=== claydoh_ is now known as claydoh [02:02] shadeslayer: If it's not on the image we've got tons of time. [07:36] sometimes I am quite the hero [07:37] good morning yall [07:37] one more hour untill demo migration feedback deadline [07:37] apachelogger: only sometimes? [07:45] other times I am unable to debug python [07:47] ovidiu-florin: is the german section going to stay? Because http://test.kubuntu.co.uk/de is completely useless [07:48] if no one will translate it, then it's useless [07:48] yofel: ^ [07:48] I've added it as an example [07:49] and to have more than just 2 languages [07:49] then leave it away. The german community is managed by kubuntu-de.org so you probably won't find anyone that'll bother [07:50] merge the german chapter [07:52] this project is not meant to impose anything to anyone [07:52] just to provide the opportinity to have a better managed community website [07:52] that is also multilingual [07:53] * ovidiu-florin will be back in ~2 hours [07:53] sure, the new website turned out really nice [08:12] * apachelogger rages a bit [08:14] * valorie passes along a cup of tea to apachelogger [08:27] morning [08:28] shadeslayer: yo, that python string problem appears to be on the qstring->pystr conversion side rather than the pystr->qstring [08:29] apachelogger: yes [08:29] because the QComboBox shows it appropriately [08:29] the qstring when arriving in qcombobox is correctly encoded as it were [08:29] yep [08:31] j6 pyqt [08:31] oom [08:32] /home/me/tmp/python-qt4-4.10.4+dfsg/build-2.7/QtGui/sipQtGuipart4.cpp:58495:1: warning: narrowing conversion of ‘(QPaintEngine::PaintEngineFeature)4294967295u’ from ‘unsigned int’ to ‘int’ inside { } is ill-formed in C++11 [-Wnarrowing] [08:32] this also makes me very confident [08:35] shadeslayer: it does not explain this though https://launchpadlibrarian.net/171287077/ole.png [08:35] why would one line be encoded correctly and the others not [08:35] it does not, yeah [08:36] apachelogger: try writing a minimal test case? [08:36] what would that test case be? Oo [08:39] extract romanian string from the po file -> make a QLabel -> display it in a QApplication? [08:39] or for that matter, get the label string and print it? [08:39] what's the point of that? [08:39] to check if it's Qt or something else ? :S [08:40] though I can't say what [08:40] though I can't say what "something else" would be here [08:40] the screenshot shows a line containing the incorrectly decoded characters and a line with the same characters correctly dedcoded [08:40] the line below the buttons where there is a 1, those are the same character as the i without the point in the first label [08:42] shoot [08:42] kde-zeroconf is seeded [08:43] nice, one also cannot gammaray it [08:43] ffs [08:44] ScottK: kde-zeroconf in unapproved [08:45] would be awesome if we could get it in [08:48] apachelogger: https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1182784/comments/25 [08:48] Launchpad bug 1182784 in ubiquity (Ubuntu) "Install with German / Swiss Keyboard fails: "ubi-console-setup failed with exit code 141" or "Installer Crashed"" [High,Triaged] [08:52] shadeslayer: what do I do with it? [08:53] thoughts on how to fix it? [08:55] shadeslayer: sure, talk to xnox [08:55] haven't heard back from him :) [08:57] * shadeslayer downloads ISO to test === shadeslayer changed the topic of #kubuntu-devel to: Kubuntu - So Sweet. | Archive: Frozen for Final | Testing for final in progress http://iso.qa.ubuntu.com/qatracker/milestones/314/builds | https://trello.com/kubuntu | Reports http://qa.kubuntu.co.uk/ | Polo Shirts are available again!! https://holvi.com/shop/Kubuntu/ (Women also) [09:00] oh fancy [09:01] the string is already borked when it arrives in the translate function [09:05] apachelogger: https://www.youtube.com/watch?v=mJ_fkw5j-t0 [09:11] what is this? [09:14] apachelogger: http://bazaar.launchpad.net/~ubuntu-installer/ubiquity/trunk/view/head:/ubiquity/plugins/ubi-language.py#L503 < code drops privs before launching ff [09:14] ah [09:14] interesting [09:14] this is via the dm [09:18] TEXT En iyi sonuç için, bilgisayarınızın şunları sağladığına emin olun: [09:18] TEXT En iyi sonuç için, bilgisayar1n1z1n _unlar1 salad11na emin olun: [09:19] shadeslayer: I know why it bricks [09:19] oh? [09:19] qstring->pystr is broken, it only affects some widgets because the initial setText works just fine (as observed with qcombobox) [09:20] the strings that are broken are processed later (after initial setText) to fiddle RELEASE into Kubuntu etc. [09:20] to do that text() needs to be used to get the previously set translation, now since text exhibits the qstring->pystr problem it returns a shit string [09:21] the post-processing takes the broken string, does the processing, and then calls setText with the broken string ultimately breaking the actual string in the qwidget [09:23] morning! I have two possible patches for https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1038522 a short one, and a larger one which results in a slightly saner infrastructure [09:23] Launchpad bug 1038522 in ubiquity (Ubuntu Trusty) "[kde] manual partitioning in installer crashes when handling many partitions" [High,Triaged] [09:23] I assume we want the short one for 14.04 [09:24] shadeslayer: http://paste.ubuntu.com/7254258/ here's your minimal example [09:24] agateau: yeah [09:25] apachelogger: yeah [09:25] ScottK, Riddell: so, we have fixes for ubiquity (very very very very important) and pam-kwallet (very very very important) [09:27] so my habit of wrapping all strings with QString.fromUtf8 is actually a good thing... [09:28] shadeslayer: proper music for you https://www.youtube.com/watch?v=7GGKS3x5XIY&list=RD7GGKS3x5XIY&index=1 [09:28] tsimpson: wouldn't actually help here :P [09:28] ^^ [09:28] it is however a good habit when you do not know whether a string might be ascii or utf8 [09:30] well I have the coding: utf-8, and I know my editor is actually using utf-8, so I tend to just make a function that just forwards to QString.fromUtf8(arg) whenever I want a QString [09:30] but that's because I don't trust the QString constructor [09:31] apachelogger: sounds important [09:31] apachelogger: what's the fixes? [09:31] the cellists seem to hold the bow weirdly [09:32] ok, some of them seem to hold the bow weirdly [09:32] Riddell: ubiquity fixes crash when there are too many partitions manually created (too many can be 4), pam-kwallet fixes lightdm crash due to too short mallocs [09:33] Riddell: test building the pam stuff right now [09:34] Riddell: zeroconf-ioslave was also updated [09:34] plz accept that too [09:35] shadeslayer: accepted! was there not another one still not at 4.13? [09:35] checking [09:35] ii kdenetwork-filesharing 4:4.12.97-0ubuntu1 amd64 network filesharing configuration module [09:35] Uploading pam-kwallet_0.0~git20140410-0ubuntu2_source.changes: done. [09:35] Successfully uploaded packages. [09:35] Riddell: ^ [09:36] agateau: are you coordinating with xnox on landing the ubiquity fix for trusty? [09:36] apachelogger: not for now, branch is not yet on LP. Riddell is usually the one landing stuff for me on ubiquity. [09:38] Riddell: kdenetwork up as well [09:42] agateau: where's the fix? [09:42] Riddell: on my laptop for now [09:42] Riddell: need to do a last test, but virtualbox is being a nuisance [09:43] let me know if I can help [09:46] Uploading baloo_4.13.0-0ubuntu2_source.changes: done. [09:46] Successfully uploaded packages. [09:46] Riddell: ^ fixes random xapian exception crashery in krunner [09:50] apachelogger: nice [09:51] argh [09:52] * Riddell accepts pam-kwallet, baloo and kdenetwork-filesharing [09:52] shadeslayer: argh? [09:52] xnox: how do I tell ubiquity-dm to start ubiquity with --debug [09:52] I ran it manually with sudo, but I can't reproduce my issue there [10:00] Riddell: so 1285705 is somewhat solved :) [10:00] http://i.imgur.com/rHOasz0.png [10:03] shadeslayer: ooh? [10:03] that doesn't look very solved [10:03] Riddell: well, firefox is the right user [10:09] Riddell: http://paste.kde.org/pzobnah43 [10:09] can you try that patch [10:11] argh wait [10:26] xnox: Riddell http://paste.kde.org/pb7nkpgsb [10:26] makes things work [10:29] shadeslayer: sweet! [10:32] Riddell: just finishing off installing on VBox [10:35] morning sgclark [10:35] Riddell: good morning [10:35] Riddell: xnox https://code.launchpad.net/~rohangarg/ubiquity/ubiquity/+merge/215837 [10:41] actually, this one https://code.launchpad.net/~rohangarg/ubiquity/ubiquity/+merge/215839 [10:42] shadeslayer: committed! [10:42] * Riddell goes to test an upgrade while waiting for agateau's ubiquity fix [10:43] Riddell: would have been nice to get a +1 from xnox too [10:43] but ok :) [10:52] * apachelogger blinks [10:53] sigh [10:53] firefox themeing is still shit, though it's the right user [10:54] well, we setup theming in startkde [10:54] but startkde is not excuted [10:54] so... [10:54] kde-config-gtk? [10:54] we don't set it up in startkde? [10:54] we do [10:54] I'm asking if it's because of the kde-gtk thingy [10:54] I dunno [10:56] pyqt 4.10.4 is broken [10:56] that's why the strings are shit [10:57] and from what I have seen so far it is because it is itering over chars but never increments the pointer, so it evaluates the same over and over again [10:57] also that appears to have arrived in the .4 update [10:57] quite the pointless versioning scheme it appears [11:00] also appears borked in their git snapshot [11:01] yeah, copying over gtk files does the trick [11:02] * apachelogger is smart without realizing it [11:02] apachelogger: /usr/share/kubuntu-default-settings/dot-gtkrc-2.0-kde4 needs copying to ~/ to get the themeing right [11:03] * Riddell had realised apachelogger was smart, does that make me smarter [11:03] just well informed ;) [11:04] fun fact about pyqt [11:04] to get a change in a cpp file adopted you actually need to rebuild everything [11:05] so [11:05] guise, listen [11:06] the python string<-qstring conversion is kaput [11:06] possible courses of action: a) revert to what it was before pyqt 4.10.4 which apparently has some conversion issues with certain encodings b) fix the shitty code (alas, I am not yet sure how) and make all conversions possibly work [11:07] I'd argue that a) makes more sense for a hotfix [11:09] hellooooo [11:12] I'm waiting for a reply from PyQt guys [11:12] I am not [11:12] if they don't reply by evening, go with a) [11:12] unjelss you want to block the release [11:12] because I am not releasing with an installer that has fucked localization [11:13] apachelogger: btw apart from hard coding /usr/share/kubuntu-default-settings in ubiquity , do you know of a better way to copy .gtkrc-2.0 > [11:13] shadeslayer: no, you need to hardcode something I reckon [11:13] hm okay [11:14] i.e. you could extract the copy logic etc from startkde put it in a standalone script, call that from both, but you still need to hardcode the script path [11:14] I just installed on my shiny new SSD hard disk and now the install is too fast to watch the slideshow, is this a release critical bug? [11:14] or you put something in bin/ [11:14] though that seems inappropriate [11:17] shadeslayer: please prep an upload http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/trusty/python-qt4/trusty/revision/1.8.1#qpy/QtCore/qpycore_qstring.cpp [11:18] the two line removal at 57/58 is what we want instaed of the present ifdef [11:18] the present ifdef comment also explains what the problem with the previous code was ^^ [11:19] Riddell: https://code.launchpad.net/~rohangarg/ubiquity/ubiquity/+merge/215844 [11:19] I do now also have a proper fix I reckon [11:19] En iyi sonuç için, bilgisayarınızın şunları sağladığına emin olun: [11:19] much magic [11:21] apachelogger: if you have a proper fix, just upload a patch? [11:21] not without review by upstream [11:22] show patch? [11:22] preping [11:23] xnox: https://code.launchpad.net/~rohangarg/ubiquity/ubiquity/+merge/215844 [11:23] I am working on rather nonexistant understanding of what the code is supposed to do and simply extrapolate what would make sense from the surrounding code [11:24] also it's very shitty code IMO [11:25] it loops in a loop, except the inner loop only gets run when a !ascii character is encountered and the results of inner will abort the outer [11:25] shadeslayer: how does that fix bug 1038522 ? [11:25] bug 1038522 in ubiquity (Ubuntu Trusty) "[kde] manual partitioning in installer crashes when handling many partitions" [High,Triaged] https://launchpad.net/bugs/1038522 [11:25] Riddell: it doesn't, Launchpad is being weird [11:26] I don't know why it thinks it does [11:27] Riddell: http://imgur.com/d4D9znk [11:29] shadeslayer: http://paste.ubuntu.com/7254739/ please be posting to the upstream thread on the topic [11:30] apachelogger: ok [11:31] general description: the previous logic didn't do no nothing, the changes make it continue iter qch to the end and adjust size values accordingly, qch gets reset later on before actually converting the data [11:32] apachelogger: already senrt [11:33] * shadeslayer is hungry [11:34] shadeslayer: http://paste.ubuntu.com/7254756/ [11:34] adjusted test clearly showing the issue [11:34] cool beans [11:35] shadeslayer: merged! [11:35] cheers [11:35] that's one whole bug down :P [11:35] agateau: how are you getting on? [11:35] Riddell: yelling rude words at virtualbox [11:37] apachelogger: oh https://translations.launchpad.net/ubuntu/trusty/+source/kubuntu-driver-manager/+imports [11:37] https://translations.launchpad.net/ubuntu/trusty/+source/kubuntu-driver-manager/+pots/desktop-kubuntu-driver-manager [11:37] neat [11:38] shadeslayer: u looking at the l10n bugs? [11:38] apachelogger: not quite, I had it open in my browser from yesterday and realized the bug had a url [11:38] lol [11:38] and clicking it allowed me to find the translations page [11:38] shadeslayer: what you doing right now? [11:39] apachelogger: just finished off ubiquity buggery, looking at bug list for next bug [11:39] and keeping an eye on https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/1294666 [11:39] Launchpad bug 1294666 in xserver-xorg-video-intel (Ubuntu) "[HSW mesa kde needs Xorg-1.15.1] Multiple tiling-esque artifacts in KDE" [High,Fix committed] [11:40] shadeslayer: pick the l10n ones :P [11:40] xD [11:40] apachelogger: I don't know what to do there tbh [11:40] when imported -> mark language team task fix released -> open new task against language-o-matic and poke pitti to look into adding them asap [11:40] ok [11:40] if a thing is not imported tell me so I can forward poke [11:42] apachelogger: https://bugs.launchpad.net/langpack-o-matic/+bug/1267763 [11:42] Launchpad bug 1267763 in langpack-o-matic "kde-config-whoopsie not localized" [Undecided,New] [11:43] apachelogger: pokery required https://bugs.launchpad.net/ubuntu/+source/kdesudo/+bug/1267765 [11:43] Launchpad bug 1267765 in kdesudo (Ubuntu) "kdesudo not localized" [High,Triaged] [11:44] shadeslayer: looks imported to me? [11:45] what do I do with whoopsie? [11:45] oh [11:45] apachelogger: I don't see them imports https://translations.launchpad.net/ubuntu/trusty/+source/kdesudo/+imports [11:45] oh [11:45] unless [11:45] right [11:45] intltool [11:46] what? [11:46] if there is nothing on the imports stuff, it should all be fine, just click translations so check [11:46] https://translations.launchpad.net/ubuntu/trusty/+source/kdesudo [11:46] s/so/to [11:46] right [11:47] there is a buttload of gst related crashes in amarok [11:47] there's no day going by with at least 2 distinct crashes [11:48] Riddell: finally https://code.launchpad.net/~agateau/ubiquity/kde-partman-fix-model-updates2/+merge/215849 [11:49] also anyone noticed cropped elements in sidebar, like this? http://agateau.com/tmp/ubiquity-cropped-sidebar.png [11:49] 'Morning folks [11:49] https://errors.ubuntu.com/problem/0386be8deb506b2c7d8df703b4ff064faf7729e6 [11:50] I am pondering catching all and everything everywhere [11:50] just because exceptions are so shitty [11:50] ::runtime-bugs:: [1299689] kwalletd crashed with SIGSEGV in atomic_exchange_and_add() @ https://bugs.launchpad.net/bugs/1299689 (by terry b) [11:50] agateau: no I've not seen cropped elements in sidebar like that [11:50] 0.o [11:51] apachelogger: those amarok crashes seem to be in gst? [11:52] yeah, except gst also likes to crash when you are using it wrong [11:52] since C doesn't do exceptions, the gst devs thought to themselfs, why not simply crash [11:54] Riddell: the bug goes away if I maximize the window, so maybe it's just a matter of making the window small enough [11:56] apachelogger: all done [11:56] down to 10 bugs [11:57] agateau: did you fix 1307291 too? [11:57] shadeslayer: I think they are the same [11:58] aha [11:59] ovidiu-florin: http://test.kubuntu.co.uk/download-kubuntu seems broken [11:59] scroll halfway through -> theme changes color [12:00] shadeslayer: noted [12:00] ovidiu-florin: http://imgur.com/Nfy5m7T [12:01] layouting also seems a bit broken [12:01] shadeslayer: saw it, no need for the screensot [12:01] ;) [12:01] I see :) [12:02] said the blind man to his deaf son... :P [12:02] * apachelogger falls off chair [12:03] apachelogger: what's wrong :S [12:07] shadeslayer: forum threads [12:07] most terrible to read [12:09] nothing interesting going on with the crashery [12:09] except ark's abortion problems being somewhat more excessive than I thought [12:11] * agateau goes off for an errand [12:13] apachelogger: https://www.kubuntuforums.net/showthread.php?65120-I-honestly-think-you-ought-to-sit-down-calmly-take-a-stress-pill&p=348896&viewfull=1#post348896 [12:14] I did not read that [12:14] wall of text are not meant to be read I believe [12:15] gosh baloo is fast and accurate on a new install with an SSD :) [12:15] jose: are you around? [12:17] shadeslayer: did you prepare a pyqt4 revert upload? [12:17] apachelogger: no, I'd rather wait for a couple of hours to hear back on the PyQt mailing list [12:18] I don't want you to upload, but have something ready :P [12:18] ok [12:19] shadeslayer: what's the story with pyqt? [12:19] le broken [12:20] in what way? does it need to get on the iso? [12:20] (could have been entirely prevented by having unit tests) [12:20] Riddell: yes [12:20] apachelogger: as a dependancy of what? [12:20] bug 1299881 [12:20] bug 1299881 in ubiquity (Ubuntu) "[kubuntu] Strange characters in different locale" [Undecided,Confirmed] https://launchpad.net/bugs/1299881 [12:20] ovidiu-florin: dependency? [12:21] most things could be prevented by having unit tests [12:21] that thing should have had a unit test [12:21] or at least QA [12:21] pyqt is the library that provides the ability to code in python using Qt, right? [12:21] but apparently pyqt is meant to be used with ascii characters to begin with [12:21] ovidiu-florin: yes [12:22] apachelogger: then do we also provide tools for development on the ISO? [12:23] or it is necessary on the iso because it's a dependency of something [12:23] ? [12:23] dependency [12:23] of a pile of things [12:23] ubuntu likes to write stuff in python for the oddest of reasons [12:24] so we end up needing to write stuff in python short of duplicating efforts [12:28] apachelogger: I'll prep a upload after lunch [12:29] Riddell: the full migration is done. I just need to do some category cleanup [12:29] Where can we place the logo of the sponsor? [12:31] ovidiu-florin: at the bottom of the donations page? http://test.kubuntu.co.uk/donate [12:31] I had that in mind [12:32] I'll make another Heading called Sponsors [12:32] We can put there Cms2Cms and Blue Systems [12:33] and canonical too to be fair :) [12:34] ok [12:34] :D [12:34] maybe even limux [12:34] I need their logos [12:34] and what link to point them to [12:37] google knows [12:44] Riddell: http://test.kubuntu.co.uk/news [12:45] ovidiu-florin: http://test.kubuntu.co.uk/news/page/2 not work [12:45] from the "older entries" link [12:47] shadeslayer, apachelogger: should I respin images now or wait for the pyqt fix? [12:47] Riddell: wait IMO [12:47] at the very least until shadeslayer returns from lunch [12:47] should I be worried that after an upgrade I got a crash from nepomuk? [12:48] what/ [12:48] ? [12:48] what what? [12:48] what crash [12:48] Riddell: upgrade from 13.10? [12:48] yep upgrade from 13.10 [12:49] oh great now apport doesn't want to show me what it was [12:49] after reboot or before? [12:50] after reboot [12:50] :O [12:50] Riddell: did you send that to errors.ubuntu? [12:51] alas no, it disappeared when I clicked on apport [12:51] Oo [12:51] Riddell: kcmshell4 whoopsie [12:51] click on previous reports [12:51] is anything listed there? [12:51] first box is ticked [12:51] there is a link at the bottom [12:51] "previous reports" === soee_ is now known as soee [12:52] baloo file broke https://errors.ubuntu.com/oops/c4f5e0cc-c498-11e3-8cbe-fa163e373683 [12:53] https://errors.ubuntu.com/problem/b1d1599689351bfbab08ad927e106614cf44bde4 [12:54] Riddell: I don't think that URL is wrong [12:54] because you would never type in that URL [12:54] vHanda: http://paste.ubuntu.com/7255118/ this is a new creash with 4.13.0 and Riddell just encountered it [12:54] have fun :P [12:54] you would use the link from the bottom of the page [13:00] ovidiu-florin: pardon? [13:00] Riddell: the link you gave me (http://test.kubuntu.co.uk/news/page/2 ) is a generated link by the CMS that points to the second page [13:01] that link will show different things [13:01] I'll try to get it to work, but IMO it's useless to do so [13:02] ovidiu-florin: why is it useless? the "older entries" link on http://test.kubuntu.co.uk/news seems quite important? [13:02] yes [13:02] go to the bottom of that page [13:02] you'll see a link: Older posts [13:03] yep [13:03] which takes me to the broken page [13:03] what? [13:03] let me check that === greyback_ is now known as greyback|lunch [13:06] apachelogger: urgh. I don't get where these exceptions keep coming from. [13:08] vHanda: perhaps Riddell can reproduce it [13:08] the backtrace certainly isn't much help [13:08] Riddell: if you can please ping me. [13:08] ping vHanda [13:08] Riddell: or, maybe the startkde.log has something about it? [13:08] oh if I can recreate it? [13:09] Riddell: yes. [13:09] unlikely, happened shortly after I logged in, I'll try logging out and in again [13:09] but it was after an upgrade from 13.10 with nepomuk so it might have to do with the nepomuk migration [13:09] Riddell: check yer startkde.log [13:10] if you can please run it under gdb and type 'catch throw' [13:10] Riddell: it's the baloo_file process that is crashing, not the migrator. But you never know. [13:10] http://starsky.19inch.net/~jr/tmp/startkde.log [13:10] also, could you please check to see if baloo_file automatically restarted itself [13:10] terminate called after throwing an instance of 'Xapian::DocNotFoundError' [13:10] "baloo_file(1751): Indexer crashed while indexing 516" [13:10] after crashing? [13:10] there, I am hero [13:11] vHanda: yep now running jr 13437 1166 1 13:23 ? 00:00:37 /usr/bin/baloo_file --nocrashhandler [13:11] apachelogger: thanks, that helps a little. [13:13] vHanda: you should write a macro XAPIAN_CALL(x) try { x } catch (Xapian::Error&) {} [13:13] ;) [13:13] apachelogger: I've written a wrapper library over most parts of xapian [13:13] :O [13:15] Riddell: found it. [13:15] apachelogger: thanks, that helped [13:19] apachelogger: what do you propose as the version for the rollback of pyqt4 === greyback|lunch is now known as greyback [13:19] 4.10.4really4.10.3+dfsg1 ? [13:20] apachelogger: any chance you can get this one into 14.04 as well? [13:21] shadeslayer: you don't roll back [13:21] shadeslayer: you selectively patch back that one function [13:21] oh? [13:21] oh [13:21] okay [13:22] vHanda: if you give me a patch probably [13:22] depends on whether Riddell accepts the upload or not ^^ [13:24] apachelogger: so roll back qpycore_qstring.cpp to python3.3 ? [13:25] erm [13:25] rollback to version 4.10.3 ? [13:26] shadeslayer: mh, yeah, I posted a rev earlier [13:26]