=== cr3_ is now known as cr3 === LordTatty is now known as Tatty === Tatty is now known as LordTatty [00:40] hrm, it seems that my eth0 device has disappeared on my intrepid thinkpad [00:41] kirkland: e1000e? [00:41] wgrant: i'm not sure [00:41] wgrant: 82566MM Gigabit Network Connection [00:41] That sounds like e1000e to me. [00:42] poop [00:48] ThinkPads are known for their intel ethernet cards. [00:48] * kirkland goes boot a live cd to see if his hardward is toasted :-/ [00:49] kirkland: Don't! [00:49] kirkland: The driver has been disabled for a few days. [00:50] But booting an Intrepid alpha might kill it, as it's not blacklisted tehre. [00:54] try a gutsy livecd [00:54] Or Hardy. [00:54] or hardy,, right [00:55] wgrant: i booted a Damn Small Linux, 2.4 kernel [00:55] wgrant: device wouldn't load [00:55] gone [00:55] kirkland: I doubt 2.4 has drivers... [00:55] Does it appear in lspci? [00:55] wgrant: yeah [00:56] Try a Hardy live CD, as ogra-maemo suggested. [00:56] wgrant: lspci says http://pastebin.ubuntu.com/52659/ [00:57] wgrant: lshw says http://pastebin.ubuntu.com/52660/ [00:58] * kirkland goes try a hardy cd [00:58] kirkland: Looks like it's getting setup, which is a good sign. [00:58] kirkland: Your hardware is probably safe; in the other cases I've heard of it has disappeared off the PCI bus. [00:58] kirkland: AFAIR, if you fried the firmware, it won't even get BIOS setup (so, no IRQ, etc) [01:02] /join #linuxhelp [01:07] infinity: wgrant: eth worked okay with hardy cd... whew [01:10] kirkland: Good news indeed. [01:11] wgrant: thanks ;-) [01:11] np [01:12] wgrant: funny thing is, i saw that announcement from slangasek and thought, "whoa, sucks for them" [01:12] :-D [01:12] kirkland: Hahah. [01:13] i hadn't used my wired ethernet on my laptop in so long [01:13] All of my hardware is far too old, fortunately. [01:13] hehe, I'm currently connected to internet using a linksys ethernet USB card :) [01:13] stgraber: good idea... i've got some old pcmia "fast" ethernet card somewhere :-) [01:14] but that sucks as my test network is using gigabit and that card is only 100Mb/s, but still better than a toasted gigabit card [01:15] kirkland: yeah, I also have a couple of 3COM PCMCIA cards but they were 10Mb/s only :) I thought I had a 100Mb/s one somewhere but couldn't find it [01:15] stgraber: yeah, i hooked it up to gig ethernet to transfer some huge kvm images, as wireless just wasn't cutting it [01:27] kirkland: I've merged your landscape-client branch. [01:27] mathiaz: thanks [01:27] mathiaz: which do you prefer? bzr or debdiff? [01:27] kirkland: I'll upload the package once BetaFreeze is over. [01:27] mathiaz: cool [01:27] kirkland: in that case, the bzr branch. [01:28] kirkland: as everything is in bzr already. [01:29] kirkland: next time you could try to use the 'propose for merging' feature in lp. [01:29] mathiaz: sure [01:30] mathiaz: want me to do that for the update-motd one? [01:31] kirkland: well not really - since Vcs-Bzr points to a branch only you can commit to [01:33] mathiaz: right but i want my branch to be merged into the ~ubuntu-core-dev branch though, right? [01:33] kirkland: sure - except that ~ubuntu-core-dev doesn't exist. [01:33] mathiaz: ;-) [01:33] mathiaz: gotcha [01:33] kirkland: sure - except that the ~ubuntu-core-dev *branch* doesn't exist. === chuck_ is now known as zul === superm1|away is now known as superm1 [01:50] Who turned my background green? [02:00] Which package do I blame for the live CD ejecting and immediately reeating itself on shutdown? [02:00] wgrant: casper possibly. [02:01] As afaik casper does the ejecting. [02:01] (only on some machines, strangely enough...) [02:01] Thanks TheMuso [02:01] np [04:11] * Hobbsee ewwwww [04:12] Is this the place for talking about putting changes to an app into Ubuntu repos? [04:17] * wgrant agrees with Hobbsee. [04:17] portablejim: Depends - you likely want to file a bug on Launchpad. [04:19] I wnt help about getting the changes i have made to an app into the repos (since the app was broken due to lack of maintanance). [04:19] portablejim: Which app, what sort of changes. [04:19] s/./?/ [04:20] App: cdcontrol. Changes: Make the app so it does not fail 90% of the time. [04:21] It's in universe, so #ubuntu-motu is likely more appropriate. [04:21] But you probably want to file a bug (if there isn't one already), attach a debdiff, and get it sponsored. [04:22] https://wiki.ubuntu.com/SponsorshipProcess [04:26] Thanks. [05:53] I think I inherited evil [05:53] I have an RS/6000 running Windows NT for PowerPC [05:55] asac: ping (NM 0.7 dhcp debugging) [05:55] NCommander: You don't Ctrl-Alt-Del to login, but you draw a Sigil using the mouse? [05:57] StevenK, probably [05:57] Its running Windows NT server 4 [05:57] and [05:57] Oh god, IIS .... [05:57] ;.; [05:57] I could run a website off this relic [05:57] * StevenK arms a cobalt bomb [05:58] YES, SAVE ME WITH MIPS [05:59] I meant Cobalt, the element, not the product, you dolt :-P [06:14] Good morning [06:15] pitti: The 25-1 ports kernel has landed in NBS [06:15] pitti: There's a new linux-ports-meta waiting in unapproved, so it can't get killed yet [06:16] StevenK: acknowledged [06:17] the linux-ports-meta in unapproved was nacked by cjwatson [06:17] Ah [06:17] slangasek: It's crack? [06:17] the package names are altogether inconsistent [06:17] * StevenK pokes at it [06:18] Ew [06:18] Actually, it's NEW, not unapprovied [06:18] Er, unapproved [06:20] pitti: Can you poke me after you've caffinated? === BenC1 is now known as BenC [06:21] StevenK: poke (I'm online, just processing IRC scrollback in all channels) [06:22] pitti: So, the scariest thing still in NBS is libltdl3 :-( [06:22] StevenK: yeah, that'll require a big bunch of rebuilds still [06:23] pitti: I think they all FTBFS [06:23] right, s/rebuilds/build dep fixes/, I mean [06:23] * StevenK is checking [06:24] Riddell: ah, so upstream most likely develops qt on Fedora then :) [06:24] Riddell: I just wonder why it has wrong debuglinks then; does it still call dh_strip with some --dbg information? [06:25] Did anybody end up working out what should happen to get lib{32,64}ffi4 and the like caught by the NBS scripts? [06:25] ion_: oh, thanks for pointing that out; it's actually supposed to ignore the cache if any of the .fdis is obsolete; otherwise you would always have to restart hal if you install any package with an .fdi [06:25] wgrant: I didn't, at least [06:25] ion_: can you please file an RC bug for this and assign it to me? I'll look into that ASAP [06:26] pitti: wgrant just reminded me. Should we NBS those two binaries out? [06:26] sure, they aren't built any more [06:26] There are probably dozens of others. === tkamppeter_ is now known as tkamppeter [06:40] pitti: liblame0 only has rdepends of mencoder and mplayer on powerpc, do you think we can kill it? [06:45] * TheMuso digs into the PowerPC mplayer FTBFS. [06:45] StevenK: it's pretty popular, but sure, we can rebuild it a bit later [06:45] oh, it's FTBFS? darn [06:45] oh. [06:46] and FTBF sbecause of a compiler segfault it seems. [06:46] Whee [06:46] An ICE? [06:46] let me check [06:46] ad_imaadpcm.c:376: internal compiler error: Segmentation fault [06:47] So, yeah, an ICE [06:47] It was given back about 9 hours ago, too [06:47] * TheMuso goes to try and build locally. [06:48] good morning [06:50] morning dholbach [06:51] hi Hobbsee [06:57] wgrant: Bug 276629, if you want to subscribe [06:57] Launchpad bug 276629 in soyuz "archive-cruft-checker doesn't pick up all cruft" [Undecided,New] https://launchpad.net/bugs/276629 [07:00] StevenK: Thanks. [07:03] Oh, this isn't cool. [07:03] guile-gnome-platform runs xvfb during the build process which spins with 100% CPU usage [07:03] Nice. [07:04] 3 failed: cacao guile-gnome-platform ksimus [07:04] 3 unaccounted for: gnash mit-scheme myodbc [07:04] Hm? [07:04] (unaccounted for is either "Build Depends failed to install" or similar erorr, or it's still building) [07:04] Ah. Some Soyuz CLI tool output? [07:05] Nope, it's a script I wrote for test builds [07:05] One of the 6 or so scripts I wrote to help with NBS [07:05] Aha. [07:05] Didn't think I'd seen it before. [07:06] Ah ha. mit-scheme is only i386. [07:07] steven@liquified:~/ubuntu/libltdl3-cruft% grep -c 'warning:' cacao/current [07:07] 481 [07:07] Bah === pgraner_ is now known as pgraner [07:10] Yep, mplayer on PowerPC fails for me locally in the same way as on the buildds. [07:10] Gar, not again. [07:10] TheMuso: With what kind of error? [07:11] wgrant: An ICE [07:11] Ah. SEP === philwyett_ is now known as philwyett [07:17] ARGH [07:17] * NCommander shivers [07:17] Launchpad's new UI burns [07:18] wgrant, its a compiler segfault. I did a little work debugging it, but its really ugly [07:19] NCommander: Bug #276618? [07:19] Launchpad bug 276618 in launchpad-foundations "New black (sub)subtabs look bad" [Undecided,New] https://launchpad.net/bugs/276618 [07:19] mplayer is in multiverse, perhaps we could build it with gcc-4.2? [07:19] TheMuso: Do you want to try that? It's a hack, but it could get a working mplayer on PPC [07:20] pitti: What do you think? ^ [07:20] StevenK: Wortha try, as soon as I work out what needs changing to do that. :) [07:21] TheMuso: export CC=gcc-4.2 and Build-Depend on gcc-4.2 [07:21] Er, export CC=gcc-4.2 in debian/rules [07:21] wgrant: Argh [07:21] wgrant: It really does reach out and slap you [07:22] StevenK: Right, will have a play. [07:22] TheMuso, I can roll those changes [07:22] StevenK: What does? The black hole? [07:22] wgrant: The black subtabs [07:22] StevenK: Right. [07:22] NCommander: I'll try here as well. [07:23] Open that page, flip to another tab, and then go back and it hits you [07:23] NCommander: Comparing LP to SF... that's low. [07:23] 6 failed: cacao guile-gnome-platform ksimus mit-scheme myodbc pacemaker [07:23] 1 unaccounted for: gnash [07:23] Sob [07:24] StevenK: Indeed. [07:24] wgrant, I could do worse [07:24] Hm. Where did libkonq4-dev go? [07:24] StevenK: Away with KDE3, maybe? It is KDE, so it can do that. [07:24] libkonq4 is KDE3? [07:24] Blink [07:24] It might be. [07:24] Probably not, though. [07:25] libkonq5-dev exists now... [07:25] Yeah, just trying that [07:27] wgrant, can I just switch it to use gcc-4.2 across the board, or would you perfer if I just made that change on PPC :-) [07:27] Ideally just on PPC, but I can't see it killing much on other archs... [07:28] Is there a handy trick to determine an architecture in the rules? [07:28] Yes [07:28] Let me dig it up [07:28] mplayer already uses it lots. [07:28] Yes, yes there is [07:29] ifeq (lpia,$(shell dpkg-architecture -qDEB_BUILD_ARCH)) [07:29] EXTRA_CONFIGURE_ARGS += --enable-hildon [07:29] endif [07:29] Yes, I am just popping the export under the configure arch setting for altivec. [07:29] * NCommander did the same :-) [07:29] TheMuso, its a race to see who finishes first [07:29] heh. [07:29] * NCommander test builds ;_0 [07:29] its not really. [07:30] NCommander: Don't forget changing debian/control. [07:30] TheMuso, already done [07:30] StevenK, BTW, do you know anyone who cares about Ubuntu SPARC? [07:30] * NCommander notes that all ports are uninstallable from scratch in intrepid [07:32] NCommander: Due to what error? [07:32] StevenK, due to lack of kernel [07:33] StevenK: linux-ports-meta [07:33] no linux-ports-meta package [07:33] The upshot is that no ports can release with intrepid [07:33] It's in NEW, it got nacked [07:33] * NCommander rolls eyes [07:33] (PowerPC has made some progress, but we won't be releasable until jaunty likely) [07:34] The problem is there is no one that I can find who can do the work on getting the kernel into shape for HPPA, IA-64, or SPARC [07:36] (LPIA sidestepped the issue with linux-lpia and linux-lpia-meta) [07:37] * StevenK is well versed on linux-lpia{,-meta} [07:37] NCommander: The difference is that there is actually a team supporting that port. [07:38] Well, maybe we should do linux-powerpc{,-meta} [07:38] That way each port can work without stepping on each others toes so to speak [07:38] No. Three copies of linux in the archive is quite enough. [07:38] Although -rt might still be separate. I forget. [07:38] wgrant: Well there is actually more, but you have a point re the metas at least. [07:38] wgrant: -lpia and -generic are now both 27 [07:39] StevenK, you won't happen to know a closet SPARC developer, would you? [07:39] lol [07:39] I ask more that I'm worried that REVU will have to switch to Debian [07:40] The irony would be overwhelming :-/ [07:43] StevenK, BTW, I got a new toy, an RS/6000 ;-) [07:43] NCommander: Running NT4 server [07:43] You know, I made a joke about it [07:43] oh right [07:43] I fail [07:44] Sorry [07:44] Windows NT for PowerPC hurts my brain [07:44] then don't use it [07:44] I'm not [07:44] Its just the mere proximity ;-) [07:45] Yay, progress: [07:45] 7 failed: cacao gnash guile-gnome-platform ksimus mit-scheme myodbc pacemaker [07:45] NCommander: Didn't you look at cacao? [07:45] why is myodbc failing? [07:45] ../driver/.libs/libmyodbc3_r.so: undefined reference to `mysql_odbc_escape_string' [07:45] StevenK, yeah, thats a FUN FTBFS [07:46] StevenK, the general issue is GCC changes its JNI mappings, the cacao SVN has a fix [07:46] But [07:46] (there is always a but) [07:46] They completely redid the JNI code [07:46] So backporting would be an absolute nightmare [07:46] StevenK: ummm, that appears to be a function provided by libmysqlclient, so where did it go? [07:46] slangasek: Would you like to see the complete build log? [07:47] no, just the linker command line immediately befor eit :) [07:47] gcc -O3 -DDBUG_OFF -I/usr/include/mysql -DBIG_JOINS=1 -fPIC -g -Wall -O2 -Wl,-Bsymbolic-functions -o .libs/my_basics my_basics.o ../driver/.libs/libmyodbc3_r.so -L/usr/lib -L/usr/lib/mysql /usr/lib/libmysqlclient_r.so -lcrypt -lnsl -lm -lz /usr/lib/libodbcinst.so /usr/lib/libltdl.so -ldl -lpthread -Wl,--rpath -Wl,/usr/lib [07:47] o_o; [07:48] ah. Stupid test suite [07:49] slangasek: Hm? [07:49] my_basics is part of the test suite; I'm blaming the test suite to avoid acknowledging that there's probably a bug elsewhere ;) [07:50] Hehe [07:50] And gnash now installs it's Build-Depends, but promptly fails in ./configure [07:51] Sigh [07:51] sladen, probably the MySQL API changed, and someone didn't bother to update the test suite [07:51] * StevenK points NCommander at slangasek [07:52] s/test suite/package/ [07:52] *sighs* [07:52] Two autocomplete failures [07:52] Damn it [07:53] * NCommander hacks it so that if sladen is in the room, he is ignored and slangasek is selected instead [07:53] poor sladen [07:54] StevenK, anyway, if you want to "fix" cacao, its not a picture [07:54] NCommander: The new version of cacao in Debian builds? [07:54] They packaged the SVN snapshot [07:55] Typical [07:55] StevenK: the fact that myodbc is FTBFS in this way means that the libmysqlclient ABI has silently changed since June, dropping library entry points [07:55] And only uploaded it to experimental [07:55] slangasek: \o/ [07:55] StevenK, yeup [07:55] It's a rather nasty problem [07:55] Since we're dealing with GCC JNI ABI [07:56] It seems you can build it if you use Sun's javac, but I didn't get too far in figuring out how to do that [07:56] slangasek: Could you be convinced to hand wave a cacao in? :-D [07:56] do what then? [07:57] slangasek: Be convinced to sync "cacao | 0.99~rc5-1 | experimental | source, amd64, i386, mips, mipsel, powerpc" [07:57] ewwww [07:58] StevenK: er... why wouldn't you do that yourself? === sladen is now known as vorlon [07:58] Haha [07:58] That's cheating [07:59] slangasek: I can, but I need a FFe and all that jazz for you to approve, don't I? === vorlon is now known as sladen [08:00] And Debian have a newer version of gnash than we do [08:00] StevenK: motu-release; it's in universe [08:01] Aren't you also on that team? :-P [08:01] not by choice ;) [08:01] * StevenK chuckles [08:01] tjaalton: JFYI, running with the older -intel 3a2.4.1-1ubuntu4 for three days now, and didn't have the "suddently turns black" issue; that's with current kernel, current X, and (few) pipe A/B underruns in Xorg.0.log [08:01] slangasek, it could be worse. You could be a member of ubuntu-hppa ;-) [08:02] pitti, how soon do you need libpam-blue by? [08:03] oh wait [08:03] That was persia who asked me to look at that [08:04] right; I was pretty confused :) [08:04] (although I'm watching the progress in the bluez 4 bug) [08:05] Argh [08:05] Bah [08:05] I hate packages that don't have a patch system [08:05] pitti, if there is no patch system, which one do you recommend I add, or should I just inline? [08:06] (I need to inline a bunch of M4 marcos to fix libtool :-/) [08:06] ls [08:06] er [08:11] NCommander: I didn't touch acpi-support for a while, but I could check whether it's ok if I commit some fixes to it [08:12] NCommander: The Debian maintainer is nice but he has been doing all the work, so I'd prefer checking with him before using my commit rights [08:12] or permissions rather [08:12] lool, I resolved the issue without needing to touch it, and since ACPI sleep is seemingly going away for jaunty [08:12] ACPI sleep is going away? [08:12] Thats what someone said [08:13] As in, the kernel will do the sleeping without speaking to ACPI, or no sleep? [08:13] Just pm-utils will be supported [08:13] * wgrant can't see either as being possible. [08:13] NCommander: Oh, you mean acpi-support is going away [08:13] yeah [08:14] Sorry if I wasn't clear [08:15] NCommander: my personal recommendation is "simple-patchsys.mk for cdbs" and "dpatch for normal debhelper", they are the least intrusive ones [08:16] NCommander: however, I wouldn't add one at all if diff.gz already has patches [08:16] * NCommander nods [08:16] Yeah, it does [08:16] * NCommander HATES autotools [08:16] even though they're good [08:16] When they break, they're miserably [08:16] ugh simple-patchsys [08:16] Ugh CDBS [08:16] Ugh quilt [08:17] ugh a gazillion different possibilities for source packages [08:17] I quite like dpatch, though. It doesn't require that I smash my head through my desk to patch something [08:17] pitti: you're right; let's eliminate the one that's most inadequate, simple-patchsys :) [08:17] StevenK, WOOOOOO [08:17] Haha [08:17] Ugh dpkg [08:17] Ugh rpm [08:17] So says Steve "Vendetta" Langasek [08:17] slangasek: /usr/share/cdbs/1/rules/patchsys-quilt.mk WFM, too [08:18] however, I found that many people have more troubles with quilt than with cdbs-edit-patch [08:18] quilt is the patch system that forces me to smash my head through my desk to patch something [08:18] quilt is insanely stupidly complicated. [08:19] slangasek: simple-patchsys' popularity is actually quite astonishing, given that it was just meant as an "example"... [08:19] With dpatch and simple-patchsys I just tell it to edit the patch. [08:19] wgrant, its the emacs of patch systems [08:19] wgrant: wait until you see bzr-loom then :) [08:19] bzr-loom? [08:19] quilt refresh \n Launching tactical nukes [08:19] bzr-loom isn't bad. [08:19] * NCommander is reminded of the Lucasarts Game [08:19] wgrant: the stupidity of quilt is mainly "quilt add" IMHO; the general design is quite good [08:20] pitti: But that is infinite stupidity. [08:20] * NCommander hates it when a package maintainer changes something for no clear reason without noting it in the changelog [08:20] wgrant: has bitten me more than once, too [08:21] I don't like the inline change, there is no way to "exit 1" if you screwed something [08:21] seb128: Me either [08:21] and it often let noise in the source after doing the refresh and pop too [08:21] quilt is the only one I finally understood how it is supposed to work :) [08:22] seb128: however, you can actually change/build/test in that mode, which isn't possible wit simple/dpatch [08:22] I usually end up binning the source tree and starting again [08:22] It always leaves something in .pc for me... or am I missing something? [08:22] siretart: you don't understand how cdbs simplepatchsys is working? [08:22] OMG WHY [08:22] WHY [08:22] WHY [08:22] Woo, caco 0.99~rc5 builds [08:22] * pitti reduces NCommander verbosity by 1 [08:22] WHY IS THIS MAINTAINER REGENERATING AUTOTOOLS IF THERE ARE NO CHANGES TO CONFIGURE.IN [08:22] Ack verbise -1 [08:22] seb128: it seems to insist to create a copy of my tree every time I want to create or modify a patch [08:22] **** - new verbose explination [08:22] which is insane for larger trees [08:23] BTW, siretart do you care about Ubuntu SPARC? [08:23] NCommander: spooky is one. so to some degree, yes [08:23] siretart, good, you just became the sparc kernel guy for linux-ports :-) [08:23] siretart: right it does a copy, I don't work on packages large enough for that to be annoying though, gtk or evolution take a couple of seconds to do that [08:24] seb128: I mainly work on my laptop with crypted everything. copying is painful for me [08:24] NCommander: I see. [08:24] NCommander: as first step, I resign :) [08:24] Well, intrepid SPARC is not installable [08:24] If you want to fix that for jaunty, you shouldn't resign [08:25] Less we make spooky run Debian ... [08:25] But that would be wrong :-) [08:28] Ok mplayer successfully built using gcc 4.2. [08:28] \o/ [08:28] pitti: Do you think that's a good call -- uploading a mplayer that uses gcc-4.2 on ppc? [08:29] StevenK: if it ICEs with 4.3, that's an acceptable workaround, yes [08:29] especially since it's not in main [08:29] TheMuso: Upload it! [08:29] argh [08:29] * NCommander wanted credit for that upload [08:29] * NCommander runs [08:30] heh [08:30] NCommander: U looz [08:30] Ok mplayer is in bzr so I'll do things that way. [08:30] Well, I don't have any uploads in multiverse or restricted [08:30] NCommander: http://www.pvponline.com/2008/06/30/interlude-the-adventures-of-lolbat [08:34] StevenK, lolbot [08:34] er [08:34] bat [08:34] # well, rm -f config.log :) [08:34] libpam-blue source: configure-generated-file-in-source config.log [08:34] o_o; [08:34] Why do people think its a good idea to disable lintian warnigns [08:34] *warnings [08:38] * TheMuso ponders how mplayer got uploaded with that maintainer field. I try and dpkg-buildpackage it, and I get the maintainer field warning. [08:38] * NCommander thinks we need to resurrect the ubuntu-powerpc group, and get someone who is active to be an admin [08:38] TheMuso: the previous upload might not have had an ubuntuish DEBEMAIL [08:39] pitti: You're totally correct. [08:39] I'm using themuso@ubuntu.com. [08:39] so then he just got a warning and ignored it [08:39] TheMuso: "update-maintainer" FTW :) [08:39] right [08:39] libtool: compile: x86_64-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I../../libbase -I.. -Wall -g -Werror-implicit-function-declaration -O2 -W -Wall -Wcast-align -Wcast-qual -Wpointer-arith -Wreturn-type -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -MT libltdlc_la-ltdl.lo -MD -MP -MF .deps/libltdlc_la-ltdl.Tpo -c ../../libltdl/ltdl.c -fPIC -DPIC -o .libs/libltdlc_la-ltdl.o [08:39] ../../libltdl/ltdl.c:32:25: error: lt__private.h: No such file or directory [08:39] * StevenK kicks libtool [08:39] pitti: I know, but the maintainer is an Ubuntu team, i.e the MOTU media team, which has an email address pointing to tauware.de. [08:40] TheMuso: oh, ok; workaround: "DEBEMAIL= debuild -S" [08:40] pitti: Right. [08:40] TheMuso: I don't think how we can make the check recognize this case without becoming totally useless [08:40] s/think/see/ [08:41] StevenK, libpam-blue fixed [08:41] \o/ [08:41] THe fix was removing the maintainers insane rules file [08:41] He rebuilt autotools incorrectly, disabled autoconf's Makefile generator and ... [08:41] Ew [08:42] The debdiff is massive [08:42] Since all the crap he added was removed [08:42] TheMuso: The solution is probably to move than to a *ubuntu* domain, like motu-reviewers was. [08:42] wgrant: Long term, I agree. [08:43] * TheMuso double-checks it on x86 to make sure he didn't inadvertantly break anything else. [08:49] superm1, ping [08:49] NCommander: likely asleep. [08:49] :-( [08:51] StevenK, care to upload libpam-blue to the bluetooth PPA [08:51] https://bugs.edge.launchpad.net/ubuntu/+source/libpam-blue/+bug/276343 [08:51] Launchpad bug 276343 in ussp-push "Rebuild for libbluetooth2 -> libbluetooth3 transition" [Undecided,Fix committed] [08:51] NCommander: I can add you to the bluetooth team [08:51] \o/! [08:51] YALPT [08:51] NCommander: Apply, I'll approve you [08:52] It's offical [08:52] at 800x500, I now have enough teams to cause a horizonal scrollbar on my profile ;-) [08:52] NCommander: heh [08:53] TheMuso, I'm in a race with cody-somerville to see who can cause a scrollbar to show up at 1024x768 first [08:53] StevenK, applied [08:53] NCommander: I just approved you [08:53] * NCommander adds ppa-bluetooth to his dput config [08:53] NCommander: *sigh* you're not really in a race are you? [08:53] \o/ [08:53] TheMuso, 29 to 30 [08:53] er [08:53] wait [08:53] 30 to 30 [08:53] WOOO [08:53] tied [08:54] !enter | NCommander [08:54] NCommander: Please try to keep your questions/responses on one line - don't use the "Enter" key as punctuation! [08:54] OK, fine. Enter [08:54] :-) [08:54] Haha [08:56] libpam-blue uploaded :-) [08:56] StevenK, so the bluez transition is complete [08:57] Now we just need to release intrepid, and pound the Debian BT guys to adopt our packages === dholbach_ is now known as dholbach [08:57] StevenK, ew, there were a few ugly bluetooth build failures [08:58] There are? [08:58] Which? [08:58] StevenK, libbtctl4-dev(inst 0.10.0-1ubuntu1~ppa1 ! >= wanted 0.10.0-1ubuntu1) [08:58] That's an anonying one :-) [08:59] same with sylpheed [08:59] That's due to the funness of ~ [09:00] 0.10.0-1ubuntu1~ppa1 isn't greater than 0.10.0-1ubuntu1 [09:00] That's why you use source:Version [09:00] StevenK, the transition for bluetooth was done damn fast O_O; [09:00] less than 19 hours [09:03] NCommander: slypheed wants pilot-link [09:03] And wants greater than ubuntu3 [09:03] Argh [09:03] Death to PPA versioning [09:06] StevenK, you could just upload sane versions ;-) [09:06] * NCommander runs [09:07] pitti: so, the no_render_suspend patch broke it for you.. [09:08] tjaalton: likely, but the changelog made it look hardware specific? [09:08] tjaalton: shall I file a bug about this now? [09:09] pitti: yep, but the patch also affects 965GM systems [09:10] pitti: which one did you have? [09:10] tjaalton: (00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03) [09:10] pitti: huh, should not make a difference then.. [09:13] * StevenK kicks gnash [09:14] tjaalton: it might be a more complicated chain of change, such as building ubuntu6 against a newer X server, or newer toolchain, or so, but that becomes really obscure then [09:14] pitti: Right, all seven of the source packages required by libltdl3 are broken in odd ways. [09:15] pitti: cacao can be fixed by sync'ing from Debian experimental, but that requires a jump from 0.98-2 to 0.99~rc5-1 [09:15] StevenK: that should get doko's blessing then [09:15] gnash is still busted, even the newer upstream version in Debian doesn't build [09:16] just because of libtool, or for other reasons? [09:17] StevenK: myodbc isn't broken; that's a mysql-dfsg-5.0 bug [09:19] (well, perhaps myodbc is also broken, if we choose the horrible option of bumping soname :P) [09:21] StevenK, what's wrong with gnash? [09:22] hi ! [09:36] StevenK: 0.8.4 will be it [09:36] (gnash) [09:36] upstream has fixed gstreamer support a few days ago [09:51] hum I accidentally uploaded a package to ubuntu that should have gone to my ppa... [09:51] this is NEW package so might not reach the archives, can somehow drop it when falling in NEW plz ? [09:51] the package is libpackagekit-qt [09:52] Tonio_: to be safe ping the archive admins directly [09:52] asac: true that, and most of them are here :) [09:52] asac: but yeah, an email or Riddell directly would be better :) [09:53] * Hobbsee looks [09:53] Tonio_: hm, it's not there yet. [09:54] * Hobbsee accepts a bit more universe stuff in the mean time [09:54] * Hobbsee %^&*&^%$#%^&*(*&^%$%^&*(* [09:54] * pitti reports a memory corruption bug against Hobbsee [09:55] you mean that's not a clever regexp? [09:55] rather a sendmail config file [09:55] pitti: i hope the title is "launchpad is being a pain in the neck" again. [09:57] tseliot: FYI, just committed a jockey change to test main window without any handlers [09:57] right. I can accept at least 4 at a time. OK. === The_Company is now known as Company [10:09] pitti: great, I'll have a look at it [10:12] Tonio_: rejected libpackagekit-qt [10:12] StevenK: 0.99~rc5 seems to be way too old [10:15] Riddell, ScottK: is there web content for kubuntu that I should link to for the beta? [10:16] Riddell: thanks :) [10:21] slangasek: will be at https://wiki.kubuntu.org/IntrepidIbex/Beta/Kubuntu [10:46] doko: Debian experimental has 0.99~rc5, should I be looking at a newer version? [10:46] Riddell: qxembed.h ... where would that be? [10:46] asac: And you'll deal with gnash when 0.8.4 is released, right? [10:46] StevenK: yes [10:46] asac: Cool [10:47] * ogra wonders what it is that makes NM keep asking for the network passphrase on his mobile image [10:47] StevenK: i will bring up a snapshot right after beta and release should happen quite close to our release [10:47] ogra: because its a bug [10:47] ogra: take the NM from PPA [10:47] slangasek: So, bug you about mysql-dfsg-5.0 after you've had 2 days sleep after the beta is released? [10:47] ogra: ~network-manager [10:47] StevenK: bug me? Bug the server team instead? :) [10:47] asac, phew, i was suspecting its a missing package or something in the gnome-keyring stack [10:47] if its a bug all is fine [10:47] ogra: no. its a regression and we had bad luck with that snapshot ;) [10:48] Hm. Who on the server team deals with mysql? [10:48] StevenK: changelog says zul uploaded it [10:48] asac, then i'm fine, i cant build from PPAs now that its built in the datacenter [10:48] but its a bit painful if you are bound to an onscreen kbd to have to put it in every boot [10:48] ogra: sounds complicated :-D [10:49] ogra: not sure. you could put it into keyring manually ;) [10:49] StevenK: http://www.complang.tuwien.ac.at/cacaojvm/download/ [10:49] asac, well, i assume its targeted to be fixed for final [10:49] ack [10:49] doko: Okay, I can look at updating to 0.99.3 ? [10:49] so i dont care for now, its not mobile specific [10:50] asac, btw feel free to test mobile on your Q1 :) http://cdimage.ubuntu.com/ubuntu-mobile/intrepid/current/ [10:50] ogra: is that an install or live image? [10:51] StevenK: there should be another release soonish. please could you add a cacao-source binary package just including the source rarball? then I don't need to copy that into cacao-oj6 [10:51] asac, installable live image :) [10:51] btw, why are you interested in cacao itself? [10:51] asac, its my side of ograsac on a livecd [10:51] ogra: yay ... byebye imagecreator? [10:51] err s/cd/usb image/ [10:51] yup [10:52] all mobile images (ubuntu-mid and ubuntu-mobile) are built in the datacenter with livecd-rootfs [10:52] doko: I'm interested in cacao because it's in the NBS list for libltdl3 [10:53] asac: We haven't touched MIC all Intrepid [10:53] \o/ [10:53] :D [10:55] doko: Hm. cacao-oj6 is just source and has a 6b... version number. I'm confused. [10:55] NBS? [10:55] Not Built from Source [10:55] libltdl7 is now provided by libtool, not libltdl3 [10:56] the versioning is done using the opnjdk source drop [10:56] doko: If cacao is on your list to deal with, then please go ahead :_) [10:57] StevenK: no, I don't care about cacao myself. why is cacao needed for libltdl7? [10:58] doko: Because the cacao binary package Depends on libltdl3. We can't remove the libltdl3 binary package from the archive until nothing depends on it, or we break them [10:59] ahh, ok. well, go ahead. would save me some typing adding this package myself. else I'll include the copy myself [11:00] doko: Well, hence this conversation. 0.98-2 in Intrepid FTBFS. 0.99~rc5 in Debian works, but I can update that to 0.99.3 if you wish [11:01] StevenK: yes, and please add the cacao-source binary package [11:03] doko: Okay, cool. I think a debdiff is worthless, but I'll point you at it when I'm done [11:09] Oh, ew. Cacao uses cdbs [11:13] asac: kdelibs4c2a (from memory) [11:13] or kdelibs4-dev [11:14] Riddell: hmm [11:15] * anilg is portinf kdelib4-dev and he hates waiting 15 minutes to see if the last change fixed the issue [11:15] Riddell: what should gnash use for kde? qt4 and kde4 ? [11:23] asac: if it has a kde 4 port then it should use that. as far as I know if only has a kde 3 port in which case it should use kdelib4-dev [11:23] anilg: porting to what? [11:23] Nexenta hardy [11:24] some thigns dont work the same way in solaris.. and things debian/rules expects arent always there [11:31] Riddell: would the kde3 plugin still be useful in kde4 knoqueror? [11:31] asac: no [11:32] Riddell: kde4 will be default right? [11:32] asac: yes, there is no kde 3 konqueror, so the kde 3 kpart can be dropped [11:32] Riddell: ok. so unless gnash devs fix kde4/qt4 support its just the standalone player we can ship [11:33] assuming that kde4 users can install kde3 apps ;) [11:33] yes that's fine [11:43] pitti: Will do === Ng_ is now known as Ng === jamesh_ is now known as jamesh [12:13] wow, shift+K now actually works when editing .py files in vim; hadn't noticed that yet [12:13] what does it do ? [12:14] Shows the help for the keyword under the cursor. [12:14] sweet [12:15] normally it shows the manpage for the word under the cursor [12:15] but now it's clever enough to actually consult Python docstrings [12:15] just tried it with "shutil" and "tempfile", and it shows the python modules [12:17] pitti, btw, i use fdi files for touchscreens, but that apparently only works if i install them to /etc/hal/fdi/policy ... if i put them into /usr/share/hal/fdi/policy/10osvendor the evdev driver overrides evtouch, any hint how to solve that ? [12:17] ogra: you might need a higher number prefix then, for correct ordering? [12:18] oh, that might help indeed, i use 10 for all of them ... which the evdev one surely does as well [12:18] and sadly touchscreens often enough match input.mouse [12:31] * ogra hugs pitti ... that seems to work [12:32] * pitti hugs ogra back, good [12:36] seb128, can you let ubuntu-mobile-default-settings through ? (universe) [12:36] or do i need a release team member to do that during the freeze ? [12:37] no, any archive admin can shove through universe [12:37] ogra: technically I can not sure if I'm supposed to though, better to ask pitti though [12:37] right, thats what i thought [12:37] ok, good to know [12:37] * pitti pokes [12:38] thanks :) [12:38] flushed all pending universe stuff [12:38] gconf is silly ... [12:38] as soon as i put quotes around a default setting it makes it a string, tsk [12:39] I would say that's a good thing [12:42] * ogra applies the same fix to ubuntu-mid-default-settings [12:43] seb128, well, depends ... many apps just cope with that, but some dont, i used to quote all my settings unlit i found that some booleans dont apply [12:43] where do you set those values? [12:43] and how? [12:44] in a file in /usr/chare/gconf/defaults [12:44] seb128, /apps/gksu/disable-grab "true" doesnt work, /apps/gksu/disable-grab true does :) [12:44] why would you quote numbers or boolean there [12:45] you are asking for trouble [12:45] right, "" is usually a string [12:45] that's the same in let's say python [12:45] well, /apps/gnome-power-manager/lock/hibernate "false" works [12:45] 1 and "1" are different things [12:45] it really depends how the app pulls it out of the db in the end it seems [12:45] that's probably because the code is smart enough to ignore incorrect types [12:45] g-p-m doesnt seem to care [12:46] or that it works by luck [12:46] no, the gconf api let you specify the object type you expect [12:46] well, gksu was the first app where i ran into trouble in 3 years ... [12:46] some software will ignore incorrect values [12:46] since it apparently does strict checks [12:46] almost nothing uses .gconf-default [12:46] which indeed is a good thing [12:46] that's a debian specific things to make easy to set system defaults [12:46] right [12:47] which i use in edubuntu, ubuntu-mobile and the mobile team uses in ubuntu-mid [12:47] ok so you have been lucky until now ;-) [12:47] and i guess xubuntu does that as well [12:47] yeah [12:47] :) [12:49] Keybuk: Hi thx for your comment on LP #274860. Any idea why this problem doesn't show up on other systems? [12:49] Launchpad bug 274860 in ubuntu-ps3-port "usplash doesn't run on boot on PS3 in intrepid" [High,In progress] https://launchpad.net/bugs/274860 [12:49] phew, gdm is a bad beast ... seems teher is no easy way to get cellwriter only shown in the login screen [12:58] munckfish: different framebuffer drivers maybe? [12:58] cjwatson: hi [12:58] cjwatson: do you think would could just revert that debian change? [12:58] so that it works again? [12:58] munckfish: seems like we should just revert that initramfs-tools change, yes [12:58] I don't see any benefit to it [12:58] ok [12:58] it's an optimisation, and correctness trumps that [12:59] ok [12:59] Our initramfs-tools source isn't in bzr I think [13:03] cjwatson: I wonder if we should revert this commit too ... [13:03] http://git.debian.org/?p=kernel/initramfs-tools.git;a=commitdiff;h=d377e38823d06d8e79341fba44ce8ea9c42daff6;hp=0aec8b0c22b7622841c4ab7a3b492b4d2657456f [13:04] as it removes the /dev/tty* setup which was done there too [13:04] cjwatson: comment indicates udev is expected to do that too [13:10] mm, possibly [13:10] TheMuso: ping === vorian is now known as stevie [13:15] * StevenK tries to figure out how to bend find(1) to his will [13:17] cjwatson: ok i'll revert the fb one first if it all works without the tty one then ... so be it [13:18] willl attach patches to the bug once I'm done [13:18] cjwatson: any thoughts on the spufs one? Seems initramfs needs to bundle the spufs.ko module and explicitly load it. [13:18] jeremy has told me that spufs should be autoloaded by the kernel [13:19] but it's not in our case. I'm going to do a build with some printk in to see if I can see what's wrong [13:19] I hadn't seen that bug [13:20] oh, I had, just forgotten about it [13:20] wouldn't be the first time we had to explicitly load a module *shrug* [13:20] LP #274854 [13:20] Launchpad bug 274854 in ubuntu-ps3-port "spufs filesystem not mounted on boot" [Medium,In progress] https://launchpad.net/bugs/274854 [13:21] There were too other bugs sort of related to the installer failing [13:21] because of spufs [13:21] I created this new one for the "on boot" version of the problem [13:23] cjwatson: my question would be where best to do the loading ... [13:23] 1) in an initramfs script, if so which package should home it? mountkernfs.sh is in the initscripts package [13:24] why does it need to be loaded in the initramfs, though? [13:24] we could add a script to that [13:24] cjwatson: well this is the other thing [13:24] I mean, it's being used outside the initramfs [13:24] doesn't make sense for the initramfs to load it [13:24] it's getting mounted in mountkernfs along with /sys and /proc [13:24] it doesn't need to be [13:24] in which case the question is where and how should it be loaded? [13:25] how about right before it gets used? [13:25] i.e. mountkernfs [13:25] where does it get used? [13:25] when mountkernfs tries to mount it? [13:25] ok I see [13:25] davmor2: pong === pedro__ is now known as pedro_ [13:25] TheMuso: has the login music in ubutnu been changed? [13:26] davmor2: No, its just that there are two themes, the freedesktop theme and the ubuntu one, and there seems to be a weird bug where sometimes the ubuntu one is not used. [13:27] munckfish: if it isn't clear, mountkernfs is run after the end of the initramfs [13:28] cjwatson: sorry I was just checking exactly that by extracting my ramfs [13:28] for some reason I thought it was bundled [13:28] cjwatson: anyway, so it would be easiest to modify mountkernfs.sh to do a modprobe if required [13:28] providing that works [13:29] yep should be easy enough I'll knock that up this week [13:29] cjwatson: thx for your advice [13:37] TheMuso: yes I thing it might be to do with metacity/compiz. I got 2 test machines and on my all intel one I get the normal ubuntu tune but on the nvidia one I get the other version (or it could be the other way around) [13:38] davmor2: I think its more random than that. Most of the time when I log in I get the Ubuntu one, but occasionally I get the freedesktop theme. [13:38] But thats only eover for the login sound. It has something to do with the gconf key for the theme not being loaded by the time it has to play. === Ziroday is now known as ziroday [13:40] /c === LucidFox_ is now known as LucidFox [13:50] asac: see bug 268005 for an interesting test case for n-m/ifupdown interaction [13:50] Launchpad bug 268005 in network-manager "live cd from nfsroot breaks the nfs mount during bootup" [Undecided,New] https://launchpad.net/bugs/268005 [13:50] asac: if you have an alternative proposal for how casper could stop network-manager from getting in the way, I'd be interested [13:55] cjwatson: only way i see is to make the device used for nfs in initramfs unmanaged. so the bug 256054 will fix it [13:55] Launchpad bug 256054 in network-manager "[intrepid] new 0.7 branch ignores /etc/network/interfaces" [Unknown,Confirmed] https://launchpad.net/bugs/256054 [13:55] like you said [13:56] asac: right, so just what casper's doing at the moment then [13:56] it does "inet manual" for it [13:57] cjwatson: is that a special image? [13:57] asac: hmm? [13:57] cjwatson: thought that we dont create anything in /etc/network/interfaces [13:57] asac: not by default, but if you're using the wacky boot-live-CD-over-NFS mode then we do [13:58] or are we doing that on livecd, but remove those lines when installing? === davmor2 is now known as davmor2_lunch [13:58] and since it's done in casper it isn't copied when installing [13:58] cjwatson: right. i think booting over nfs will not work with NM at least until NM 0.8 - if ever [13:59] ok, that's quite an important regression so I'd like to do something about it. What do you think is the most practical option for the live CD? [13:59] cjwatson: oh. so the install will not boot over nfs? [13:59] just disable n-m altogether if you're booting over NFS/ [13:59] ? [14:00] cjwatson: the entries in /etc/network/interfaces should be enough. if you need something before bug 256054 is fixed you have to disable NM [14:00] Launchpad bug 256054 in network-manager "[intrepid] new 0.7 branch ignores /etc/network/interfaces" [Unknown,Confirmed] https://launchpad.net/bugs/256054 [14:02] but that fix has to land anyway [14:02] the touch NetworkManager in "cow" workaround looks like a proper hack ;) [14:03] asac: hmm, ok [14:03] yeah, I was quite impressed by that hack :) [14:04] cjwatson: Impressed or sickened? :-) [14:04] *shrug* I've seen worse [14:04] Heh === thunderstruck is now known as gnomefreak === davmor2_lunch is now known as davmor2 === thunderstruck is now known as gnomefreak === thunderstruck is now known as gnomefreak [15:29] * norsetto wonders if TheMuso already went to bed [15:38] siretart: yt? [15:41] Anybody any idea about being dropped to a busy box during boot with an error message saying a raid partion wasn't found? After doing a dmraid -ay I can proceed but the fs is mounted read-only [15:43] asac: about to leave. what's up? [15:45] siretart: not sure if i already asked that, but what would it take to make ifupdown wpasupplicant integration use the dbus service? [15:45] or is that mission impossible ;) [15:46] tkamppeter: ping, regarding https://bugs.launchpad.net/ubuntu/+source/cups/+bug/276573 [15:46] Launchpad bug 276573 in cups "P2POutputStream: write error" [High,New] [15:47] asac: I'd think some appropriate tweaks to /etc/wpa_supplicant/functions.sh. why? [15:47] it is written in shell, but you could call some python helper, if that would help you [15:48] siretart: does wpa_cli have a dbus variant? [15:48] asac: wpa_cli does not deal with dbus. wpasupplican does, AFAIUI [15:50] siretart: hmm. isnt wpa_cli the command line tool to send instructions to wpasupplicant (which could listen on socket or dbus) [15:50] asac: wpasupplicant does already listen on a socket. that's what wpa_cli is using [15:51] jouni isn't that in favor of dbus, but favors a custom protocol [15:51] yeah. right. thats why i would think that teaching wpa_cli to speak dbus instead of socket would make it easy :) [15:51] dan is pushing the dbus interface AFAIUI [15:51] ok [15:51] asac: I don't see what that would buy you [15:51] perhaps you should talk to Dan about this [15:52] you seem to be in good contact with him anyways [15:52] anyway, need to run now. cu! [15:53] siretart: thanks cu [15:54] siretart: it would buy us that ifupdown would just work (TM) with dbus activated supplicant :) [15:54] what causes an error like this: "** (gdm-binary:6390): Warning **: Failed to acquire org.gnome.DisplayManager: Connection ":1.950" is not allowed to own the service "org.gnome.DisplayManager" due to security policies in the configuraion file" [15:54] Missing some dbus- or pk-config-file there? [15:58] MacSlow: Yeah, I'd also guess that perhaps there are two GDMs running? Someone already has the name. [15:58] MacSlow: I realize that's not what the error says, but that would cause being unable to own the name also. [16:00] tedg, on my intrepid-box it was missing /etc/dbus-1/system.d/gdm.conf [16:00] cjwatson: re LP: #274860. Reverting just that one initramfs commit did the trick. usplash is working again on PS3. Is it too late to get that one included for the beta? [16:00] https://bugs.launchpad.net/ubuntu-ps3-port/+bug/274860 [16:00] munckfish: too late for beta, but it can be queued up for right afterwards [16:00] Launchpad bug 274860 in ubuntu-ps3-port "usplash doesn't run on boot on PS3 in intrepid" [High,Fix committed] [16:01] cjwatson: ok understood. Anything I need to do make sure it doesn't get missed (other than bug you on here)? [16:01] munckfish: ps3 can't release with beta anyway since linux-ports-meta isn't in yet [16:01] cjwatson: doh! [16:01] munckfish: I'll do it right now [16:01] cjwatson: still not in? [16:02] So there won't be a PS3 beta [16:02] no, it was uploaded, but all the package names were wrong [16:02] doh (weeps ....) [16:02] superm1: rotate X crash is bug #276782 [16:02] Launchpad bug 276782 in xserver-xorg-driver-ati "X server crashes on rotation with ATI driver." [Undecided,New] https://launchpad.net/bugs/276782 [16:03] cjwatson: so our """beta""" will have to be the next daily build afterwards [16:04] munckfish: uploaded, it'll sit in the queue [16:04] munckfish: yeah, sorry [16:04] cjwatson: which one is uploaded? [16:04] superm1: powertop is complaining: a USB device is active 100.0% of the time: /sys/bus/usb/devices/2-2 [16:04] ports meta? [16:05] munckfish: initramfs-tools [16:05] oh great thx [16:05] Stefan Bader is doing linux-ports-metqa [16:05] -q === thunders1ruck is now known as gnomefreak [16:05] yep ok [16:05] cjwatson: thx [16:07] seb128: hi seb [16:08] seb128: this is the packages list I'm currently using in generating the desktop packages list: http://bazaar.launchpad.net/~norsetto/+junk/main/annotate/20?file_id=packages.list-20080929175829-yjv3cr6q0lssue9a-1 [16:08] hey norsetto [16:09] norsetto: what is the file format? [16:09] seb128: csv [16:10] norsetto: right, but the columns? the 0 and 1 value for example [16:11] seems to be source name, upstream name, something, usual uploader, watch? [16:11] seb128: ubuntu package name, upstream package name (if different), blacklist (0 will download from ftp.gnome.org, 1 will not), maintainer, watch line [16:12] seb128: yes, that something is if you don't want to consider the package in ftp.gnome.org (for instance gimp, or ubuntulook) [16:12] norsetto: I'm wondering if we should not simply have watches for everything, we need at least a way to specify the serie we are tracking [16:13] seb128: for the series its a command line argument [16:13] seb128: --gnome 2.24 or --gnome 2.25 etc. [16:14] norsetto: well, we need that in the file, because we might track gdm 2.20.n [16:14] norsetto: and libcairo 1.7 and pango 1.23 [16:14] seb128: in that case you simply add the correct watch file [16:14] seb128: and you blacklist ftp.gnome.org for those packages [16:15] norsetto: ah, that makes sense, so 2.24 by default and we overwrite those where we need a different version [16:15] seb128: yep [16:15] good [16:15] we should probably move that somewhere in the ubuntu-desktop or desktop-bugs bzr [16:15] seb128: that would be cool [16:16] norsetto: what's the idea behind that list? [16:17] nxvl: to generate this: http://norsetto.890m.com/desktop_packages.php [16:18] norsetto: let me ask on #ubuntu-desktop if people have an opinion on the bzr to use [16:18] nice [16:18] mathiaz: ^^ [16:20] norsetto: What about "last Ubuntu uploader" and "number of open sponsorship bugs"? [16:20] Or some kind of indicator that there are bugs awaiting sponsorship [16:21] Laney: the first column has the usual uploader [16:21] seb128: I see that, but it's not filled in very much. Or can we assume that if there is nobody there then it's free for anyone to do? [16:21] Laney: if you give a bug number in the comment field, you will get a link through the package name, othewrise that will point to the relative sponsoring queue for bugs on that package [16:22] pitti, can I bother you a bit with ConsoleKit-related questions? [16:22] Laney: right that's the idea [16:22] norsetto: I'm just thinking that it might make it easier for sponsors if there was a way to scan down the page and see if there is anything pending. Maybe use the LP API to colour the links or something? [16:22] MacSlow: what's up? [16:22] Laney: the page is still under work, we will file informations [16:23] seb128: I know, just giving you guys suggestions :) [16:23] ok ;-) [16:23] this page is mainly for tracking versions [16:23] Personally I'd always be wary of doing an update without pinging someone first, hence the last uploader field. [16:23] I think we should rather use harvest or similar to track sponsor request, etc [16:23] pitti, I wonder if $PREFIX/sbin/ck-log-system-restart & Co are meant to be called directly via a terminal [16:24] pitti, I'm trying to figure out why my graphical-greeter isn't able to trigger a reboot or shutdown [16:24] Laney: I can add a last uploader if its required, but note that its just one click away now [16:24] Laney: consider #ubuntu-desktop as being the contact when nobody is specifically listed, maybe that should be mentionned somewhere on the page too [16:25] MacSlow: ck-*log*-? that won't actually restart [16:25] pitti, I understood taht to be logging also ... not only logging [16:25] MacSlow: gdm should use the d-bus calls [16:25] * Laney nods [16:25] pitti, that's what I do in my greeter... but nothing happens [16:26] MacSlow: what's the result of the call, PermissionDeniedByPolicy or so? [16:26] MacSlow: did you try in a terminal with dbus-send? do you have a CK session (ck-list-sessions)? [16:26] pitti, I don't know... there's nothing logged [16:27] pitti, dbus-send ... uff... what a monster [16:27] one sec [16:27] MacSlow: hm, I'm not actually sure which d-bus call is used for reboot/shutdown [16:28] MacSlow: there's Restart() and Stop(), not sure what they do (I don't want to reboot right now) [16:28] seb128: what about adding: "If no usual updater is listed, consider pinging the last updater or #ubuntu-desktop." ? [16:28] MacSlow: you can use d-feet :) [16:28] MacSlow: nice interactive d-bus browser [16:28] pitti, I'll try to salvage the needed dbus-path/interface from the sources [16:28] MacSlow: f-u-s-a and gnome-session do the same CK calls, so shouldn't be hard [16:29] norsetto: looks good [16:29] MacSlow: give a try to d-feet first, that's much easier, and once it works, you can copy&pate the object paths from it [16:29] seb128: added [16:29] norsetto: thanks [16:30] seb128: welcome, I really hope this tool can be usefull [16:31] norsetto: no doubt about that, it's already useful now since it lists things which are not update and that we need to update after beta [16:31] norsetto: we are doing quite ok looking at the list ;-) [16:32] seb128: cool :-) let me know if you get any problem, I'm sure there are still bugs lurking somewhere [16:33] Could anybody suggest my a Python package that uses distutils and cdbs, and another which uses distutils and debhelper, to study, please? [16:33] * me === superm1 is now known as superm1|away [16:35] Heh. It seems like every time i look at this window, there’s a “superm1 is now known as superm1|away” on the screen. [16:43] tedg: I've updated bug 261084 for you, let me know if I didn't get what you needed [16:43] Launchpad bug 261084 in gnome-power-manager "Suspends again right after resume" [High,Confirmed] https://launchpad.net/bugs/261084 === thunderstruck is now known as gnomefreak [16:48] bdmurray: Hmm, so the code that was stopping you from unplugging your laptop and then doing suspend, which got removed, was probably hiding a debounce issue. [16:48] bdmurray: Seems silly that GPM should have to debouce keys. [16:48] what module is used for hotkeys? [16:48] or better: keyboard shortcuts [16:50] tedg: what code? [16:51] ogra: you recently added 10-samsung-Q1-keymap.patch to hal-info; that isn't upstream yet, can you please send it to the hal list? [16:51] bdmurray: GPM used to have a mis-feature that it wouldn't allow power events to happen within a couple seconds of each other. Supposedly this was to avoid breakage on some obscure laptops. The result of that was that if you unplugged your power adapter and then closed the lid your computer, it wouldn't suspend. That code was removed. But it seems that it was likely hiding this debounce issue. [16:53] tedg: there's always something exciting === asac_ is now known as asac [16:56] It's still odd to me that a computer would have a suspend button. I mean, isn't that what the monitor is for? Seems PC manufacturers get paid by the LED/Key. [17:00] tedg: if the computer and monitor were integrated, like Apples did ... [17:01] bah, who needs monitors? [17:02] all the geeks who otherwise don't get a tan [17:05] kirkland, hi [17:06] tkamppeter: hey, hit a cups regression, was trying to debug [17:06] kitkland I have seen your bug report [17:06] tkamppeter: k ... the gutenprint driver works for me [17:06] kirkland, it is most probably not CUPS, but the driver, HPLIP. [17:06] tkamppeter: ah, right [17:07] kirkland what you should do is at first to update to the current state of Intrepid. [17:07] tkamppeter: i'm already there ;-) [17:07] Then start system-config-printer and there call the troubleshooting wizard in the Help menu. [17:08] tkamppeter: cool, i'll do that [17:08] This sets CUPS to debug log mode and coillect a lot of useful info in one text file. [17:09] Before starting the wizard, have a print queue handy which uses the default driver, where the bug occurs. [17:09] tkamppeter: k [17:10] You do not need to try with the Generic PostScript driver any more, as your printer is not a generic PostScript printer. [17:10] tkamppeter: ;-) right [17:10] tkamppeter: that was my second attempt to get something to go to the printer [17:11] when you reach the step of the wizard where you get asked for printing something, print a job which causes the bug. [17:11] tkamppeter: shortly after "telnet officejet 9100; asdfasdfasdfashjas;dihfwjfnaskdjf" [17:11] :-) [17:11] k [17:11] To get data sent to a network printer so that the printer prints it, you must do [17:12] nc -w1 9100 < [17:13] Perhaps also the telnet works. The text must have an end-of-page in the end to let the printer print the plain text page. [17:14] Now back to the wizard. Doing investigations with the wizard only works if you send the job through the queue with the broken driver, so that the errors of the job get captured in the error_log of CUPS. [17:15] tkamppeter: nah, the telnet just sent it garbage, but it proved i could talk to it over 9100 [17:15] * kirkland changes the driver back to the hplip foomatic [17:16] foomatic/hpijs [17:16] When the job has finished (or got stuck) then you complete the wizard which generates a file. This file please attach to the bug report. [17:17] Riddell, have you already ported the troubleshooting wizard of system-config-printer to the KDE version? [17:18] tkamppeter: the troubleshooter error messages dialog shows: http://pastebin.ubuntu.com/52894/ [17:19] kirkland, nothing else? Therer must be other messages before that concerning the Job 59. [17:20] tkamppeter: i just attached the troubleshoot.txt to https://bugs.launchpad.net/ubuntu/+source/cups/+bug/276573 [17:20] Launchpad bug 276573 in cups "P2POutputStream: write error" [High,New] [17:21] tedg: when did this change happen? [17:21] tkamppeter: shall I'll move the bug from 'cups' to 'hpijs' ? [17:21] tkamppeter: no [17:22] <\sh> can someone let ia32-libs*ubuntu15 go through the buildds pls? [17:22] <\sh> will be up on ftp. in no time [17:28] kirkland, yes. [17:28] tkamppeter: done [17:28] Riddell. the troubleshooting wizard is very useful. [17:29] Riddell, what about s-c-p-kde in general, will it be the default printer setup tool in Intrepid? Now where KDE Printing Manager is gone? === robbiew is now known as robbiew-away === superm1|away is now known as superm1 [17:31] kirkland, it seems that the troubleshooting wizard did not cope well with your job. the piece of error_log it supplied did not help much. [17:32] i just printed a test page... should i print something more complex? [17:33] Please do the following: In s-c-p, select "Settings" in the "Server" menu and activate the debugging info (last checkbox). [17:33] kirkland, then print a job, the test page is enough. [17:34] Or better, to make sure that all works right, do "cancel -a" before sending the job. [17:35] After your new job is ready or stuck, copy the /var/log/cups/error_log to your home directory. [17:35] Open the copy with an editor and look for the last occurrence of "P2POutputStream: write error". [17:37] Look for the [Job xx] signature in this line and then go back to the first line with this job signature. Go some more lines back to see the two [Job ??] lines. [17:37] tkamppeter: better error log posted to https://bugs.launchpad.net/ubuntu/+source/hplip/+bug/276573 [17:37] Launchpad bug 276573 in hplip "P2POutputStream: write error" [High,New] [17:37] Delete all before these lines and save. [17:38] tkamppeter: yes, s-c-p-kde is the printer setup tool we're using, it misses quite some features still but the basics are there [17:40] Riddell, can PPD option defaults (PageSize, input tray, ...) be set? [17:40] kirkland, I have found the problem, see the line [17:40] D [01/Oct/2008:11:36:06 -0500] [Job 61] Could not create temporary file: Permission denied [17:41] tkamppeter: yup [17:41] tkamppeter: what is that path, where it can't create the temp file? [17:41] tkamppeter: not yet I'm afraid [17:41] It seems CUPS stopped setting the TMPDIR variable. I will add a foomatic-rip task asking to support this case. [17:43] kirkland, this requires a fix in CUPS or a compatibility enhancement in foomatic-rip, I think the latter is easier to do for Intrepid (there I am upstream). [17:43] tkamppeter: nice ;-) [17:46] tkamppeter: my wife says she's excited that she found a "real bug" :-) [18:05] mathiaz: sorry for bugging you, but we have another source bug fix in landscape-client, this one much smaller than the last: Bug #260230 [18:05] Launchpad bug 260230 in landscape "sysinfo disk plugin prints unnecessary warnings about the / device when over capacity" [Critical,Fix committed] https://launchpad.net/bugs/260230 [18:05] I've already done the upstream release and prepared the branch for the debian update [18:06] radix: ok - could you look at lp:~ubuntu-core-dev/landscape-client/ubuntu ? [18:06] mathiaz: sure, that's the branch I based mine on [18:06] radix: I merged a fix from kirkland yesterday [18:07] radix: great ! [18:07] mathiaz: oh, you want me to review that change? [18:07] mathiaz: I think I already did, but I'll check the branch to make sure it's the same [18:07] radix: ok - it adds a timestamp to the -sysinfo wrapper [18:09] mathiaz: yep, that's the change I already reviewed, looks good [18:11] ok, lunch time [18:22] bdmurray: In the Intrepid cycle, I don't remember the version exactly. [18:23] bdmurray: We actually discussed distro patching it at UDS, and then upstream did it :) [18:23] tedg: you think it is in g-p-m now though? [18:23] bdmurray: Correct, currently it does not suppress duplicate events. [18:24] bdmurray: So I think we now need to add code to debounce the keys. [18:25] bdmurray: Or, perhaps it would be better to just clear all keys entered before a suspend. Hmm, not sure how to do that one. [18:27] slangasek: Due to being briefly included in kdepim, kmobiletools got binary promoted to Main. I just did an upload (not noticing that). It needs to get demoted to Universe and then accepted if you would ... === njpatel is now known as njpatel_away [19:02] ScottK: kmobiletools is still a recommends: of kubuntu-desktop in the archive and in the seed; how does an upload of kmobiletools change any of that? [19:04] hello, how can I bypass the graceful shutdown and make it just die when the power button is pressed? (without disabling acpi) [19:06] hwilde: edit /etc/acpi/events/powerbtn [19:06] Or rather, the shell script that it calls. [19:09] hmm [19:09] it calls /etc/acpi/powerbtn.sh [19:09] that just says [19:09] /sbin/shutdown -H now "Power button pressed" [19:09] but why does it take 7-9seconds [19:09] I want it to shutdown -H now [19:16] shutdown changes the runlevel, for a graceful shutdown [19:17] you could do 'poweroff -f' instead, I guess [19:17] not that this is good for your data integrity [19:18] yeah i know i know [19:18] :/ [19:18] not good for the electrical system if people expect immediate shutdown [19:19] then they start hitting the button repeatedly and power flickers [19:28] \sh: you mentioned a new ia32-libs upload does that fix bug 271550? [19:28] Launchpad bug 271550 in ia32-libs "ia32-libs missing libQtDBus, others?" [Unknown,Fix released] https://launchpad.net/bugs/271550 [19:32] hwilde: if the button is under ACPI control, how does hitting it repeatedly cause the power to flicker? :) [19:35] its probably a special button that accumulates the shutting down ... and in the end you shut down your whole house/office, who knows [19:35] bryce: re. bug 177658, I'm currently unable to boot into intrepid, so, can't help I'm afraid [19:35] Launchpad bug 177658 in xserver-xorg-driver-ati "[hardy] Blank GDM screen with ATI R420 [X800XT PE]" [Unknown,Confirmed] https://launchpad.net/bugs/177658 [19:36] hwilde: "kill -9 0" [19:36] tedg heh [19:36] poweroff works good [19:36] norsetto: ok thanks for letting me know [19:37] bryce: Is the "X not finding HAL" issue an already reported bug? [19:39] norsetto: could you comment onto the bug to this effect, so we can keep it open until you're able to boot? Else since it's been a while since you commented someone might close it inadvertantly [19:39] bryce: sure [19:40] tedg: m, not sure. what in particular is the problem? [19:40] tedg: LP #275825 [19:40] Launchpad bug 275825 in hal "After an upgrade to intrepid, HAL prefers old fdi-cache to new FDI files" [High,Confirmed] https://launchpad.net/bugs/275825 [19:40] tedg: in general it doesn't hurt to file extra bugs; duping is cheap [19:40] bryce: Boot up, in GDM but not keyboard or mouse activity. Then I can drop to a console, restart X, and everything is happy-happy again. But it's annoying to do that after every reboot. [19:41] bryce: The problem is that X doesn't find HAL, and so it thinks there are no input devices. [19:41] Oh, if restarting X makes it work, it’s probably a different bug. [19:41] tedg, /etc/rc.local sleep 5 && /etc/init.d/gdm restart :) [19:41] tedg: is this after an upgrade or a fresh install? If it's after an upgrade then yeah sounds like the bug ion pointed to [19:41] "doesnt find" ? [19:41] bryce: So X needs to listen on DBus for when HAL comes on. [19:42] ogra, I'm guessing it hasn't started to the point of publishing a name on DBus yet. [19:42] tedg: I've not seen that issue myself on any of my machines, but go ahead and file a bug on it. [19:42] bryce: It was much worse with nvidia drivers 177, better with 173. I guess the 177 ones are faster ;) [19:43] probably gdm needs a later bootsequence number (until we have full upstart) sounds hard to implement ot make X listen on dbus [19:44] tedg: oh this is with -nvidia, hmm [19:44] tedg: can you reproduce the issue with -nv or -vesa? [19:44] ogra, well it's already listening on DBus to get device insertions. It is just a matter of adding listening for DBus name change events. [19:45] That is, if I'm correct in assessing the problem :) [19:46] oh, i wasnt aware, i thought it leaves that part completely to hal [19:46] It does, but HAL reports that as a dbus signal. [19:47] bryce: I can try. Painful drivers those be. I can't think of a time it's happened with bulletproof X though. [19:48] bryce: Which seems to come up on every kernel upgrade. I'm unsure why DKMS isn't rebuilding my modules -- but I haven't looked into that at all. [19:48] bryce: BTW, nice work on the new menu system there. Lots more useful stuff, I was playing with it last kernel update :) [19:54] tedg: thanks [19:54] tedg: yeah I have a system that's been failing to rebuild a module for a while too; haven't had a chance to investigate [19:56] tedg, bryce generally if you are missing the matching -headers package for a kernel [19:56] but the latest meta should always be pulling them in [19:56] I'm finding both gb.archive and cdimage very slow at the moment; is that true for others as well? [19:59] mdz_: I've been finding archive very slow - from 2200K/s (my limit) to ~100K/s in the last two days [19:59] dont' know if that's related [20:06] ieeee hyptonic brown poka-dot spiral! [20:07] kees: sounds like a new exploit ;-) [20:07] seb128, slangasek: should I raise an FFe for totem 2.24.1? It fixes 2 bugs + 1 typo. [20:08] norsetto: typoes are also bugs, so that doesn't sound like there are any new features that need excepting [20:08] slangasek: ok, thx for confirming [20:13] Nafallo: "must ... install ... malware" [20:13] lol [20:14] kees: that's better than "must ... upload ... ubuntu" :-) === nxvl_ is now known as nxvl === dwatson is now known as davewatson [20:24] I love ubuntu :) === mvo_ is now known as mvo [20:45] <\sh> bdmurray: the problem was libqt4-{core,gui} -> has no content then "transitional"...I fixed that now again [20:49] \sh: okay, great! I just hadn't seen the bug updated but saw you mention a new package here [20:50] <\sh> bddebian: libqt4-dbus etc. were included in -14...but I wasn't realizing that half of the packages of libqt are now named differently [20:50] <\sh> it should be working now..(at least I hope so) [20:51] Sweet thanks. ;-P [20:52] jcastro: yo, i linked as many virt-manager bugs to upstream bugs in rh/fedora as i could [20:56] slangasek: just a heads up, I have some patches for -ati pending independent validation [20:57] slangasek: all patches are cherrypicked from upstream and correspond to LP bugs. Just waiting on testing confirmation that they do fix the issues. http://www.bryceharrington.org/ubuntu/Ati/xserver-xorg-video-ati_6.9.0+git20080826.a3cc1d7a-2ubuntu4~bwh2.debdiff === apachelogger_ is now known as apachelogger === Adri2000_ is now known as Adri2000 [21:11] slangasek: Urgh. I'll work with Riddell on seeds then. [21:42] hi [21:42] What's the best way to produce a deb from a ruby script? [21:43] Who should I talk to for weird udev behavior ? === ivoks2 is now known as ivoks === iulian is now known as spider [22:00] kirkland: thanks for that, it's no longer at the bottom. :D [22:00] jcastro: \o/ [22:01] A question for packaging a deb [22:01] plz [22:01] kirkland: in cases where you're sure it's upstream but there is no corresponding bug, you can open an upstream task, and leave it as "I have no way to link it" - that generates a target list to upstream so we can batch submit later [22:02] kirkland: basically, if you can just make a determination that it is upstream I can go in and file those for you [22:02] jcastro: ah.... well there's a ***ton*** i can do that for [22:02] What's the best way to produce a deb from a ruby script? === spider is now known as iulian [22:02] jcastro: about 20+ of those are trivial gui interfacy changes that should be upstreamed [22:02] I used checkinstall and dpkg-deb which one is better? [22:02] oh, I can bust those out easily if you just mark them as you see them. :D [22:05] jcastro: okay, i moved onto a couple of other things at the moment, but i'll get back to it ;-) [22:09] no worries. [22:27] slangasek: I just unseeded kmobiletools, so it should demote itself eventually. [22:28] ScottK: ok [22:33] good evening [22:33] i am an able linux programmer and i might want to contribute into ubuntu project even thought i dont use ubuntu myself [22:34] i would just need a push to right direction from a friendly mentor [22:34] i can do C, C++, C#, Java, Python, Lua, Bash, HTML, MySQL, Javascript ... [22:34] PHP [22:35] i have used Linux for many years and i have contributed to some open source projects before [22:35] ok, now come on, why did you even admit that last one :) [22:35] what is it you're hoping to contribute? [22:35] i dont know [22:35] anything [22:35] which involved programming [22:36] *involves [22:36] im from Finland [22:36] its actually the middle of the night here [22:37] i think i would like to start from small stuff [22:37] then if it goes well, go into bigger stuff [22:37] i dont expect to be paid or even thanked for [22:38] henux: well, there's a list of bugs that have been marked as 'bitesize', implying that they should be good places to get started: https://bugs.launchpad.net/ubuntu/+bugs?field.tag=bitesize [22:40] okay [22:41] i will take a look of that list [22:43] i would have to install ubuntu linux into my system to fix these [22:44] and these do not actually involve real programming, only scripting mostly [22:45] i was thinking if there were actually any small tools perhaps which the ubuntu community would need and i could do [22:46] but i will nonetheless take a look of these ... [23:02] *** New desktop images are appearing on http://iso.qa.ubuntu.com - any testing help would be appreciated! *** [23:02] * Ng blinks at console-kit-daemon having 1024 file handles [23:03] * slangasek blinks as well, and wonders if that hints at the cause of the fclose() crasher bug [23:03] it's all /etc/ConsoleKit/run-session.d and /usr/lib/ConsoleKit/run-session.d over and over and over [23:04] stop using your console :P use X [23:04] ogra: er? [23:04] only 199 fds here, but yes, definitely leaking [23:04] * ogra is making silly jokes ... [23:06] tedg, do you have a bug about the broken g-p-m icon ? (its not showing AC status at all here, i always have a battery) [23:07] ogra, I haven't see one about the icon broken 100%, but there is a preference that it'll hide when there is no action. Do you have that set? [23:08] ogra: have you restarted g-p-m lately? there was a bug last week about the icon not reflecting reality [23:08] no, i use my icon in tablet mode on my laptop, so i have it always on [23:08] (i use it for suspend) [23:09] slangasek, well, i think i rebooted after the last upgrade [23:09] hmm, I'm seeing the last thing ogra is [23:10] ok, well, I can confirm a bug that fits that description in the previous g-p-m, and it's fixed for me in the current one [23:10] I noticed the other day that it wasn't updting to reflect AC changes well enough [23:10] -0ubuntu3, uploaded 29Sep [23:10] it might be a theme issue though [23:10] i sse it right on ubuntu-mobile where i use the gnome theme [23:11] (human doesnt scale properly on 48px panels) [23:11] * Ng stabs it in the TERM [23:11] since it's possible I've not restarted with the new one [23:12] Ng: facile bug in console-kit's directory iteration; thanks for noticing that, I wonder if that doesn't fix bug #263245 [23:12] Launchpad bug 263245 in consolekit "console-kit-daemon crashed with SIGSEGV in fclose()" [Medium,In progress] https://launchpad.net/bugs/263245 [23:12] slangasek: np :) [23:12] hmm, should just killing g-p-m and re-running it be enough? doesn't seem to have helped [23:12] it showed the correct status to begin with, but didn't change when I plugged AC in [23:13] yeah, that was enough to fix it for me [23:13] Ng, kill it, run the power preferences [23:13] that will start it again [23:13] and as i just saw that resets your gconf preference [23:13] so the icon is only shown on battery again [23:14] meh, why does it reset [23:15] it still doesn't seem to be updating the icon properly [23:16] no, it doesnt [23:16] it properly switches to battery if i unplug AC [23:16] but doesnt switch back [23:17] yup, not a theme issue, i see the same on ubuntu-mobile [23:17] oh, no, its just slow there [23:18] works with the gnome theme, just takes about 5 sec [23:18] it was slow to switch to battery, but after it did that and I put AC back in it's not changed the icon [23:18] and that was about a minute ago [23:18] yeah [23:18] same here on my laptop with human theme [23:18] * ogra tries to switch themes on the laptop [23:19] Ng, ogra: It freezes for me for 42.5 seconds on an AC state change - but then recovers after that timeout. [23:19] Ng: If you put your mouse over does it say that it's on AC? [23:19] tedg: yep [23:19] tedg: and if I bring up the full battery info it's correct [23:19] tedg, same here, its only the icon [23:19] . [23:19] Er. [23:19] Oh. [23:20] but switching themes didnt change it on the laptop [23:20] i wonder why it works on ubuntu-mobile [23:21] I have a problem with launching matlab, using gnome-termal it works, but when I use ALT-F2 or a launcher it does not, it only shows the splash screen for half a second and disappears. I asked allready in ubuntu+1, but no-one was able to help, maybe someone here knows what is going on? [23:22] * ogra points thvdburgt to #ubuntu-motu [23:22] ogra: Uh? NO. [23:22] slangasek: do you want a bug about the consolekit thing btw? [23:22] wgrant, ? [23:22] ogra: We are not support for proprietary software in there. [23:22] wgrant, but motu packages it, no ? [23:22] Ng: I am assuming, with no evidence whatsoever, that it's the same as bug #263245 [23:22] Launchpad bug 263245 in consolekit "console-kit-daemon crashed with SIGSEGV in fclose()" [Medium,In progress] https://launchpad.net/bugs/263245 [23:22] ogra: MATLAB is an expensive inverse clone of Octave. [23:23] wgrant, oh, right, i thought we had it in multiverse, sorry [23:23] We have Scilab. [23:23] and octave [23:23] Right. [23:24] Scilab is the one in multiverse (though it won't be there much longer) [23:24] right, i reviewed it once for edubuntu [23:24] wgrant, I understand you do not suport matlab, but you might be able to tell what the difference is between launching via a launcher or via a terminal [23:24] * wgrant has no clue about that. [23:25] the difference is whether your process has a terminal [23:25] using the same command I expected the same result [23:25] slangasek: on the basis that I know virtually nothing about consolekit, fair enough ;) [23:25] * Ng looks for the g-p-m one instead [23:27] slangasek: (fwiw c-k-daemon hasn't crashed on the machine I noticed it on, it's just erroring any time something asks it to do something) [23:27] Ng: pff, facts [23:28] haha :) [23:28] * ogra really doesnt get why he is seeing g-p-m DTRT on a samsung Q1 with ubuntu-mobile while it fails on the laptop [23:29] ogra: does it ever update the icon on your laptop? [23:29] Does anybody know why consolekit is reluctant to switch to a non-X VT the first couple of times it is tried? [23:29] The screen goes black, but then it jumps straight by to VT7. [23:30] Er, straight *back*. [23:30] haven't seen this [23:31] slangasek: It's easily reproducable for me. [23:31] wgrant: I've seen that [23:31] brb rebooting [23:31] wgrant, nope [23:31] it stays on battery [23:32] ogra: Huh. As I said above, mine hangs for 42.5 seconds (no tooltip, no context menu, tray icon doesn't rerender, etc.), but then it's perfect. [23:32] * Ng confirms bug 275855 [23:32] Launchpad bug 275855 in gnome-power-manager "on charge/discharge icon doesn't change" [Undecided,Confirmed] https://launchpad.net/bugs/275855 [23:32] wgrant, i have a proper tooltip and proper notificaions [23:32] just the icon doesnt change [23:33] i'll leave it on battery for a while to see if it shows the power state [23:33] slangasek: It seems that once it lets me change, it'll be OK for a few minutes. But after those few minutes it will be reluctant again. [23:33] Both of my g-p-m bugs seem to be brightness-related hangs, but I'd still expect to see your bug... [23:34] I don't know; it worked fine for me, other than the fact that it borked the permissions on my sound devices, woohoo [23:34] slangasek: Any hints on how to debug it? [23:35] wgrant, dbus-monitor might help [23:35] and the debug ode of g-p-m [23:35] ogra: I was speaking of the VT switching stuff, actually. [23:35] oh, sorry [23:35] I've commented with the g-p-m debug log snippets on my bug, though. [23:35] * ogra is talking in several channels [23:36] ah, and now I've tried to switch vts again and my machine has locked up [23:36] * slangasek glares at wgr [23:36] wgrant, too [23:37] slangasek: Oh, lovely! [23:37] locked up hard, it doesn't ping anymore [23:37] Niice. [23:37] woah [23:37] Wow. [23:37] wgrant, re bug 261721, what machine you doing this on? I've seen it a few times too, but i haven't been sure where to blame it [23:37] Launchpad bug 261721 in gnome-power-manager "LCD brightness OSD steals keyboard focus" [Undecided,New] https://launchpad.net/bugs/261721 [23:38] superm1: It's a Dell Inspiron 630m. [23:38] I presume it's separate from (but made more noticable by) the hanging bug. [23:38] wgrant, interesting. i wonder if gpm is making too many smbios calls [23:39] or moreover calls to the hal daemon for dell backlights [23:39] if you kill of hal-addon-dell-backlight, do they persist? [23:39] superm1: In one case it makes what appears to be one 42.5 second call, in another (using brightness keys) it seems to receieve the events thousands of times. [23:40] wgrant, looking over the time frame of when you started to see this, that bodes about when i was seeing it too [23:41] superm1: It still steals focus for ever, I think, but it doesn't hang. [23:41] and VT switching was the only solution; same assertion I had [23:41] (I without hald-addon-dell-backlight running, I can now get a tooltip from the tray icon - before I couldn't) [23:41] s/I // [23:42] superm1: What hardware are you on? [23:42] wgrant, this is unreleased hardware, but Dell hardware too [23:42] which is what made me think gpm is calling that hal daemon too much [23:42] Ah. [23:43] Bug #261724 [23:43] Launchpad bug 261724 in gnome-power-manager "Brightness OSD doesn't stop adjusting" [Medium,New] https://launchpad.net/bugs/261724 [23:43] it's hard to provide a useful data point with unreleased hardware though unfortunately [23:43] (no stable BIOS yet etc) [23:43] can an archive admin please pocket-copy openssh from -security to -updates? (and anything else outstanding too?) [23:44] I'm away from AC power right now, but I'll try the state change without hald-addon-dell-backlight in a few minutes. [23:48] wgrant, well fwiw, even without the hal daemon running, i can reproduce the really slow changing and loss of focus that's only recoverable from a VT switch [23:49] superm1: When you do what? Brightness keys, or AC state change? [23:50] TheMuso: same problem as before with PA, after a reboot; I suspect a race condition due to the new sound theme [23:51] wgrant, brightness keys [23:51] superm1: Right. [23:52] OK, even without hald-addon-dell-backlight it hangs for 42.5 seconds when trying to set the brightness on startup or AC state change... [23:52] i dont have the proper batteries for these machines, so I can't test an AC state change without the machine going down :) [23:52] Doing this: [23:52] TI:08:50:56 TH:0x8d7b640 FI:gpm-brightness-xrandr.c FN:gpm_brightness_xrandr_output_set,324 - hard value=15125, min=0, max=15125 [23:52] superm1: I just realised that I can see it on starting up g-p-m too. [23:53] between this and the loss of my power button turning the machine off, i'm really not happy with gpm in intrepid right now :( [23:53] It's the only thing that's giving me trouble, but it's giving me lots of trouble. [23:54] superm1, why do you blame g-p-m fo ryour power button ? [23:55] that's what the bug triagers set it to [23:55] not gpm's fault [23:55] the power button is probably acpi not gpm [23:55] yeah [23:55] g-p-m's preferences dialog is the one with the option for what to do when the power button is pressed. [23:55] ogra, bug 252795 [23:55] gpm is a silly frontend ... its either hl or acpi in the kernel [23:55] Launchpad bug 252795 in fedora "pressing the "Power" button shows a logout dialog" [Unknown,In progress] https://launchpad.net/bugs/252795 [23:55] the button itself works [23:55] superm1, heh, thats gnome-session [23:55] but the wrong dialog comes up now [23:56] I can never remember whether it's gnome-session or gnome-panel. [23:56] on (upstreams) purpose [23:56] ogra: Or is g-p-m invoking the wrong one? [23:56] no, its gnome-session [23:56] upstream made that change [23:56] to make the logout dialog come up when pressing the power button? [23:56] yeah [23:56] sorry, but that makes no sense? [23:57] Ng: well, of course I installed the package with the fd leak fixed before my laptop hung, so now it won't tell me whether my guess was right about the fd leak being connected to the crashes :) [23:57] i want to turn off the computer so i press the power button [23:57] slangasek: Right, I'll dig into it. I don't see it at all here, but I can probably put something in place to make sure things sort themselves out cleanly. [23:57] /apps/gnome-session/options/logout_option [23:57] change it :P [23:57] TheMuso: ok. It really does seem to be the case that PA never tries again to open the device, after the initial failure [23:57] slangasek: doh! [23:58] superm1: Any idea why g-p-m seems to be trying to use xrandr to change my brightness, and failing? Do you see that too? [23:58] xrandr controls brightness? [23:58] slangasek: I cheekily snuck it in as a new bug anyway, just in case it's not a dupe. I'm not sure how to reproduce it, but both my intrepid installs show it and have <2days uptime, so I guess it wil rear its head again if that doesn't fix it [23:58] It's calling gpm_brightness_xrandr_output_set, so I presume so [23:59] Or maybe it's changing the brightness of an xrandr output. [23:59] Ng: how to reproduce the fd leak? Trivial: let cron run [23:59] ogra, changing that option to shutdown and logging back in didn't appear to change the behavior when pressing the power button. you sure it's the same thing? [23:59] wgrant, btw, my battery seems to reflect the loss of power in the icon [23:59] i'm at 89% and i see the icon has changed