternal was an op from memory because I remember him booting someone the first month I started coming to #kubuntu :P [00:18] you're not an real op untill your first kick/ban :) [00:18] haha === jpetso_away is now known as jpetso [00:28] mhb: sorry, my internet broke [00:28] * Jucato read that as "bloke".... [00:29] nosrednaekim: so you need a script in /etc/X11/Xsession.d/ which reads a setting you write somewhere [00:29] maybe ~/.kde/ [00:29] or ~/.compiz [00:29] doesn't really matter [00:29] and that sets $KDEWM which gets picked up by startkde [00:29] nosrednaekim: compiz ought to fall back to kwin, that it sometimes doesn't is a bug bug that can be fixed [00:30] Riddell: ok... thanks! === Jucato is now known as Jucato_ === Jucato_ is now known as Jucato === \sh is now known as \sh_away [01:57] right. i should be able to install kde4 now :) [01:57] is it worht installing? [01:58] oh, shit. i missed the meeting again [01:59] oh frack! [02:00] no stuff to the ML about it either. i wonder if it even happened [02:01] me too [02:02] I don't think it was kinda nice to have it scheduled the day after new year :) [02:03] It did happen [02:04] pfft :( [02:08] it happened.. [02:09] yuriy was accepted.. and that about all that happened [02:09] later all [03:01] * Hobbsee ponders waiting until the kde4 actual release [03:03] 8 days :) [03:03] or you can wait when it's tagged at least? :) [03:04] anyone know an easy way to add a line edit to a toolbar in qt4-designer? [03:09] :( [03:10] * Jucato doesn't know.... [03:40] Hobbsee: You didn't miss much as far as the meeting went. One new member and one item on the agenda :P [03:40] nivek: fair enough [03:40] You want a log? [03:43] http://pastebin.com/m4f4b64dc If so. [03:45] thanks [04:17] * dasKreech waves [04:20] heya dasKreech [04:20] hiya dasKreech! :) === awen_ is now known as a [04:22] Hey Jucato! === a is now known as awen_ === BobbySponge is now known as krwk === apachelogger__ is now known as apachelogger [09:09] yuriy: Well, eventually, I may move the methods to .cpp files. It's just more comfortable this way. [09:09] yuriy: At least for now. [10:27] hi there [10:27] rebonjour Tonio_ [10:29] bonne année Riddell ! [10:30] Riddell: I'm back in the work, fixing bugs for kde 3.5 etc... [10:30] Riddell: most people seem to focus on 4.0 so I'll do my best on the 3.5 part.... [10:30] Riddell: like fixing ark, broken [10:31] Riddell: I found the problem with guidance power manager [10:31] Riddell: the problem is outdated .pot file [10:31] Riddell: I have a few questions on that point [10:32] Riddell: is there a way to import the pot in rosetta for the package's page ? [10:32] rosetta will import it when you upload [10:32] Riddell: no admin/ folder in rosetta, so no way to patch [10:32] Riddell: it was never imported in fact [10:33] Riddell: I also have to fix powermanager code a bit to include some missing i18n() things [10:33] https://edge.launchpad.net/ubuntu/+source/kde-guidance [10:34] Riddell: as you can see, there is no translation entry in it [10:34] https://translations.edge.launchpad.net/ubuntu/gutsy/+source/kde-guidance/ [10:34] Riddell: hum weird... [10:35] Riddell: oki let's patch guidance then [10:35] and include an updated pot file [10:40] Riddell: just a little question concerning guidance, again [10:41] Riddell: I see both __tr() and i18n() are used, but never together, is there a reason to this ? [10:43] self.screensavergroup.setTitle(self.__tr("Enable screensaver")) [10:43] here an example [10:43] I suspect we should use self.screensavergroup.setTitle(self.__tr(i18n("Enable screensaver"))) [10:47] Tonio_: is the tr() used in files made from .ui files? [10:47] not only [10:48] Riddell: you have it in the standard guidance-power-manager.py for example [10:48] Riddell: I try to use i18n in it, and if it works, then will patch everything [10:49] Tonio_: probably a mistake then [10:49] yup [11:43] Riddell: patch is done, but I can seem to build the .mo file manually [11:43] Riddell: any clue for this ? [11:43] Riddell: I have the patched .py files installed, seems to work, but I can hardly test if I can't compile the .po files [11:44] Riddell: even local debuild -nc doesn't built it... [11:46] Tonio_: I don't know if they've ever built as part of the package [11:46] they'll build as part of the language pack of course [11:47] you can try and work out the gettext command to compile them [11:47] Riddell: the debian package should contain the commands right ? [11:47] Riddell: debian doesn't have our langpacks mecanism afaik ? [11:48] maybe kbabel can do it btw === \sh_away is now known as \sh [11:53] the debian package shouldn't be much different from ours [11:54] oki [12:11] * Tonio_ is happy, no new bugs posted with the latest kdesudo [12:11] fdoving: ^_^ === claydoh_ is now known as claydoh [12:29] Riddell: the problem looks more complex in fact [12:29] the .ui file generates a ui.py file automatically, which doesn't contain any i18n strings [12:30] I did ask if that's where they came from [12:30] uic has a flag to change that translation function [12:30] tht's it indeed [12:30] not sure if pyuic does [13:04] Riddell: what is the flag for the standard uic please ? [13:07] -tr use func() for i18n [13:07] Tonio_: ^^ [13:08] okay suprt [13:09] Riddell: I was searching for i18n and not internationalisation, shame on me..... [13:09] Riddell: you can blame me :) [13:17] Riddell: seems to work, testing [13:27] Riddell: looking at guidance packaging, there are things missing : [13:27] -XGETTEXT=/usr/bin/kde-xgettext EXTRACTATTR=/usr/bin/extractattr sh $(DEB_SRCDIR)/admin/cvs.sh extract-messages [13:27] this cannot work, as there is no admin folder in it [13:28] btw this isn't supposed to work with .py afaik [13:30] wouldn't imagine it could [13:31] Riddell: that's from debian/rules ;) [13:32] Riddell: fixing to something that will work [13:47] Riddell: seems to work ;) [13:47] Riddell: btw, there are fixes that I'd rather have in the tarball [13:47] Riddell: can you update the svn branch if I send you a fixed tarball, and then I'll prepare a package with the debian/rules fix ? === \sh is now known as \sh_away === keffie_jayx is now known as effie_jayx [14:14] Tonio_: can do [14:14] Tonio_: there's a patch in dholbach's patch queue too i think [14:17] Riddell: great [14:18] Riddell: http://pastebin.com/m62574e65 [14:18] Riddell: here-s the patch [14:19] Riddell: I'll ping dholbach once available, but I'll release a new package as soon as you have commited [14:19] as well as a new kdebluetooth ans a new knm [14:30] Lure: wasn't knetworkmanager 2.1 in the pipe for month ? ;) [14:31] Tonio_: I thought so too, but not sure where it got stuck. You may want to ask on mailing list [14:31] sure [15:07] grmpf, bad internet connexion today === Tonio__ is now known as Tonio_ [15:16] https://bugs.edge.launchpad.net/ubuntu/+source/kde-guidance/+bug/177366t [15:16] Launchpad bug 177366 in kde-guidance "misspelling found - lowercase 'w' in 'wine'" [Low,Triaged] [15:16] https://bugs.edge.launchpad.net/ubuntu/+source/kde-guidance/+bug/177366 [15:23] Tonio_: I've applied your patch and that one to svn [15:29] Riddell: super [15:29] Riddell: building the new package, and then trying to fix ark :/ [15:30] Riddell: I noticed the current gutsy version of kdesudo causes lots of problems, as lots of bugs have been fix [15:30] Riddell: shouldn't we consider a security update ? [15:30] or even a backport [15:30] backport sure [15:30] I doubt it'll pass security [15:31] Riddell: well the bad args parsing was a potential security issue... [15:31] but yeah, few changes it passes security ;) [15:33] Riddell: also with security, the minimum code changes are required right ? then no need to even think about it :) [15:37] Tonio_: right [15:43] Tonio_: kdebase/4:3.5.8-2ubuntu10 broke [15:47] Riddell: yeah, the patch was doubled in the patch/ fodler [15:47] Riddell: already uploaded a ubuntu1 version that builds [16:04] Heya [16:04] boo :) [16:05] EEEK! [16:07] hi everyone [16:07] hello Tonio_ [16:08] :-) [16:11] Riddell: is there a delay before I can svn update on the anonsvn server ? [16:11] Riddell: I can't seem to ugrade now.... I'll wait a bit [16:12] Tonio_: 10 minutes usually [16:12] okay [16:12] Riddell: well it was 38 minutes ago ;) === neversfelde_ is now known as neversfelde [16:18] Tonio_: definately there http://websvn.kde.org/branches/extragear/kde3/utils/guidance/ === \sh_away is now known as \sh [16:20] Riddell: yeah, doing manually is another option ;) [16:25] Riddell: http://websvn.kde.org/branches/extragear/kde3/utils/guidance/wineconfig/ [16:25] Riddell: I can't see the wine typo fix uploaded... [16:31] hrm [16:33] Tonio_: it's definatly in svn [16:33] http://kubuntu.org/~jriddell/tmp/wineconfig.py [16:34] Riddell: http://websvn.kde.org/*checkout*/branches/extragear/kde3/utils/guidance/wineconfig/wineconfig.py [16:34] Riddell: not updated here at least, I can't see the fix in it [16:34] shrug, slow sync [16:35] maybe due to the mega icon commits [16:35] Riddell: looks like.... especially since the .pot fix is in for more about an hour now ;) [16:35] Riddell: I applied the patch manually in the sources, it'll be the same as the svn tree [16:36] Riddell: so could you answer my question now, please? I asked about the KDE4 Hardy CD and when it is ready (to be tested) [16:40] mhb: it's blocked on seeds being reorganised [16:41] Riddell: that is so last year :o) just joking, but the answer was the same before Christmas. [16:42] mhb: I doubt colin spent his christmas rearranging seeds [16:43] mhb: is mhb.ath.cx down? [16:44] ryanakca: no [16:44] Riddell: right, here's hoping that he might spens some time in January about it :o) [16:50] Riddell: don't we want the grubconfig guidance tool ? [16:50] Riddell: it is commented in rules [16:51] dunno, ask mhb [16:51] I believe it wasn't ready last I looked [16:52] k [16:52] mhb: ping ? ;) [17:00] Riddell: last (sorry) question about the rules file [17:00] Riddell: why did you use kdepyuic instead of the standard pyuic ? [17:01] Riddell: kdepyuic doesn't seem to allow the usage of parameters like -tr for example [17:01] Riddell: is it of any use for the program ? [17:01] Tonio_: well they're KDE apps, I suspect they don't work with the qt pyuic [17:02] yeah, the standard pyuic doesn't work with KDE classes/eidgets [17:04] hum, I then have to find a way to make kdepyuic use -tr instead [17:04] Riddell: afaiu that to generate pykde .py compatible files right ? [17:05] Tonio_: you can just use sed [17:05] s/tr\(/i18n\(/ [17:06] Riddell: right [17:28] <\sh> Tonio_: pyuic has a switch for that...where you can set this...and pykdeuic is just an addon to it, so it should use the same switch when I remember correctly [17:29] <\sh> Tonio_: if not, it's a bug ,-) [17:30] \sh: then, a bug it is :) [17:30] \sh: kdepyuic doesn't allow to set options in fact, only 3 are available [17:31] <\sh> Tonio_: kdepyuic has the -tr switch [17:31] <\sh> (pykde3) [17:32] \sh: it hasn't, the switch is for pyuic [17:32] <\sh> Tonio_: http://pastebin.ubuntu-nl.org/50608/ [17:33] \sh: the output you see is the pyuic output [17:33] <\sh> Tonio_: hehe...kdepyuic is just a wrapper around pyuic...it inherits all stuff from pyuic and adds kde stuff too [17:33] \sh: look carefully to the output, you have both in a raw, buggy [17:33] <\sh> Tonio_: yes...but it should call pyuic [17:33] \sh: basically kdepyuic allows 3 options : -a -i and -p [17:34] \sh: it does, but there is no way to add specific options for pyuic [17:35] Riddell: interesting, the .py files you generate with kdepyuic are not installed [17:35] Riddell: those in the current package are done with the build system with pyuic [17:36] Riddell: that's why I couldn't get my patched files in the package...... [17:36] Riddell: I'll probably patch the build system to use kdepyuic instead [17:36] <\sh> Tonio_: give me a sec...tryingf [17:36] \sh: sure [17:37] \sh: would be nice if I could avoid the sed :) [17:38] \sh: already tried -p "pyuic --tr i18n" fyi [17:38] Riddell: looks like our powermanager app is not a kde but a qt application :) [17:39] at least the ui [17:39] <\sh> Tonio_: -p is just the path to pyuic [17:39] <\sh> Tonio_: I see now...it's a bug in kdepyuic...we should fix it ,-) [17:40] \sh: ;) [17:40] \sh: I don't have time to wait for this, I'll fix in guidance waiting for a fix instead ! [17:41] <\sh> Tonio_: I would say, regarding the fact that kde is just using i18n, we should add a "-tr i18n" to opts = ' -o ' in doPyuic [17:41] <\sh> or adding an option to it directly and adding a conditional to doPyUic... [17:42] <\sh> Riddell: any objections? [17:43] \sh: well it isn't that simple [17:43] \sh: that'll generate i18n entries in the code, but the catalog should be loaded too [17:44] \sh means something like "from kdecore import kglobal" [17:44] \sh: and also add the kglobal.locale() somewhere in the code :) [17:44] <\sh> Tonio_: so lets do it...i didn't check on pykde4 this time...but fixing pykde3 is essential...we can push the patch as well to upstream [17:45] \sh: and that I don't know how to do it correctly [17:45] <\sh> Tonio_: that should be done at application lvl... [17:45] then just one line after loading kglobal should be enough [17:46] \sh: I'm all for that [17:46] \sh: I can wait for the fix to release guidance then [17:47] <\sh> Tonio_: well, I'll propose a patch for it...and you can test it...I'll upload pykde3 to my ppa [17:47] \sh: stupid question, but how will you now the catalog name to load ? ;) [17:47] \sh: KGlobal.locale().insertCatalogue("guidance") for example [17:47] <\sh> Tonio_: as I said..this should be done at app lvl...not in pykdeuic ... [17:48] \sh: but how will that be done for a py file generated from a ui file ? [17:49] ho you mean that this file is imported and then it should work ? [17:49] <\sh> Tonio_: as it is global, it should inherit everything from top...yes [17:49] * Tonio_ isn't a python expert [17:49] oki [17:49] * \sh neither regarding locales .-) [17:50] <\sh> but thinking of qt/kde libs in general, where you do this as well in main somehow...it should be the same [17:50] yep, understood [17:50] \sh: I'll wait for your patch then [17:53] \sh: [17:53] if addApp: [17:53] opts = ' -x -o ' [17:53] else: [17:53] opts = ' -o ' [17:53] that's the 2 lines to patch right ? [17:54] <\sh> Tonio_: well, I'm doing it correctly..adding the opt to the toplvl options...and feed it as conditional to DoPyUic [17:55] \sh: oki super, please ping me when done so that I can test it and fix guidance build system to use this instead of pyuic :) [18:02] <\sh> Tonio_: sure [18:11] <\sh> Tonio_: pykde3 is in my ppa...just waiting to build...python-kde3 - 3.16.0-4ubuntu1~ppa1 is the version...for hardy...add [18:11] <\sh> deb http://ppa.launchpad.net/shermann/ubuntu hardy main [18:11] <\sh> deb-src http://ppa.launchpad.net/shermann/ubuntu hardy main [18:11] \sh: that will not globally fix our issues since pykdeextensions uses pyuic to build ui files [18:11] \sh: grep -re uic /usr/share/python-support/pykdeextensions [18:11] <\sh> Tonio_: if it works ok...just use the kubuntu_04_kdepyuic_i18n.dpatch from it [18:11] <\sh> Tonio_: we need to fix this too ,-) [18:12] \sh: I'll test [18:12] \sh: I guess it should check for kdepyuic in the first place [18:13] <\sh> Tonio_: it needs a dep on python-kde3-dev ;) [18:13] what needs that dep ? [18:14] <\sh> Tonio_: everything what needs to use kdepyuic [18:14] so if I have python-kde3-dev installed, the setup.py, importing kdedistutils, should use kdepyuic instead ? [18:15] \sh: I already have it installed [18:15] \sh: but pyuic is when launching setup.py [18:17] <\sh> yepp [18:17] \sh: doesn't seem to work... [18:18] the installed .py generated files are done using pyuic (no kde import things in it) [18:18] \sh: I can fix that overwritting the file in debian/rules, but that sounds a bit ugly :) [18:19] <\sh> Tonio_: patch is at http://www.sourcecode.de/~shermann/python-kde3-patches/ [18:20] \sh: testing this right now [18:21] <\sh> pyuic_exe = find_executable('pyuic',pyqt_configuration.pyqt_bin_dir) [18:21] <\sh> hmmm.... === stanz_ is now known as stanza [18:23] <\sh> give me some time to think about a fix... [18:25] \sh: your patch seems incomplete [18:25] I can see the -tr option, but the command doesn't work [18:26] http://pastebin.com/m3ee140d5 === stanza is now known as stanz [18:30] is it possible to download a KDE mailing list archive? [18:31] you can do it with Mailman, but KDE uses MARC for their archives and I can't figure it out yet [18:32] <\sh> Tonio_: pykdeuic -tr i18n ? [18:33] \sh: yep, fails for me [18:33] <\sh> grmpf...give me some minutes..need to wash dishes first ,-) [18:34] \sh: [18:34] shortOptions = "aip:" [18:34] longOptions = "noapp noimport pyuic=" [18:34] \sh: look lik python doesn't want a 2 letters based short option btw [18:34] I guess aiptr: wouldn't be welcome :) [18:37] \sh: I can just test with -t and let you find the fix, gimme a moment [18:39] <\sh> Tonio_: argl...yeah will fix it in a few... [18:40] <\sh> Tonio_: fetch it now from the former location at sourcecode.de === neversfelde_ is now known as neversfelde [18:42] <\sh> argl...adjusting the help :( [18:43] \sh: I was about to ay it ;) [18:43] <\sh> now [18:43] <\sh> ,-) [18:44] k [18:46] \sh: seems to work [18:46] now the second fix to use kdepyuic and it'll be perfect ;) [18:47] \sh: wait no there is a problem, looks like i18n hasn't been passed to pyuic [18:47] I still have __tr based translation in the .py file [18:48] \sh: [18:48] if i18nFunc: [18:48] opts = opts + ' -tr ' + i18nFunc + ' ' [18:48] if addApp: [18:48] opts = opts + ' -x ' [18:48] opts = ' -o ' [18:48] here is the problem, last line should be opts = opts + ' -o' no ? [18:49] \sh: that's it, works this time [18:50] <\sh> Tonio_: bah...lI'm doomed [18:51] \sh: super, I'll test an ugly hack in guidance to override the usage of pyuic by default, I'll wait for another fix on that point :) [18:51] <\sh> well, if the fix works now in general...we should add it to the hardy package... [18:52] <\sh> and try then to fix the other things [18:53] \sh: I tested this on 10 ui files, works perfectly, so I think you can upload right now :) [18:57] <\sh> Tonio_: I can't upload it right now :) please apply the patch to python-kde3 and add it to the 00list file and upload :) [19:01] oki no pb [19:02] <\sh> Tonio_: actually I can't upload in general :) [19:03] \sh: lost your gpgp key ? [19:03] gpg [19:04] <\sh> nope [19:05] <\sh> I'm not a motu or core-dev anymore :) [19:05] \sh: NameError: global name 'i18n' is not defined [19:05] \sh: that's what I get using the generated file, I think we should import something else when using i18n [19:09] <\sh> Tonio_: checking the doc...what needs to be done [19:11] <\sh> Tonio_: from kdecore import i18n [19:11] <\sh> Tonio_: check the examples in python-kde3-doc [19:12] \sh: tested, but that doesn't change anything [19:13] \sh: shouldn't we just import * ? [19:13] \sh: how does this impact performances ? [19:14] <\sh> Tonio_: well, quoting Riddell: it doesn't matter ;) [19:14] <\sh> Tonio_: depends on what you want to do...I think... [19:14] \sh: then we should patch your patch :) [19:14] <\sh> Tonio_: let's see what it does actually [19:15] \sh: well if we want to use i18n, we can either load just i18n (which works in fact), or load everything and ensure it works [19:15] I'd vote for the second option [19:15] dunno if upstream will then accept the patch, that's the problem [19:15] <\sh> well kdepyuic does n.write ('from kdecore import KCmdLineArgs, KApplication\n') [19:15] <\sh> and that's only if we have an app [19:16] <\sh> so I need to add some code magic [19:17] <\sh> Tonio_: thinking about it...it's only be needed if you use i18n() for the locales [19:18] yep [19:18] <\sh> Tonio_: so I could add a switch (instead of the introduced one) like --useKDEi18n which adds i18n to the import line for all generated .py files and it would leave it alone when you don't need it [19:19] <\sh> and adds the -tr i18n to the pyuic call [19:21] yep, that would be the idea [19:21] <\sh> ok...will do it in a few... [19:27] <\sh> Tonio_: should we default it to true or false? [19:27] I'd say yes [19:27] i18n is kde default right ? ;) [19:28] % strace -e trace=open python /var/lib/python-support/python2.5/guidance-power-manager.py 2>&1 | grep LC_MESSAGES [0.11 0.12 - 31% 0%] [19:28] <\sh> ok..default is true...no problem ,-) [19:28] open("/usr/lib/locale/fr_FR.UTF-8/LC_MESSAGES", O_RDONLY) = -1 ENOENT (No such file or directory) [19:28] open("/usr/lib/locale/fr_FR.utf8/LC_MESSAGES", O_RDONLY) = 3 [19:28] open("/usr/lib/locale/fr_FR.utf8/LC_MESSAGES/SYS_LC_MESSAGES", O_RDONLY) = 3 [19:28] open("/usr/share/locale/fr_FR/LC_MESSAGES/libc.mo", O_RDONLY) = -1 ENOENT (No such file or directory) [19:28] open("/usr/share/locale/fr/LC_MESSAGES/libc.mo", O_RDONLY) = -1 ENOENT (No such file or directory) [19:28] open("/usr/share/locale-langpack/fr_FR/LC_MESSAGES/libc.mo", O_RDONLY) = -1 ENOENT (No such file or directory) [19:28] open("/usr/share/locale-langpack/fr/LC_MESSAGES/libc.mo", O_RDONLY) = 10 [19:28] WHY DOESN4T IT WANT TO LOAD THAT STUPID CATALOG FILE ?????????, !!!!!!!!!!! [19:28] that's in the code, but it doesn't work..... [19:28] now I understand why it wans't translated.... [19:28] bad .pot file -> incomplete .po files [19:29] bad .py files not using i18n [19:29] bad kdepyuic [19:29] and finally the catalog loading fails :) [19:29] \sh: it was, at least unlikelly to work ^_^ [19:32] <\sh> Tonio_: catch the patch from www.sourcecode.de/~shermann/python-kde3-patches/ and check if this works better :) [19:32] oki ;) [19:35] <\sh> Tonio_: if it works...I'll prepare a debdiff for python-kde3 [19:35] sure, will do [19:35] atm I'm eating, brb and will let you know === meduxa is now known as toscalix [20:09] <\sh> Tonio_: family is calling ... so send me an email or wait for tomorrow with the patch... === \sh is now known as \sh_away [20:09] nixternal: maybe get the flashgot plugin (for firefox) and go flashgot all at the archive page you want? [20:09] \sh_away: sure [20:21] Riddell: ping ? [20:21] hi Tonio_ [20:21] Riddell: re ;) [20:22] Riddell: I went to the point I can say there is something weird in power-manager [20:22] Riddell: I have a working .po file with french translations for each guidance component [20:22] but power-manager refuses to load ANY kde Catalog [20:22] Riddell: I suspect there is something weird in the program structure [20:23] Riddell: look by yourself : [20:23] ho and btw I don't understand how are the translations loaded [20:23] klocale loads the .mo file [20:23] Riddell: I disabled (removed) the displayconfig insertCatalog line, but the translation is still loaded [20:23] > [20:24] strace can tell you which it tries to load [20:24] Riddell: done and here is the result, you'll be amazed I guess, look at that : [20:25] % grep -re Catalog displayconfig.py [0.01 0.08 - 31% 0%] [20:25] # KGlobal.locale().insertCatalogue("guidance") [20:25] as you can see it is disabled [20:25] % strace -e trace=open python /var/lib/python-support/python2.5/displayconfig.py 2>&1 | grep guidance.mo [0.01 0.07 - 31% 0%] [20:25] open("/usr/share/locale-langpack/fr/LC_MESSAGES/guidance.mo", O_RDONLY|O_LARGEFILE) = 10 [20:25] but it's loaded [20:26] I can see lots of kde catalogs loaded : [20:26] http://pastebin.com/m640b762e [20:26] here's the all list [20:26] with powermanager : [20:27] http://pastebin.com/m261babe5 [20:27] only libc.mo is loaded [20:27] not even kdelibs.mo..... [20:28] Riddell: that's why I suspect some other problem with it [20:28] also I don't understand how can displayconfig load the catalog if disabled [20:28] that's a non-sense for me [20:29] Riddell: so if you have an idea................. that would be nice, since I have done I think all I could to make that to work.... [20:31] Riddell: I am now wondering is there is some kind of cache with python (pyc files already removed), cause I don't understand how a catalog can be loaded without any kind of locale() call.... [20:32] Tonio_: it depends on how it's loaded, e.g. if it's through system settings, or kcontrol or kcmshell [20:32] .pyo files? [20:34] Riddell: pyc, not pyo ? === apachelogger_ is now known as apachelogger [20:35] Riddell: loaded from the shell only, starting the .py file manually [20:35] strace -e trace=open python /var/lib/python-support/python2.5/displayconfig.py 2>&1 | grep guidance.mo [20:36] Riddell: as you can see, just starting the script with python loads the catalog, even if the code doesn't...... [20:36] that's where I get lost [20:38] Tonio_: Do you have any interest in trying to maintain Klamav in Debian? It looks to me like the Debian maintainer has abandoned the package and may be MIA. [20:38] I'm interested to do it, but since you packaged it, I thought I ought to check and see if you were interested first. [20:39] ScottK: I have no upload rights in debian, and to be honnest, bad experience working with a sponsor :) [20:39] Tonio_: ah well there you go [20:40] I'll take that as a no then. [20:40] Tonio_: if you load it through kcmshell etc then it tries to load the .mo files with that associated name [20:40] ScottK: I will still maintain the package in debian [20:40] ScottK: charge for debian to merge ;) [20:40] ScottK: s/in debian/in ubuntu [20:41] Riddell: oki but how can that work if I load it from nowhere, just command line to python myscript.py [20:41] OK. Good enough. I'll let you know if I get something worked out for Debian. [20:41] btw, the mo file name is "guidance" not "displayconfig" [20:41] Riddell: that's just completly crazy for me..... I hate automagic things [20:42] ScottK: sure [20:48] Tonio_: there will be a KApplication() [20:48] that has an app name passed to it [20:48] maybe as part of KAboutData() [20:48] that name is used to load the .mo file [20:48] Riddell: I was just testing this right now [20:49] Riddell: I completly disabled the catalog loading in the all code -> still works, now the about data thing :) [20:50] Riddell: that's it [20:52] Riddell: looks like the all insertCatalog things are useless then [20:52] Tonio_: umm, why? [20:52] Riddell: I'll try to get that fixed tomorrow [20:52] Riddell: they are ignored [20:52] it's still needed for loading through system settings [20:53] hum lemme test [20:53] hum right [20:53] Riddell: all I should do then is adding the correct about infos to power-manager then ? [20:54] ping mhb [20:54] Riddell: but why aren't the other mo files (like kdelibs.mo) loaded ? [20:54] Riddell: not only the guidance.mo is needed [20:55] Riddell: I suspect for example at least the "ok" and "cancel" buttons should be translated no ? [20:55] dunno that [20:55] what do other kde apps load? [21:04] mornfall: ping [21:06] yuriy: semi-pong [21:06] Be quick, I have to go soonish. [21:08] mornfall: ok then i'll skip the more specific stuff. i don't have somewhere to serve a darcs branch off of, so what would be a good way for me to start making changes and commit them somewhere you could easily get to? [21:09] yuriy: I will set up darcs send tomorrow, is that okey? (That would mail the patches to me in an applicable form.) [21:09] mornfall: ok [21:10] yuriy: You could shoot the more specific stuff as well, if you are willing to risk that answers may come only tomorrow. [21:11] Riddell: it works ! [21:11] finally...... [21:12] Riddell: would you suggest a separate po file or patching power-manager to have "guidance" in KAboutdata ? [21:13]