[00:00] cjwatson: around? [00:01] mneptok: hah, yeah, but then you have to fix the stuff he ships you. ;) [00:01] soren: sort of [00:01] cjwatson: You forgot to commit or push to the eucalyptus packaging branch when you uploaded it. [00:01] * cjwatson blinks [00:02] turnabout is fair play, I suppose :) [00:02] * soren blushes at cjwatson's blinking [00:02] :) [00:02] fixed [00:02] ta [00:02] odd, normally I use a wrapper script which causes me to remember to do that, in this case I must have forgotten ... [00:04] I'm still trying to get used to this setup. [00:05] I still haven't quite worked out when to use dch (and which heuristic to use) and when to use debcommit. [00:09] I use DEBCHANGE_RELEASE_HEURISTIC=changelog, DEBCHANGE_MULTIMAINT_MERGE=yes, dch to add changelog entries (which I do any time I make a non-trivial change), debcommit to commit anything other than trivial changes [00:09] DEBCHANGE_RELEASE_HEURISTIC=changelog makes dch put the "UNRELEASED" as the release, right? [00:10] So when you're releasing, you just "dch --release", save, "debcommit --release"? [00:10] Or is there a shortcut I'm not aware of? [00:11] soren: that's the way [00:11] (and DEBCHANGE_FORCE_SAVE_ON_RELEASE=no is nice, but that hasn't been merged to Ubuntu yet :() [00:13] Cool, thanks. [00:13] * soren adds to his .bashrc [00:14] soren: or .devscripts [00:18] slangasek: Ah. Thanks. [00:33] yes, you want it in .devscripts, I don't think it works right in .bashrc [00:34] I have http://paste.ubuntu.com/260110/ as ~/bin/ubuntu-release [00:34] watch out for dch -r's slightly coarse time granularity - you have to wait a second :-/ [00:39] * soren hasn't heard of debrelease before [00:39] it is for very, very lazy people like me [00:41] cjwatson: Ah, I see what it does. That's convenient. [00:43] * soren wonders if there's more he can reasonably manage to do before FF kicks in. [00:45] I have a new grub2 building here, I do hope it works ... [00:47] * soren throws in the towel and wanders off to bed [00:48] 11 months old babies don't care much when their parents go to bed. They'll happily wake up at 6 AM even though their father didn't stagger to bed until 2 AM. :) [00:55] I noticed that [00:55] GOOD MORNING DADDY ISN'T IT A LOVELY DAY [00:55] *grunt* [02:40] StevenK: That is a close one. [02:44] TheMuso: Which? window-picker-applet? It's all bug fixes. [02:44] StevenK: ah ok [02:48] https://bugs.launchpad.net/ubuntu/+source/jhbuild/+bug/419363 would require packaging and getting up on REVU and getting 2 advocates by tomorrow, huh? [02:48] Launchpad bug 419363 in jhbuild "jhbuild is heavily outdated, update it" [Undecided,New] === akgraner_ is now known as akgraner [03:01] maco - if you're talking about uploading to REVU and getting 2 advocates by tomorrow in order to meet the FF deadline, then it's too late ;) [03:01] chrisccoulson: ok [03:01] and FF covers everything or just main? [03:02] maco - it's everything. I haven't looked at whats changed in jhbuild though [03:02] major version releases, it seems [03:05] maco - that would need to follow the process at https://wiki.ubuntu.com/FreezeExceptionProcess [03:05] so i guess that means it freezes at the *start* of thursday, not the end of it [03:06] yeah, it gets all confusing due to timezones ;) [03:06] but the e-mail went out about 1hour ago [03:09] kees, what's up? [03:37] superm1: I'd like to test some changes to fglrx-installer and nvidia, but I don't have ATI hardware that fglrx works on [03:38] superm1: it's as simple as: sudo apt-get install execstack; sudo execstack -c /usr/lib/dri/fglrx_dri.so /usr/lib/libGL.so.* /usr/bin/amdcccle and then making sure X still loads, and ccc still works, etc [03:40] kees, has to be on karmic? [03:42] superm1: better if it is -- jaunty i386 -desktop kernel won't break if the executable stack is executed. [03:42] kees, well i've got a jaunty box right here actually thats i386. how do i undo this if this breaks? [03:42] sudo execstack -s /usr/lib/dri/fglrx_dri.so /usr/lib/libGL.so.* /usr/bin/amdcccle [03:43] if you're running jaunty i386 -desktop it won't break if it's broken. i386 -server will break if it's broken, though. [03:44] for jaunty it'd be better to test on amd64 [03:44] i've only got i386 -desktop [03:44] so still a good test case? [03:45] unfortunately not, the markings will have not affect there (since -desktop in jaunty lacks PAE and nx-emulation) [03:45] I can test nvidia amd64, and nvidia i386, but I've got no ATI anywhere. [03:45] (well, no fglrx-supported-ATI) [03:46] kees, i'd say throw something out on ubuntu-devel ML then and see if you can grab someone to help out [03:46] yeah, it's probably more a bug-fix than a feature, so I probably don't need to rush it in tonight. [03:56] kees: nvidia actually works with that turned on now? [03:57] Amaranth: dunno, haven't tested it yet. [03:58] Amaranth: nvidia is actually worse than fglrx; they're compiling with such an old gcc that it doesn't include ELF header with the stack flags. :P [03:58] kees: yay RHEL 2 support [03:58] heh [04:15] Amaranth: yup, seems to work fine with nvidia (at least on jaunty 180) [04:18] kees: One of their customers must have made it a requirement :) [04:18] yay [04:19] Amaranth: now I wonder if it was just fixed in newer drivers, I don't have -173 -96 or -71 hardware... [04:19] kees: Pretty sure it didn't work a couple years ago [04:20] well, I'll try to find some people with older nvidia hardware to test those. /me prepares a ppa [04:46] And now gnome-orca is broken [04:51] TheMuso: Right, python-speechd is in universe, and gnome-orca is in main. [04:53] TheMuso: This makes ubuntu-netbook-remix and ubuntu-desktop livefses unbuildable [04:56] StevenK: right I think pitti promoted speech-dispatcher but didn't promote python-speechd. [04:57] Not according to LP, he didn't. The source for speech-dispatcher is still in universe. [04:58] TheMuso: Is there an MIR bug for this? [04:58] StevenK: yes just a sec. [04:59] bug 419036 [04:59] Launchpad bug 419036 in speech-dispatcher "[MIR] speech-dispatcher" [Undecided,Fix committed] https://launchpad.net/bugs/419036 [04:59] hrm maybe I misread the bug... [05:00] although it is marked fix committed [05:00] Fix Commited means "MIR approved", Fix Released means "It's been promoted" [05:01] right [05:01] Anyway, I'll promote it now [05:01] ok [05:02] StevenK: thanks I owe you one. [05:02] TheMuso: Actually, there are a whole bunch of binary packages - do they all need promotion? [05:02] StevenK: No, you should only need to promote speech-dispatcher, python-speechd, libspeechd2, and libspeechd-ev. [05:02] libspeechd-dev even [05:03] Bah. Just missed the publisher. [05:03] Heh, no, I haven't. [05:03] haha [05:05] TheMuso: Done. [05:05] StevenK: thanks again [05:06] TheMuso: Should I demote gnome-speech while I'm at it? [05:06] StevenK: If you have a few moments, there are a couple of backports sitting in Jaunty New that it'd be nice to get pushed through. [05:07] * StevenK grumbles at ScottK [05:07] StevenK: may as well [05:07] I can do kontrolpack, but quassel's my upload, so if you could take that one, I'd appreciate it. [05:08] ScottK: If I'm going to do one, I may as well do them both [05:08] StevenK: Thanks. [05:14] * StevenK fiddles the overrides of quassel-dbg [05:17] ScottK: Accepted. [05:17] StevenK: Thank you. [05:18] ScottK: I should rescue protobuf from hardy-backports, too? [05:18] Please. [05:19] * StevenK blinks [05:20] libprotobuf3 is in main in jaunty, karmic and jaunty-backports, but universe in intrepid-backports? [05:21] Doesn't really matter for backports. [05:21] StevenK: It was in universe pre-Jaunty... [05:21] They are all equally unsupported and Main can build on Universe in backports. [05:28] ScottK: Crackports ignores the OGRE model? Ugh [05:29] StevenK: It pretty much has to or you'd end up having to do backports only promotions to backports stuff with new depends and we don't have a mechanism for that. [06:38] good morning [06:46] guten morgen dholbach! [06:46] hey highvoltage! [06:49] dholbach: how are things? [06:49] good good - I'm just waking up slowly :-) === porthose is now known as porthose|AFK [06:57] dholbach: heh, same. [07:06] Does anyone plan on filing an MIR for liblqr, as it seems imagemagick has not been built for ages due to requiring liblqr-1-0-dev [07:06] which puts it in dep wait. [07:07] * TheMuso checks to see if imagemagick can be built without it. [07:09] Good morning [07:09] TheMuso, StevenK: no, I just approved speech-dispatcher; I still had some questions for it, and seeds need changing [07:09] Morning pitti. [07:09] pitti: right. The issues pointed out were addressed, and questions answered. [07:09] ah, read the further backlog [07:10] StevenK promoted, was the MIR bug closed then? [07:10] yes [07:10] and was gnome-speech demoted along? [07:10] yes [07:10] great [07:10] I didn't demote gnome-speech, acutally [07:11] Should I drop the lot to universe? [07:11] oh ok [07:11] Yes nothing should be depending on it. [07:11] in main at least [07:12] pitti: good morning [07:12] Right, confirmed with checkrdepends. [07:13] hey ttx [07:13] pitti, TheMuso: gnome-speech dumped to universe. [07:13] ok great. [07:13] \o/ [07:14] ok seems imagemagick is doing fine without liblqr. Will upload assuming the build is successful, unless someone does plan to file an MIR for liblqr. [07:14] s/assuming/if/ [07:15] pitti: do you know what was the outcome yesterday night for axis2c and rampart ? Would you consider preemptively promoting them as well to let Eucalyptus reach the CD ? [07:16] ttx: oh, sure; sorry that this got dropped, yesterday was too crazy [07:17] pitti: I can only imagine, it was already crazy for me :) [07:19] ttx: promoted [07:19] pitti: thanks ! [07:20] kirkland: did you see this in component-mismatches: [07:20] o kvm: kvm [07:20] [Reverse-Depends: Ubuntu.Karmic virt-host seed] [07:20] * TheMuso is a little surprised at hearing his CPU fan spin up with the imagemagick build. [07:21] kirkland: seems this needs some seed fix? [07:35] * soren hugs pitti [07:35] pitti: Thanks for the axis2c and rampart promotions. [07:36] yw :) [07:38] * soren is puzzled why apport decided to bombard him with crash reports just now [07:39] soren: you're past-FF, now you can concentrate on crash reports. [07:40] ttx: heh :) [07:40] how consderate of apport [07:40] highvoltage: that's part of its AI adaptive features. [07:41] I pushed the crash nagging lever up two notches yesterday [07:48] pitti: I just tried the Eucalyptus build again, and it ended up in depwait? [07:48] pitti: Oh,I need to wait for a publisher run, don't I? === tkamppeter_ is now known as tkamppeter [08:02] superm1: hi [08:14] soren: yes [08:23] Does someone know why we pull things like ubiquity, casper, or even lupin-casper in livecd-rootfs instead of the live seeds? [08:23] I mean why LIVELIST="ubuntu-live^ laptop-detect casper lupin-casper" instead of LIVELIST="ubuntu-live^" + seeding? [08:26] I guess cjwatson or lamont might have an answer [08:27] lool: hysterical raisins, I suspect [09:03] If there are any main-sponsors or ubuntu-release members who would like to weigh in on LP 406245 it would be appreciated - it's missed FeatureFreeze whilst waiting for sponsorship, but I think it's important for Karmic [09:03] Launchpad bug 406245 in subversion "Merge subversion 1.6.5dfsg-1 (main) from Debian unstable (main)." [Undecided,Confirmed] https://launchpad.net/bugs/406245 [09:13] cjwatson, u bumped the dove ABI in d-i (just in case you are rolling a package right now, there is a pending change) [09:13] s/u/i/ [09:24] * pitti blinks [09:24] ValueError: Invalid value '"Medium"' for parameter 'importance': valid values are: "Unknown", "Critical", "High", "Medium", "Low", "Wishlist", "Undecided" [09:24] go launchpadlib [09:25] james_w: did you ever happen to see this by chance? happens since recently, probably with new wadllib or so [09:25] * pitti files bug [09:25] '"Medium"' != "Medium" [09:25] or am I wrong? [09:25] right, seems wadllib quotes it once too often or so [09:25] dholbach: right, it's probably that [09:25] my code didn't change in ages, though, and I don't see what's wrong [09:25] task.transitionToImportance(importance='Medium') [09:27] pitti: I've seen this yesterday, bug #418802 [09:27] Launchpad bug 418802 in python-wadllib "requestsync crashed with ValueError in validate_param_values()" [Critical,Fix released] https://launchpad.net/bugs/418802 [09:28] geser: ah, thanks [09:29] yeah, I need to update the bug [09:29] I forgot that it was blocked by the ugly python-oauth MIR [09:30] and I'm not sure what to do about it === Quintasan_ is now known as Quintasan [09:54] cjwatson: eucalyptus should reach the CD now that all of it was promoted... when should the cloud installer part hit the daily CD ? [09:59] ttx: it's in the archive, you know the rules I'm sure :) [10:00] cjwatson: cool, thanks :) [10:00] the node installation isn't really right yet though [10:00] and need a small change to CD bits to get a boot option in place [10:01] * cjwatson promotes eucalyptus-udeb [10:06] mvo: salut, apturl-kde additions pushed to trunk branch (not yet uploaded), please have a quick look (especially at the translation stuff, I am not quite sure how to test pot creation) and upload if you are good with it [10:07] apachelogger: thanks, will do [10:07] pitti: I'm slightly confused by bug 400179. It says you promoted pywebkitgtk, but it lists udev as the package and I don't see it in the overrrides file. [10:07] Launchpad bug 400179 in pywebkitgtk "pywebkitgtk main inclusion review" [Medium,Fix released] https://launchpad.net/bugs/400179 [10:09] package is still pywebkitgtk; I'll re-promote it [10:09] it's not the first time that a change-override.py command was just silently not working :/ [10:10] * cjwatson sighs at people projecting single bugs in grub2 that affect them into "it shouldn't be the default" [10:10] evand: done [10:11] BTW is anyone here successfully running GRUB2 with Windows ME? [10:11] there's a really really weird bug report about WinME thinking it's a virus [10:11] pitti: thanks! [10:12] Anyone have a Karmic machine and a USB disk (that they don't care about) handy and want to test for me whether DeviceKit-disks' PartitionTableCreate dbus method is broken? [10:15] evand: o/ [10:15] what do you want me to do? [10:16] I have an 1 GB USB stick which I can trash at will, thath should suffice? [10:16] pitti: cool. I'm seeing "Error creating partition table: timeout (10s) waiting for change" when calling PartitionTableCreate with 'none', [] in d-feet. [10:16] pitti: can you reproduce that? [10:16] 'none' literally? [10:16] yes [10:17] it will clear the partition table === dholbach_ is now known as dholbach [10:18] evand: is 'none' even legal? shouldn't that be "mbr" or so? [10:18] evand: right, I get the same error [10:18] evand: it does work fine with 'mbr', [] [10:18] if you say mbr when a partition table exists, you get an error, I believe [10:19] indeed, "device already has a partition table of given scheme" [10:19] * pitti tries [10:19] oh nice, palimpsest can also partition now [10:20] ok, I have an existing partition table now [10:20] (I'm implementing a "format the device" method here, so it needs to wipe everything (including the boot sector), and then create a msdos style partition table and a single fat32 partition spanning the disk) [10:20] okay [10:21] thanks! [10:21] 'org.freedesktop.DeviceKit.Disks.Error.Failed: device already has a partition table of given scheme' [10:21] confirmed [10:21] that's "mbr", [] with an already existing mbr [10:22] superm1: hi [10:22] evand: so it seems the bug is that 'none', [] works, but doesn't give a proper return code? [10:22] pitti: indeed, it's timing out, but otherwise works [10:22] ok, that should be fixed then [10:23] indeed, I'll file a bug report [10:29] pitti, have you ever submitted a patch to a vger.kernel.org mailing list? [10:29] tkamppeter: no, I didn't [10:29] the #ubuntu-kernel folks certainly did, though [10:34] pitti: filed as bug 419796 [10:34] Launchpad bug 419796 in devicekit-disks "PartitionTableCreate method times out when 'none' is specified as a parameter." [Undecided,New] https://launchpad.net/bugs/419796 [10:36] Hi [10:37] The console application won't display user information when running under X Session [10:37] It is correct behavior? [10:52] Which would be the correct mailing-list to ask a very technical question about python-support, CDBS, and adding a "python-mlt" binary package to "mlt" - the main issue being how to run python-support only for python-mlt (not all the mlt binary packages), or how to recreate python-support's binary-arch output manually? [10:52] Hi. [10:52] hm. it's not clear to me from xorg.0.log which input device is used in the event that xinput thinks there are "multiple" mice [10:52] Could anybody look on my library and ask god to include it into Ubuntu? [10:54] https://sourceforge.net/projects/cli2gui/ [10:55] directhex: all of them? [10:56] slangasek, hm, then i need to debug it. assuming xev isn't lying to me, it's not seeing any events at all from the side buttons or side-wheel [10:58] (II) Microsoft Microsoft Wireless Optical Mouse? 1.00: Found 5 mouse buttons [10:58] lies [11:02] directhex: no wonders it doesn't work with ubuntu [11:02] slangasek: now in NEW (again) :) [11:05] sebner: yep, accepted [11:05] * sebner hugs slangasek and showers with with cookies :) === happyaron_ is now known as happyaron [11:10] * hyperair wishes mdadm and lvm2 are installed on the livecd by default [11:23] hyperair: you can use the alternate CD to install with them [11:23] or you can just apt-get install them if you need to mount an LV or MD device [11:23] Keybuk: yes, yes i know. but i'd like to avoid having to apt-get install them all the time =\ [11:23] "all the time" ? [11:24] you break your MD arrays *that often*? :p [11:24] for recovery purposes [11:24] more like i'm trouble shooting grub, and it's refusing to read anything [11:24] Loading Stage1.5Read Error [11:24] remember that if they were on the Live CD, they'd be on the default install for everyone [11:24] so i tried grub2, and i got Loading kernel.. Read error [11:24] * hyperair whacks grub [11:25] eh? really? [11:25] Keybuk: not actually quite true ... [11:25] Keybuk: is it possible that e2fsprogs broke anything? Since the updates fsck fails on boot. Telling me root partition needs a manual fsck. I do and reboot, works. On the next day I start the laptop again, same issue [11:25] the installer would have to remove them for everyone [11:25] cjwatson: well, you could uninstall them, but that's just wasting effort ;-) [11:25] Keybuk: consider that ubiquity is not on the default install for everyone :) [11:25] sebner: fsck will give you detailed output, what is it? [11:25] the real reason we don't have them on the live CD yet is that it would cause quite a bit of confusion since ubiquity doesn't support LVM or RAID yet [11:26] cjwatson: actually we *removed* them from the set a while back [11:26] Keybuk: on bootup it just tells me, fsck died on /dev/sda and needs a manual fsck [11:26] we didn't want LVM or RAID on desktop installs because they're expensive [11:26] *dev/sda3 [11:26] Keybuk: right, I know [11:26] Keybuk: but nevertheless that's the reason they aren't in the live seed [11:26] what's in the desktop install is a red herring I think [11:26] sebner: and what does the output of fsck -y /dev/sda3 say? [11:26] cjwatson: I'd say that u6y shouldn't support them [11:27] I want it to (at least LVM, maybe not RAID), just haven't had time [11:27] the experience would be much better if we waited to support btrfs instead [11:27] Keybuk: Checking inodes etc, result is always 0.4% non-contiguous [11:28] sebner: see, what you're not doing here is pasting me the exact output [11:28] Keybuk: well, next time I boot up I'll make notes ;) [11:30] sebner: there's a particular e2fsck change that outputs a particular piece of information [11:30] if you don't have that, then no, I don't think the update will have "broken" it [11:30] you could simply be failing to unmount the filesystem properly on shutdown, etc. [11:31] ArneGoetje: was ttf-vlgothic meant to directly replace tt-sazanamic-gothic? because the bzr log says they're meant to be smaller, but vlgothic is 2MB larger than sazanami-gothic [11:31] Keybuk: Uhhh, that may be a point. since some days it hangs on shutdown [11:31] sebner: there you go then, unclean/dirty filesystem -> fsck required [11:31] Keybuk: not sure if it's because of the -7 kernel. I'll try [11:33] slangasek: ttf-vlgothic is preferred over ttf-sazanami-gothic by Japanese users. [11:34] ArneGoetje: but it cost us 2MB on the CDs that we already didn't have... :/ [11:34] slangasek: together with the other changes I made, we actually gained space according to the Size: numbers [11:35] ArneGoetje: which changes? [11:35] the arabic / chinese changes? [11:35] it's a slight net gain, yes [11:36] slangasek: http://pastebin.ubuntu.com/259752/ [11:39] slangasek: I might reshuffle and repackage fonts further to gain extra space while supporting as many scripts as possible on the LiveCD. [11:39] sebner: there are a couple bug reports blaming -7 kernel to hang on shutdown === torkel_ is now known as torkel [11:39] ArneGoetje: hmm, yep, those numbers look good [11:39] and in fact, alternates /are/ down in size vs. alpha-4 [11:39] so why are the liveCDs up [11:40] possibly livefs build failures [11:40] * ArneGoetje doesn't know how size calculation for the LiveCD works [11:40] much less precisely :-P [11:41] sebner: for instance #418509 [11:41] f-spot 0.6.1.1 should give you a couple of meg [11:43] ArneGoetje: right, this libgd2 conflict is breaking the livefs builds. Why does libm17n-0 care about xpm support? [11:43] slangasek: no idea [11:43] hmm, I'll have a closer look [11:43] slangasek: thanks === kapipi_ is now known as kapizilla [11:47] StevenK: would you have a minute to source NEW media-player-id? It's a small data-only package, but needed by new rhythmbox [11:47] hello all: I am using preseeds to make unattended ubuntu install. Well, someone knows how to tell installer NOT to install a package (Network-manager :-). Thanks! === mdz_ is now known as mdz [11:47] StevenK: it's more or less converted data to hal, and I'll maintain it, so it's fine for main [11:47] StevenK: s/to/from/, of course [11:51] torkel: cool, thx for the hint [11:53] james_w: retracers just failed with [11:53] File "/usr/lib/python2.6/dist-packages/launchpadlib/errors.py", line 20, in [11:53] from lazr.restfulclient.errors import * [11:53] File "/usr/lib/python2.6/dist-packages/lazr/restfulclient/__init__.py", line 19, in [11:53] import pkg_resources [11:53] ImportError: No module named pkg_resources [11:53] james_w: that's just a missing dep from -launchpadlib to python-pkg-resources, do you agree? I'll do the change now [11:54] restfulclient [11:54] I'll upload, sorrt [11:54] ah, ok [11:55] james_w: don't worry, I can do it, just checking with you [11:55] already half way through [11:55] heh, ok :) (so was I) [11:55] * pitti drops and fixes retracers instead [11:55] james_w: thansk! [11:56] thank you [11:56] I guess pkg_resources isn't something that people are required to put in setup.py? [11:56] as requires? [11:56] they shold, it's not a built in module [11:59] lazr.restfulclient_0.9.3-0ubuntu2 uploaded [11:59] yay you [12:00] and bug filed upstream [12:00] I'm concerned about python-oauth though [12:01] I thought using the other API version was okay? [12:01] i really should work out how to make my mouse not suck [12:01] stop using a dyson [12:01] it worked before the new hrad disk, so that implies it worked with manual xorg.conf manglement [12:03] pitti: I thought so, dobey has objections to using it though [12:03] pitti: the issue isn't entirely fixed for server implementations, but we have none of those in Ubuntu [12:03] james_w: is it strictly required in p-launchpadlib? it wasn't necessary so far [12:04] the old version had an embedded copy === MacSlow is now known as MacSlow|lunch [12:04] u1 still does [12:04] ugh [12:04] and dobey has objections to us using the system copy instead of the embedded one, for reasons that aren't entirely clear to me [12:04] but then we can just as well promote it, if it already has been in main covertly [12:05] an explicit package is so much easier to support than trying to find hidden copies [12:05] he has valid concerns about the code, but I don't see why that stops us using the exact same code from a single package [12:05] exactly [12:05] he is proposing a fork to fix the issues as he finds upstream to be too slow responding to his concerns and patches [12:06] agutierr: err, it's tricky to exclude a package from a task, easiest way is to use preseed/late_command to remove the package after it's installed [12:06] cjwatson, the problem, on the installer I only have apt-install, not apt-remove command [12:06] it's your call as to what goes in, I have made it clear to him that we are not shipping karmic with an embedded copy of oauth.py when it is packaged separately [12:07] I'm not sure what the situation will look like when we release though, potentially both projects will use the fork and we drop the python-oauth package [12:07] ArneGoetje: ok, no reason m17n-lib needs libgd2-xpm specifically - fix uploaded [12:08] ArneGoetje: thanks for your work on getting the CDs smaller :) [12:10] agutierr: apt-install is just a wrapper [12:10] agutierr: 'in-target apt-get purge network-manager' or similar ought to do it [12:12] ahh [12:12] ok!!! [12:12] so much thanks [12:14] TheMuso: why does speech-dispatcher use libaudio-dev? Surely nothing should be using that today? [12:14] TheMuso: (pulls it into the CDs now by default) [12:26] oh my, pulse as it is hacked up for arm support upstream now will never work if you dont compile it on the machine you want to use it on :/ [12:26] (reads /proc/cpuinfo inline from the code at buildtime) === Amaranth__ is now known as Amaranth [12:42] doko: does dh_pycentral change a #!/usr/bin/python2.6 shebang line into #!/usr/bin/python ? [12:42] doko: can I tell it not to? (it really only works with 2.6, and I would like to upload it to debian experimental) [12:43] pitti: no, not automatically. do you call python2.6 setup.py explicitely? [12:43] some build systems replace it automatically [12:43] doko: I don't, should I? [12:43] * pitti tries, thanks === Amaranth is now known as Amaranth__ [12:45] *blink* [12:45] /dev/sda1: Superblock last mount time (Thu Aug 27 14:34:07 2009, [12:45] now = Thu Aug 27 12:44:59 2009) is in the future [12:47] doko: it's correct in debian/tmp, though, just changed in debian/calibre [12:47] so it's not the upstream build system [12:48] pitti: which package [12:48] doko: calibre [12:48] doko: that's why I thought it was dh_pycentral [12:48] pitti: unstable, karmic? [12:49] doko: right [12:49] right what? [12:49] oops, karmic [12:49] :) [12:49] can't upload to sid (ENOPYTHON2.6) [12:54] pitti: well, you call python, not python2.6. that's like setuptools doing this === MacSlow|lunch is now known as MacSlow [12:55] doko: ah, thanks [12:55] likely even [12:57] doko: ah, no, the install command installs into debian/tmp/, and that is correct [12:58] the shebangs get changed in the debian//tmp -> debian/calibre mangling (and it's not dh_install, I'm sure) [12:58] but anyway, python2.6 setup.py install still does the trick [12:58] doko: thanks! [13:04] pitti: is bug 320743 filed with the help of apport? IMO apport should not file this kind of report, and point the user to remove the corrupt archive [13:04] Launchpad bug 320743 in java-access-bridge "package libaccess-bridge-java None [modified: /var/lib/dpkg/info/libaccess-bridge-java.list] failed to install/upgrade: corrupted filesystem tarfile - corrupted package archive" [Undecided,New] https://launchpad.net/bugs/320743 [13:05] doko: ah, can't build it for 2.6 anyway, since all the dependencies are only available for 2.5 [13:05] doko: looking [13:07] doko: indeed; apt shouldn't create those in the first place, but I'll filter them out in the hooks for now; thanks for pointing out [13:07] \o/ [13:09] is arky an automated tester? [13:25] pitti, a question about bug 419834 [13:25]