[00:00] lol [00:00] I didn't notice the pycairo changes had been done [00:00] I would have uploaded the pygtk changes otherwise [00:00] ember: why? [00:03] they can be accepted after beta if those are stable updates [00:05] liferea is, and has a bug fix for 1.0 -> 1.4 cache updates, which is interesting for dapper->hardy upgrades [00:06] I think I'll call to Dell tomorrow to make sure everything's going fine ;) [01:50] anyone of a software application that records a phone conversation utilizing the internal modem on a PC? [01:50] I hope this is the correct channel to post this question [08:24] hey hey mvo [08:26] hwy seb128 [08:27] (that was a bavarian hey) [08:27] ;-) [08:52] seb128: http://launchpadlibrarian.net/12662959/buildlog_ubuntu-hardy-i386.gtk%2B2.0_2.12.9-2_FAILEDTOBUILD.txt.gz <--- something is unhappy... known problem? :) [08:53] kwwii: did human-theme build for you locally? http://launchpadlibrarian.net/12625757/buildlog_ubuntu-hardy-i386.human-theme_0.13_FAILEDTOBUILD.txt.gz [08:54] kwwii: "error: can't copy 'build/share/themes/Human/index.theme': doesn't exist or not a regular file [09:11] Amaranth: can you reproduce bug #99508 ? [09:11] Launchpad bug 99508 in compiz "Window titlebar displayed not right with compiz enabled" [Medium,Confirmed] https://launchpad.net/bugs/99508 [09:12] hmm. I was sure I registered a bug regarding the new clock applets "Adjust date and time" menu item not using the default dialog for doing so in Hardy, but I can't find it.. [09:13] is the old clock applet still available in hardy, does anyone know? [09:13] good morning [09:14] hey dholbach [09:15] hey mvo [09:16] Amaranth: meh, I can not reproduce #150702 either [09:46] lool, slomo: do you know if public modules using python-support are available during upgrade now? [09:47] seb128: no, i know nothing about the new python policy and everything related to it... i'm trying to not touch packaging of such stuff and if i have to i'm looking at other packages ;) [09:49] seb128: I don't know [09:50] I got complain from pygtk upstream than pygtk is broken during upgrades [09:50] and I think the new pycentral fixes this issue and pysupport is still buggy [09:50] I'm pondering if we should switch to pycentral for hardy [09:50] seb128: I would think not as I see update-python-modules -c in prerm [09:50] lool: not what? [09:50] seb128: But then python-support is a bit special [09:51] seb128: I would think it's not available [09:51] Would it be available, it would be in postrm [10:17] seb128: I see libglib-perl has been uploaded in Debian [10:18] lool: a new version you mean? something we want to use? [10:20] I guess so [10:21] seb128: libglib-perl (1:1.180-1) unstable; urgency=low [10:21] . [10:21] * New upstream release [10:21] That's all [10:21] ok [10:21] will be for after beta then [10:21] Sure [10:22] slomo: btw you might want to use the shared-mime-info change I did yesterday for debian [10:25] seb128: will do, thanks === pedro__ is now known as pedro_ [11:05] asac: the firefox restart required informatio is displayed for me unconditionally even if ff is not running. is that a knwon issue? [11:07] mvo: can you see whats wrong? [11:07] (no its not known) [11:07] asac: I think the firefox-3.0-restart-required is lacking the "DisplaIf" line AFAICS [11:08] strange. i took it from the old package [11:09] mvo: how should it read? [11:09] I can have a look after lunch [11:09] DisplayIf: $STARTUPWMCLASS ? [11:09] i thnk i found it: DisplayIf: ps -C firefox-bin [11:10] asac: yes, thats it [11:10] asac: probably with a ps -C firefox-bin >/dev/null to avoid spittng stuff to ~/.xsession-errors [11:11] thanks [11:11] thank you ! [11:12] mvo: is there a pattern that also takes into account that firefox is running for the current user? [11:13] asac: update-notifier runs with the same uid as the logged in user, so we need to limited the ps to the users processes and we are good [11:13] ok [11:14] mvo: you have an expression at hand (its friday and i feel botty :)) [11:14] ok fixed with the above in bzr [11:14] asac: give me a sec (friday for me too ;) [11:15] asac: pgrep firefox-bin -u $USER [11:16] that works [11:16] or pgrep firefox-bin -U $(id -u) [11:16] :) [11:16] ok [11:17] yeah the last one looks good [11:18] thanks [11:18] cheers [11:19] fixed in bzr [11:30] mvo: i would like to discuss if its possible to make firefox/xulrunner upgrades safe in intrepid and if we think it might be possible discuss this at UDS. [11:31] mvo: you think we can do a pre-call to discuss some ideas i have .... lets say next week? [11:37] asac: "safe"? [11:42] seb128: can sync shared-mime-info from debian now [11:42] slomo: ok, will do after the beta freeze [11:47] Hobbsee: yes its not safe atm :) [11:49] asac: i'm trying to figure out what you're defining as "safe", though [11:57] no crashes ... no unexpected behaviour and so on [11:57] (that do eventually happen - in case you had been lucky and never experienced that :)) [11:59] ah right [11:59] yeah, true [11:59] not a trivial thing to fix though === cody-somerville_ is now known as cody-somerville [12:40] I wished somebody killed scrollkeeper for good [12:53] dholbach: why? [12:53] the update runs take ages on my laptop :) [12:53] dholbach: I think rarian compat thing is slower [12:54] really? [12:54] I did try by then and that's one of the reason why I didn't do the switch [12:54] yes, the compat is not optimized, the goal is to ride of it [12:54] to get rather [12:55] that'd be the best thing since sliced bread [12:55] ;-) [12:58] asac: sounds great [12:58] asac: (sorry, lunch took longer) [12:58] hmmmm lunch [13:01] pitti: do you do freeze exception or that's only steve now? [13:03] seb128: afaik, anyone can do them [13:03] Hobbsee: anyone? [13:03] seb128: well, anyone in -release [13:04] Hobbsee: that includes you? ;-) [13:04] seb128: in some mutant sense, yes. [13:04] * seb128 hugs Hobbsee [13:05] * Hobbsee hugs seb128 bakc, and waves her antlers around [13:06] seb128: it's the kind of mutant-cant-do-much-due-to-no-drescher-access, so.... [13:06] * Hobbsee may or may not be therefore useful. [13:08] Hobbsee: that's alright, I'm just going to do some bug fix upload and I was wondering if they were likely to be accepted today [13:08] seb128: tbh, i'm not overly comfortable doing high-impact accepts, as i'm aware that a lot of the release decisions get taken in private areas, which i don't have access to. [13:09] seb128: i had proof of that after UDS, and, afaik, it still happens, so [13:09] I don't think that's a true statement [13:09] seb128: it's safer not to work with obsolete info. [13:17] anyway I've no high impact uploads [13:17] I've just uploaded a one liner fix to gvfs which is waiting [13:17] and the patch is from the upstream svn [13:18] seb128: in unapproved? [13:18] seb128: dunno, in wherever things we upload during a freeze land ;-) [13:19] seb128: approved. [13:19] * Hobbsee waits to see if this all still works [13:20] grrr [13:20] * Hobbsee grumbles at this bug, just as she did last cycle. [13:21] Hobbsee: thanks [13:21] seb128: you're welcome [13:21] and a timeout. [13:26] the workspace switcher applet thing is still broken in hardy [13:26] NARF! [13:27] so no compiz for me on my laptop [13:41] dholbach: how is it broken now? [13:42] the same as before [13:42] hang on, I'll post a screenshot [13:44] http://daniel.holba.ch/temp/compiz-b0rk.png [13:45] dholbach: iz undocumented feature. [13:45] dholbach: it's now up to you what you would like to see in that bottom box :) [13:45] Hobbsee: ahhhh nice :) [13:45] dholbach: reset your compiz config [13:45] (MacSlow), mvo, seb128: ^ [13:46] dholbach: gconftool --recursive-unset /apps/compiz [13:46] OK... here we go again [13:47] dholbach: not the first time you do that? ;-) [13:48] seb128, mvo: http://daniel.holba.ch/temp/compiz-b0rk2.png [13:48] new variation :) [13:48] dholbach: cat .config/compiz/compizconfig/Default.ini [13:48] daniel@lovegood:~$ cat .config/compiz/compizconfig/Default.ini [13:48] cat: .config/compiz/compizconfig/Default.ini: No such file or directory [13:48] daniel@lovegood:~$ [13:48] dholbach: what do you have in .compiz and .config/compiz? [13:50] daniel@lovegood:~$ find .config/compiz/ [13:50] .config/compiz/ [13:50] .config/compiz/compizconfig [13:50] .config/compiz/compizconfig/config [13:50] daniel@lovegood:~$ find .compiz/ [13:50] find: .compiz/: No such file or directory [13:50] daniel@lovegood:~$ [13:50] daniel@lovegood:~$ cat .config/compiz/compizconfig/config [13:50] [gnome_session] [13:50] profile = [13:50] plugin_list_autosort = true [13:50] daniel@lovegood:~$ [13:51] dholbach: my, bugger, can you please run ccsm [13:51] and go to general options [13:51] desktop size [13:51] and check the "number of desktops" property there? [13:51] mvo: where should that be written? [13:51] in gconf [13:51] mvo: and I though that is was not possible to have a different setting that 1 workspace now? [13:51] and it deactivated my Ctrl-< key for open terminal [13:52] seb128: yes, I want to make sure that dholbach has it too [13:52] mvo: what key is the corresponding one? [13:52] seb128: it should not be possible anymore, but the screenshot suggests that it is still for some reason :/ [13:52] mvo: I suspect something like your issue this week, wrong gconf default to etc or something [13:52] horizontal 3, vertical 2, number of desktops 6? [13:52] does that make sense? [13:53] it was at 2, 1, 1 [13:53] oh, I can't change number of desktops [13:53] dholbach: but it is still 6 even though you can not change it? [13:53] no, it's 1 [13:53] mvo: what is the gconf key involved? [13:54] dholbach: grep compiz /etc/gconf/* -r [13:54] seb128: nothing [13:54] bah, ok, I'll let mvo debug it [13:54] seb128: it should be "number_of_desktops" in /apps/compiz/general/screen0/options [13:55] also ccsm gives me GConf backend: There is an unsupported value at path /apps/compiz/plugins/scale/allscreens/options/initiate_edge. Settings from this path won't be read. Try to remove that value so that operation can continue properly. [13:55] GConf backend: There is an unsupported value at path /apps/compiz/plugins/scale/allscreens/options/initiate_edge. Settings from this path won't be read. Try to remove that value so that operation can continue properly. [13:55] dholbach: thanks! the vnumber_of_desktops 6 is the culprit, it should no longer be possible, but apparently it is, it looks like a bug in compiz to me [13:55] dholbach: you have restarted or loged in loged out? [13:56] mvo: I'll remove all the settings again, then re-login, then switch to compiz again [13:56] let's see what happens [13:57] dholbach: I had hoped the limited the number of desktops would fix it. oh well [13:57] mvo: you should remove this if [ $user == "dholbach" ] joke now ;-) [13:57] haha [13:57] brb [14:00] mvo: no dice [14:01] same as in b0rk2.png [14:01] anyway... I'm going out for lunch now :) [14:03] dholbach: does it happen with an another user? [14:03] dholbach: enjoy your lunch [14:42] dholbach: did you ever figure out the problem with the cookie stuff? [14:53] bug 150702 [14:53] Launchpad bug 150702 in compiz "alt shift tab stopped navigating windows (gutsy)" [Low,Confirmed] https://launchpad.net/bugs/150702 [14:53] mvo: i can preproduce the titlebar thing, sometimes [14:53] haven't tested recently but a bunch of other people have :P [14:54] only happens to maximized windows and only happens with clearlooks-based themes [14:54] and i think it doesn't happen to the new clearlooks so only human, unless human has been updated as well [14:58] seb128: I'll try later on === dholbach_ is now known as dholbach [15:04] <_MMA_> seb128: Has the little GDM drum sound vanished for you on Hardy? Has here and On Ubuntu Studio. I'm still looking into it but I figured Id ask if its known or just me. [15:10] good afternoon slomo [15:13] pitti: say, does jockey have a list of pciids that fglrx works with? === onestone_ is now known as onestone [15:23] _MMA_: before you got your CDs made 'officially' how were you making them? UCK? [15:25] <_MMA_> Amaranth: The old fashion way. You would have to talk to joejaxx for exact details. But is wasnt with any app. [15:26] _MMA_: did it involve modifying an existing disc or did you make one from scratch? [15:26] <_MMA_> An Alt disk from scratch. [15:26] ah, alt disc [15:26] yeah, those are easy :P [15:27] <_MMA_> Suuurre. :) [15:38] seb128: I can help out for easier cases [15:38] Amaranth: no, jockey itself does not have such lists; they are shipped by l-r-m [15:38] yeah, i found that out [15:39] but they don't look like something i can use [15:39] if you're using a laptop with a chip supported by fglrx i want to block compiz loading if you're using ati [15:40] Amaranth: I don't quite understand why you would do this? [15:41] pitti: because those chips have problems with compiz and the ati driver [15:41] Amaranth: shouldn't you rather test for models where compiz works with the free ati driver? (independent of fglrx)? [15:41] random lockups [15:41] the only ones that seem to work right are the ones old enough that fglrx no longer supports them [15:42] right now i've just got all laptops using the ati driver blocked [15:42] but shouldn't this be a pretty static set? [15:42] it's a pretty large set too :P [15:42] and i don't know what all of them are [15:43] so i need to blacklist a bazillion pciids or whitelist a bazillion pciids [15:43] i'd like to have one of those lists autogenerated, if i can [15:48] right; however the autogeneration needs to have a solid criterium as well [15:48] what do you mean? [15:48] so at one point you have to input external knowledge [15:48] (like testing results, or feature tests, etc.) [15:48] If you're on a laptop and fglrx supports your chip I want to block compiz from starting if you're using ati [15:49] until fglrx drops support for more cards this should work fine [15:49] I thought about doing a feature check to see what series of chip you had but the xpress 200 is one of the broken ones and it doesn't have shaders [15:49] so I ran out of things to chck [15:50] check* [15:50] Amaranth: so, for that you could check whether /usr/share/linux-restricted-modules/2.6.24-12-generic/modules.alias.override/fglrx has the graphics card [15:50] but that's a very ubuntu specific hack [15:50] not an upstream solution [15:50] Yeah, but so was my last hack :P [15:51] the only problem is i don't know how to turn that file into something i can use :P [15:51] what do i read to compare to it? our current system is based on pciids as reported by lspci -n [15:51] right [15:52] grep -q 'pci:0000$VID:0000$PID' /usr/share/linux-restricted-modules/2.6.24-12-generic/modules.alias.override/fglrx [15:52] somehting like that [15:53] the better way is of course something like printf("%08X", vid) [15:53] oh, duh [15:53] Amaranth: those are modaliases, and for pci they are composed from vendor ID (v), product ID (p), subvendor (sv), subdevice (sd) [15:53] and device class, subclass, and interface [15:54] Amaranth: this entire hack is only necessary because both the fglrx and the nvidia kernel module do not have proper modaliases [15:54] yeah [15:54] they just always try to load, don't they? [15:54] Amaranth: e. g. if you look at "modinfo b43", it has [15:55] but modinfo nvidia ist just a catch-all [15:55] ooh [15:55] modinfo fglrx -> that actually looks sensible with the latest version [15:55] before it didn't have any [15:55] \o/ [15:55] hey, that looks like your list :P [15:55] yay, so we can remove the list frim l-r-m [15:56] Amaranth: right, except that the nvidia and fglrx lists from l-r-m are created by some hackish parsing of the shared libraries :) [15:56] Amaranth: modinfo fglrx | grep -q is at least a hack which is not distro dependent [15:57] yeah, i remember when you first wrote this it didn't detect my chip and i got a different list when i ran the script than you [15:57] right, that's why we now have per-arch lists [15:57] (in gutsy, too) [15:57] pitti: except in can't guarantee fglrx (or l-r-m) will be installed so i'll probably just turn this into something i stuff in the compiz package [15:58] Amaranth: true that [15:58] Amaranth: but at least it's a script which you could run when you wrap a release [15:58] just run a bit of regex magic over it to turn it into a regular pciid list so i don't have to write more code :) [15:58] it's just a handwavy heuristics anyway [15:58] oh, but i need separate arch versions [15:58] hrm [15:58] right, that should be easy [15:59] sure, except seb128 thinks we already call too many things in this shell script :) [16:00] modinfo fglrx | perl -ne 'if (/^alias:.*pci:v(\d+)d(\d+)/) { print "$1 $2\n" }' [16:01] erm, not \d [16:02] [0-9A-F] of course [16:02] anyway, you get the idea [16:02] Amaranth: per-arch> well, as you said it's only a handwavy hack, not a precise whitelist anyway [16:04] modinfo fglrx | perl -ne 'if (/^alias:.*pci:v0000([0-9A-F]+)d0000([0-9A-F]+)/) { print "$1:$2\n" }' [16:04] Amaranth: anyway: http://paste.ubuntu.com/5687/ <= amd64 list, for your comparison with i386 [16:04] that should get me a list of pciids in the format lspci returns, right? [16:04] right [16:04] lspci -n | grep -w 0300: [16:04] ^ graphics cards only [16:06] ah, they are the same thing [16:07] same list on i386 and amd64 for fglrx [16:37] isn't gpm to request password after suspend? [16:56] ember: no, gnome-screensaver locks the screen on resume [16:56] or should === Ubulette_ is now known as fta [17:33] wtf, gnome-screensaver was not installed anymore [17:35] put it out of its misery [17:48] re [17:49] pitti: sorry I was away, you were looking for work to do? ;-) I guess it'll be for next week now [18:04] anyone have an idea of how to fix compiz when it stops working? [18:19] 203 bugs in compiz [18:19] so close [18:19] goal was under 200 by the end of today [18:20] Amaranth, You can do it! :) [18:21] I can get one of them into "Fix Committed" state [18:21] but I dunno how to get rid of 3 others [18:24] Amaranth: randomly reassign to libwnck as you do usually? ;-) [18:24] haha [18:24] hey, i _fixed_ all the bugs i assigned to libwnck :P [18:25] Amaranth: btw do you think you could look at bug #118936 before hardy? [18:25] Launchpad bug 118936 in alacarte "Alacarte does not recover deleted menu items" [High,Triaged] https://launchpad.net/bugs/118936 [18:25] Amaranth: right, thanks for that ;-) [18:25] seb128: that's a tricky problem [18:25] gnome-menus offers no way to get items are are Hidden [18:25] and I mark things as Hidden to "delete" them [18:25] isn't the menu reset just a rm .config/something? [18:26] or wouldn't that work rather [18:26] well no in this case i've changed the .desktop file [18:26] and you might have .desktop files in ~/.local/share/applications that alacarte didn't put there [18:26] right [18:26] nautilus uses that too ;-) [18:26] for mime stuff, yes [18:26] and WINE uses it [18:27] ok, so not as trivial as I though$ [18:27] although i think all the files i put there are called "alacarte-madeX.desktop" [18:27] ah [18:27] i'll have to look again, i know that was the idea [18:27] in which case that's easy to delete those [18:27] to work around a couple bugs [18:27] in gnome-menus :P [18:28] I don't think the menu editor is that important [18:28] but I was trying to figure bugs that would be nice to fix for hardy [18:28] i could just remove all alacarte-made items [18:28] and that's the only alacarte one which seemed to be annoying [18:28] it might not recover everything but it should work 99% of the time [18:28] better than current ;-) [18:31] i think the rest of the alacarte bugs are it not responding well to your .menu files being broken [18:31] which means you're in trouble anyway [18:34] seb128: can you tell to your big chief to buy laucnhpad.net and all approximative domains around launchpad; [18:34] ah, no, that won't work [18:34] my fingers thanks you [18:34] to "delete" something I have to have it be named the same as the original [18:35] so no easy fix [18:36] crevette: ;-) [18:37] Amaranth: alright [18:37] seb128: are you aware of http://paste.ubuntu.com/5693/ ? (and/or the right person to talk to :) ? [18:37] probably requires an ABI breaking patch to gnome-menus [18:37] * seb128 teachs crevette how to use bookmarks [18:38] down to 202, wonder who closed another bug [18:38] * Amaranth guesses pedro_ [18:38] just koffice-dev and libgphoto2-2-dev file overwrite problem for most of hardy, that is pretty impressive [18:38] mvo: i've got a better fix for ati laptops [18:38] well, i have all the stuff i need to code the better fix, just need to do it :) [18:39] also, are you going to wait until after beta to update to git? [18:39] mvo: no, I'm not but I'm happy to fix it, pitti does update this package usually [18:42] uhm, it's just me or there are problems with compiz+nvidia and the notification area? [18:43] seb128: I'm happy to do it, but I need to leave now for the evening so I can do it either tonight (late) or monday [18:43] lapo: just you? [18:43] mvo: ok, I can do it if you want [18:43] please don't file a bug, i'll never get below 200 :P [18:43] eheh, ok, it's a known bug :-) [18:43] seb128: also, i know how you feel about libwnck, calc sends me all OOo bugs that he can't reproduce when the reporter is using compiz :P [18:44] Amaranth: ;-) [18:44] one of them got upstreamed twice [18:44] I used to do that too with libwnck bugs [18:44] but you kicked all those back :-p [18:44] i sent the bug to compiz-fusion and they send it to OOo [18:45] OOo is really cool, it requests 200x200 windows for dialogs much larger than 200x200 [18:45] and requests that they are not resizeable [18:47] seb128: yeah, go ahead if you have time for it, if not, no problem, just let me know [18:49] mvo: ok, I'll go for dinner soon but might look at it after that [18:50] thanks [18:52] Amaranth: 200 ;-) [19:03] pedro_: yay === asac_ is now known as asac [19:27] pedro_: ping [19:35] Amaranth: just below the 200 now woohoo [19:35] LaserJock: hello [19:35] yep :) [19:35] * Amaranth hugs pedro_ [19:35] * pedro_ hugs Amaranth back [19:35] now it can't go back up, don't let anyone file new bugs [19:35] pedro_: I just wondered why you marked bug #183060 as Invalid [19:35] Launchpad bug 183060 in gnumeric "gnumeric crashed with SIGSEGV" [Medium,Invalid] https://launchpad.net/bugs/183060 [19:35] hahaha [19:35] LaserJock: looking [19:36] i've doing a lot of clean up lately [19:36] pedro_: I would think the LP Janitor would take care of it after 60 days [19:37] LaserJock: more than a month in a incomplete state without a response from the reporter [19:37] and that's enough to get it marked "Invalid"? [19:37] as i said i was doing some clean up lately, so that's why [19:37] yes it is [19:37] ok then [19:38] i'm trying to make compiz bugs look less depressing [19:38] so i'm not waiting 60 days for the janitor :P [19:38] alright, so we make it look like we dont' have bugs by just closing them? [19:39] 30 days without a response for a bug i can't reproduce means no one is ever going to help me figure out how to fix it [19:39] with compiz a driver point release or new compiz snapshot can cause and/or fix a million different problems [19:40] does the janitor already close bugs? or is still disabled? [19:40] it claims it will [19:40] i haven't seen it do so [19:40] haven't seen the janitor lately either [19:40] well, in this case though, somebody just asked for a backtrace [19:41] but shouldn't apport be providing those? [19:41] ah, sometimes apport does a bad job [19:41] or the coredump is bad [19:41] right, so we're just gonna ignore it? [19:42] if you can't get a good backtrace how can you fix it? [19:42] and as far as I can see nobody has *not* confirmed the bug [19:43] that's of course what needs to be worked on [19:43] but marking it "Invalid" basically is saying "We're not gonna look at it" [19:45] Right, it means "we're ignoring this unless you give us more info" [19:46] and why should we require the reporter to necessarily do that? [19:46]