[02:28] perl module Curses::UI::Grid errors out. Could not load Curses::UI::0 from Curses/UI/0.pm Any ideas? [03:12] slangasek: Just saw your ping, I'm in HK for Connect and have a couple of days off after returning; I also miss a testcase for it because I found the bug by reading the code, I didn't really face any symptom; that said, I'll try testing this end of next week by running pieces of the new code manually === dendro-afk is now known as dendrobates [03:35] Good morning [03:36] bdmurray: ah thanks; NB that you committed this to the quantal branch :) [03:36] bdmurray: I'll reject and reupload [03:38] bdmurray: ok, reuploaded [03:47] cjohnston: hourly updates - niiice! [03:48] :-) [03:54] cjwatson: remove-package> \o/, thanks! [03:54] very slowly we approach sanity [03:55] Oooh, where's that? [03:55] lp:ubuntu-archive-tools [03:56] StevenK: see the ubuntu-archive@ announcement [03:56] cjwatson: wow, I hadn't actually expected you to be awake at this hour.. [03:57] I woke up at 1:30am and couldn't get back to sleep until I'd dealt with some of the stuff in my head. I'm about to go back to bed [03:58] StevenK: I have an MP up with the obvious nuke-from-orbit, too :) [03:59] cjwatson: Ah, I was wondering about that. [03:59] (hardly a rush, but why not) [04:00] cjwatson: Approved. [04:01] cjwatson: One thing to jot down somewhere is you can probably kill SoyuzScript at some point too. [04:55] cjwatson: reverted https://code.launchpad.net/~brian-murray/ubuntu-archive-tools/queue-tables/+merge/108073 FYI (see my followup) [04:55] just in case you wonder where it was gone [05:07] does anybody have an example how i can get the codename for a specific package with python-apt ? i had a look at http://apt.alioth.debian.org/python-apt-doc/ but found nothing. [05:07] toabctl: what is a "code name"? [05:10] pitti, afaik "precise" is a codename. i use reprepro to setup some private repositories and try to get the codename of a package. [05:10] oh, the release name [05:10] toabctl: there really is no way to tell from a .deb [05:11] pitti, eg here is the codename set: http://archive.ubuntu.com/ubuntu/dists/precise/Release [05:11] toabctl: right, usually you set up debmirror, reprepro and other mirroring by release, which will use the dists/.../Pacakges.gz indexes [05:11] pitti, imho apt should know the codename. [05:12] toabctl: you can, by looking at the .origins property [05:12] e. g. [05:12] cache['coreutils'].candidate.origins [05:12] there can be multiple origins, potentially from multiple releases [05:13] but that really sounds backwards [05:13] pitti, i did this but there is no codename [05:13] you want to ask your mirroring script "give me all precise packages" [05:13] not iterate over all pacakges and pick out the precise one [05:13] s [05:13] toabctl: sure there is [05:13] >>> c['coreutils'].candidate.origins[0] [05:13] [05:14] >>> c['coreutils'].candidate.origins[0].archive [05:14] 'quantal' [05:14] but really, this is not what you are looking for for mirroring.. [05:14] pitti, i don't want to mirror. i use this for private stuff [05:15] toabctl: ok; so the origins are the place to look for then [05:15] either the .candidate or .installed, depending on what you are looking for [05:16] pitti, where is the codename in your example? there's an archive. is that the same? [05:17] yes [05:17] hm. that's not set in my repository. maybe there's something wrong with reprepro then. [05:17] toabctl: it's taken from the "Suite:" field in e. g. http://archive.ubuntu.com/ubuntu/dists/precise/Release [05:17] toabctl: you need the Release file, perhaps you don't have this? [05:18] pitti, i have one, but Suite is missing. there is only a "Codename" field. [05:18] lool: ok; I'll ask for the package to be promoted in the meantime because the other bugs are high-impact, so maybe we'll drop the un-verified fix from the SRU [05:20] toabctl: just checked the source; it only looks at Suite:, not Codename: [05:21] pitti, where's the upstream source? imho that should be fixed. [05:22] Vcs-Bzr: says http://anonscm.debian.org/bzr/apt/python-apt/debian-sid/ [05:25] pitti, i added the Suite-Field to the Release file and i the the archive now. thanks! [05:25] works now? yw === dendrobates is now known as dendro-afk === Ursinha` is now known as Ursinha === Ursinha is now known as Guest90314 [07:02] good morning === smb` is now known as smb [07:14] doko: Stop doing toolchain uploads from conferences. [07:14] ? [07:15] ;) [07:15] doko: I was just checking on the buildds after I uploaded eglibc to see that you'd done gcc-4.7 just before that. [07:16] If both break the world, can we blame the tropical heat? [07:17] infinity: will your eglibc upload fix all the broken armel bits? :) [07:17] micahg: Which bits are those? [07:17] micahg: It fixed the last broken bit on armel that I know of... If you know of more, speak up. [07:19] infinity: I'm not sure if it's stuff still trying to do armv7 stuff or something else (and the porter boxen for armel are currently unhappy, rt open) [07:19] micahg: Stuff doing stuff or other stuff isn't all that specific. What issue were you seeing? [07:20] {standard input}:135: Error: selected processor does not support ARM mode `smulbb r9,r9,r3' [07:20] micahg: libc.so and ld.so on armel were both armv7 by accident (fixed now), but that's perhaps unrelated to your issue. [07:20] micahg: What build was that in? [07:20] that was firefox, ghc seems to have some similar does not support errors [07:21] Oh, right, GHC was pointed out to me earlier. Probably won't be fixed by my upload, no. I'll have to look at it later. [07:21] Unsure about firefox too, but when I find some round tuits, I can help investigate. [07:21] firefox builds should be fixed with the recent upload [07:21] chrisccoulson, ^^^ [07:21] firefox can wait for later, ghc is more important as I'm not sure if we should start the rebuilds until it's built on all archs [07:22] doko: chrisccoulson will be relieved to play angry birds again [07:22] I wouldn't like to play happy birds ... [07:22] micahg: I'll be looking at ghc shortly. [07:22] * infinity runs off now. [07:23] infinity: thanks, we'll rebuild firefox soon enough anyways and see if it cures itself [07:25] doko: sorry, I didn't get to chromium yesterday, hope to look at it today, are you blocked on anything? [07:25] micahg, no [08:29] StevenK: Yeah, I grep for stuff in it every time I kill something. Killing change-override.py will probably manage to drop a method or two from it. [08:30] pitti: Ah, hmm, I'll see if I can find a workaround [08:30] cjwatson: I already had a workaround for the non-working len() [08:31] cjwatson: but it still didn't work, all the queues reported 0, at least back in the days when I tried it [08:31] perhaps LP got fixed in the meantime to deal with lucid's lplib [08:53] ogra_: hey Oliver [08:54] ogra_: FYI, I gave you a WI on https://blueprints.launchpad.net/ubuntu/+spec/desktop-q-third-party-driver-installation for pvr-omap4, mostly because you know best how to detect whether the driver is necessary, and you can check "modinfo omapdrm_pvr" [08:55] ogra_: if that doesn't have any modaliases, we can add a custom detection plugin, like we had in jockey; but if it can be detected with modaliases, that'd be a lot more flexible [08:57] ogra_: of course I can help you wit this, but I'll need the initial investigation [09:09] I want to start developing for ubuntu.. can someone walk me through it?? [09:10] is anyone even here?? [09:11] skunk, sure. Most of them are probably asleep, at work, or doing other daily routine things. People tend to read scrollback though. [09:14] skunk, I can't walk you through it, but perhaps pointing you to the http://wiki.ubuntu.com/UbuntuDevelopment link in the topic would be a good starting point? [09:17] http://developer.ubuntu.com/packaging/html/ probably too :) [09:18] kk one more question [09:19] I have a java background.. should just touch up on it before learning Vala and C? i mean.. I do have to relearn how all these algorithms works eh?? [09:20] that entirely depends on what you want to do - there's java work to be done in Ubuntu as well :) [09:21] I want to be working with pixel perfect scrolling.. and improving that [09:21] as well as work to better develop multitouch [09:21] especially with synaptics touchpads [09:25] pitti, i plan to do some work on the panda anyway today, will get you the bits, np [09:25] ogra_: danke [09:29] what can i do if i have a synaptics touchpad, but ubuntu doesn't recognize it as one [09:42] pitti: http://bazaar.launchpad.net/~ubuntu-archive/ubuntu-archive-tools/trunk/revision/464 - you may vomit now [09:42] (note you still can't actually index the members of getPackageUploads() without authenticating) [09:42] haha [09:43] cjwatson: ah, that was it then -- this should work with login_anonymously(), and queue access doesn't work with that [09:43] No, that's incidental [09:43] cjwatson: thanks for fixing [09:43] This code doesn't attempt to index the members, just to get the length - I'm just noting that indexing the members still won't work [09:44] The wadllib monkey-patch is what fixes getting the length [09:45] Though that said one of my near-term work items involves creating a bot account on lillypilly, so all this can be authenticated; the LP team isn't actually desperately keen on anonymous access for various reasons, mostly monitoring/load-related [09:45] cjwatson: right, but my previous attempt on a workaround used len(list(getPackageUploads()), which requires accessing the members [09:45] Right [09:45] so that failed [09:45] This one's tested [09:45] screenscraping-b-gone [09:45] And deployed now [09:46] Maybe I should do that bot account now actually, after the next cup of coffee; we need that to move copy-report off cocoplum, and other suchlike things [09:47] I did check with the LP team about it; it's a bit alarming 'cos it's a privileged account, but it's essentially no worse than the alarming privilege of cocoplum [09:47] and at least oauth tokens are revocable [09:47] the old levels (such as "readonly") are gone now, right? (too bad, it was handy for things like that) [09:48] or do you want to automatically copy anyway? [09:48] copy-report doesn't want readonly [09:48] I want to preserve the automatic copy, it at least used to save Canonical roughly my salary per year in bandwidth :-P [09:48] *nod* [09:49] cjwatson: I still wonder whether it wouldn't be easier to just integrate this into whatever release script the security team uses (i. e. also copy to -updates, instead of mass-copying from cron) [09:50] pitti: you would need to give the security team access to -updates then ;) [09:50] We need the bot account for other reasons anyway [09:50] micahg: don't you? [09:50] nope, just -security AIUI [09:50] micahg: I wasn't aware that you need special upload privileges for -updates [09:50] The privilege around -security is specially hardcoded in LP [09:51] ah [09:51] Other reasons> for example it would allow auto-syncs to be actually automated [09:51] Rather than cron.cjwatson [10:46] Can I rely on update-manager / do-release-upgrade to remove `oldlibs`, i.e. dummy transitional packages? Or shall I write a hook for update-manager because I am reverting the transitional package rename which was done in Lucid. [10:51] cjwatson, are you still here?\ [10:54] well, cjwatson. I don't know if you're gonna see this message. But out of the six years i've used ubuntu the work you've done with the installer is incridble [10:54] skunk: thanks, and you're welcome [10:55] xnox, check with mvo but I don't think it does magically clean old stuff [10:55] 15 minute install, then im surfing the web, writing documents.. this kinda sound like Mac territory [10:56] cjwatson, who do I contact for uTouch issues?? [10:56] skunk: I'd start with cnd [10:56] seb128: i distincly remember the 'Remove old unneeded packages: Y/n' question from do-release-upgrade, But I usually dist-upgrade to +1 release, and end up not using the upgrade-manager. [10:56] (though it's not a field I know much about) [10:57] xnox, right, I just don't know if that list is computed in a smart way or if you need to teach update-manager about it [10:57] mvo: ^^^ can you check my message at :46 above. [10:57] cjwatson: honestly I want my synaptics touchpad to work like one.. instead of a standard touchpad [10:57] seb128: ok. I will test the upgrade in chroot =) [11:04] xnox: its not looking at transitional packages specifically, it does look at unused dependencies though, that might (or might not) help you [11:06] pitti: am I reading this correctly that we do not have the ability to retrace VmCore files yet? [11:06] in apport-retrace, that is [11:07] xnox: it could do heurisitc like looking at "transitional" or "dummy"in the description summary, but that is not deal, ideally there would be something formal identifying the transitional pkg [11:08] mvo: 'Section: oldlibs' is standard way to identify transitional packages, unless i am wrong. let me check the policy. [11:08] mvo: and they should be cruft to be removed. [11:08] Yeah, though we could do better about actually moving transitional packages there in Ubuntu in practice. But that's eminently fixable. [11:09] hmm policy says 'kept for legacy' [11:09] and some are not actually transitional nor dummy [11:09] xnox: right, iirc oldlibs was overloaded with multiple meanings [11:10] yeap [11:10] looks like so [11:10] I whish there was "section: transitional" or something similar or a debtag [11:10] I will do a test and hopefully I will be ok with depends/replaces/conflicts etc. [11:10] thanks [11:13] pitti, apw: also, we don't seem to have a close to unique identifier for KernelOops failures. Any objection to parsing the call trace out of OopsText (https://launchpadlibrarian.net/79771442/OopsText.txt) and generating a signature in the same manner as crash_signature (so concatenate the first few)? [11:13] * ev lunches [11:14] * xnox read * ev launches [11:15] but where to ? [11:15] * xnox To the cloud! And beyond! [11:15] :) [11:15] charming :) [11:15] * xnox loves ToyStory [11:16] ev, that doesn't sound like a bad plan, though in theory ignore the ? entries in the list ... also include the type of OOPS, the top line hints at th type, and the IP: as part of it [11:16] isnt that a requirement to become a DD ? : [11:16] xnox: launching stuff at poor infinity.. that is just mean. [11:16] Daviey, worse leaving him behind ... "and beyond" [11:17] nah, that just means infinity+1 :) [11:17] Daviey: I was certian that infinity is deployed with juju these days === MacSlow is now known as MacSlow|lunch [11:56] chrisccoulson, bug #1003733 is fixed, you can't verify when building with 4.6 :-/ [11:56] Launchpad bug 1003733 in gcc-4.7 (Ubuntu Quantal) "Angry birds does not work (test failures in js/src/jit-test/tests/jaeger/testSetTypedFloatArray.js with gcc4.7)" [Medium,Triaged] https://launchpad.net/bugs/1003733 [11:58] oh my angry birds =) [11:58] that must be critical priority =) [12:06] ev: that's right [12:07] ev: sure, if the succession of addresses in OopsText is reasonably reliable [12:22] xnox: Our offices are within 300 feet to their HQ, should I pay them a visit? :D [12:23] how would you pay this ... in icecream ? [12:24] I dunno, but I could sell them an ERP at the same time. :D [12:24] astraljava: yes, plese. show up and say that `doko is not happy and wants an angry bird stuffed animal shipped to germany` [12:24] xnox: lol [12:24] xnox: some of their employees walk around in pig caps, would that do? [12:25] pitti: awesome [12:26] apw: cheers for that - once I have an algorithm I'll run it by you [12:26] astraljava: yeap, don't forget another one to be shipped for me to uk offices =) [12:28] :) [12:33] ev, excellent thanks [12:34] micahg, ping === _salem is now known as salem_ === Guest90314 is now known as Ursinha === MacSlow|lunch is now known as MacSlow [12:46] doko, the new gcc appears to fix the issues btw (just did a test build here) [12:47] \o/ [12:48] thanks! :) [13:12] doko, any chance you could take a look at https://bugs.launchpad.net/ubuntu/+source/xmlgraphics-commons/+bug/888129/+attachment/3169273/+files/fix_icc_handling.debdiff before I upload? [13:12] Launchpad bug 888129 in openjdk-7 (Ubuntu Quantal) "xmlgraphics-commons version 1.4.dfsg-3ubuntu1 failed to build with openjdk-7" [Undecided,In progress] [13:14] jamespage, looks fine. sorry for the delay [13:15] doko, ack - thanks [13:20] mterry, I have no strong opinion either way. But if the badge is kept, maybe it should count the top-level groups of updates, not the individual packages. [13:20] mpt, hmm, k [13:21] mterry, but on the other hand, that seems a bit weird [13:38] mvo, let me ask you the other question from last night in this quieter channel [13:39] mvo, so update-manager --no-update. What is it supposed to do? I see the code it affects, but still not sure what exactly it means [13:44] mterry: my memory is not the best, but I think this got added for automatic testing via mago [13:45] mterry: I think its ok to kill it off [13:46] mvo, I want to recycle it. The new spec requires doing an apt-get update when a user opens update-manager, but we need a way to turn that off, for testing and for when we pop up automatic updates. --no-update seemed logical [13:48] mterry: yeah, that sounds reasonable [13:48] mvo, also, what's the story with the DBus API that update-manager exposes? Who consumes that? [13:50] mterry: update-manager itself to implement a poor mans singleton window [13:51] mterry: see setupDbus() but meh, its pretty rusty, need proper dbus names for example [13:52] mvo, OK, so it uses whether the name is claimed or not, much like modern GtkApplication. But does anyone consume the actual calls like "update()" or "upgrade()" that you know of? [13:53] mterry: those were added for mago as well [13:53] mvo, ah OK [13:54] mvo, (in the branch where I'm doing the update-on-start, I'm removing the ability to initiate updates elsewhere in the UI and was wondering about the dbus method to do so. Sounds like I should drop it too) [13:54] brendand: hi, iirc you added the update dbus stuff, is that actually being used these days in the test environment? [13:55] mvo - i'd have to check [13:55] mvo - i do know we have an apt-get-update test in our SRU suitew [13:56] mvo - whether it actually uses that functionality i can't recall [13:56] mvo, probably not [13:56] mvo, brendand: well regardless, if the purpose of the tests is to mimic the UI, if the UI is dropping the ability to kick of an update willy-nilly, the dbus API should to I suppose [13:57] mvo - what's the ui doing? [13:57] brendand, how do you mean that? The UI is moving to a "do an apt-get update on startup" and dropping the buttons to initiate one manually [13:58] mterry, oh right, i see [13:59] mvo, mterry - so i can say that *i* won't complain if it's removed. i can't promise no-one else will. i seem to recall that stuff was put in at the request of the QA team as well [14:03] NCommander: 12.04.1 meeting in #ubuntu-meeting === sagaci_ is now known as sagaci === tumbleweed_ is now known as tumbleweed [15:06] * ogra_ sighs about linux-base [15:06] that postinst is a pain [15:11] slangasek, FYI support for the arm server arches is in flash-kernel now, the only thing i'm struggling with is linux-base atm which needs to be MIRed (its postinst tries to rewrite all config files that could use UUIDs though, and there is no debconf option to switch that off) [15:13] ogra_: hmm, if we don't need/want that rewriting (which should be obsolete for years now), why not drop the dep? [15:13] is linux-base needed for something else? [15:14] slangasek, it ships linux-version (which is actually the only binary in there after tgardner dropped perf from it), which flash-kernel uses to determine the laters kernel version [15:15] *latest [15:17] i'm incliend to just put an exit 0 at the top of the postinst [15:18] ogra_, slangasek: I filed a MIR for linux-base this morning, FYI [15:18] * ogra_ hugs pitti [15:18] it seems fairly harmless [15:18] pitti, what else does use it ? [15:18] but I subscribed you and infinity to the bug for further comments [15:18] ogra_: just flash-kernel [15:18] bug 1006717 [15:18] Launchpad bug 1006717 in linux-base (Ubuntu) "[MIR] linux-base" [Undecided,New] https://launchpad.net/bugs/1006717 [15:19] it would be if the postinst wouldnt fiddle with hdparm.conf or udev rule rewriting [15:19] or bootloader configs [15:20] ogra_: so anything that would still be using /dev/[hs]d[a-z]* style names in any of these config files is already buggy, and *should* have been rewritten long ago (like, before we started having an armel port)... so I wouldn't worry about disabling it [15:20] erk, yes [15:22] slangasek, well, the prob is that the script doesnt take any arm bootloaders into account, so it will always show a debconf error that you have to rewrite your configs by hand [15:22] we need to get at least this bit quiet [15:23] (and i really dont like the idea that it touches your grub config if you manually forced it to i.e. /dev/sdx1) [15:23] it was sane for transitional purposes, but as slangasek says it sounds obsolete now [15:24] ogra_: really? The code looks like it's not supposed to trigger on first install [15:24] is_fresh_installation() [15:25] well, i just had that message on both, precise (using --force-overwrite to overcome the perf conflict) as well as on quantal [15:27] oh, sorry, the definition of is_fresh_installation() keys on whether there are linux-image-* packages installed [15:28] rather than on whether this is the first installation of linux-base [15:28] phooey [15:28] and if fstab exists :) [15:29] sigh, my perl is so rusty ... thats what years of ubuntu work do to you [15:30] funnily i indeed have an fstab file as well as a linux-image-* package installed on both test systems i tested on [15:30] which somewhat indicates it doesnt even work [15:31] well, no, the check is to not migrate if you *don't* have an fstab (or don't have linux-image-* packages) [15:31] so that makes sense.. I just made a bad assumption based on the function name :) [15:32] oh [15:32] yeah, and i read it vbackwards :) [15:33] ogra_: yeah, I would just disable this then [15:33] you mean the whole postinst ? [15:33] yeah [15:33] it's definitely not worth porting it to other bootloaders [15:34] right, i'll just remove it from the package completely then === lamont` is now known as lamont === ogasawara_ is now known as ogasawara === ScottK2 is now known as ScottK [16:13] stokachu: hi. I'm currently looking at your isc-dhcp SRU. I merged the pid bug with the one I used for the other SRUs (bug 985417) and am now looking at bug 588635 [16:13] Launchpad bug 985417 in isc-dhcp (Ubuntu Oneiric) "dhcpd cannot write /var/run/dhcpd.pid" [Undecided,New] https://launchpad.net/bugs/985417 [16:13] Launchpad bug 588635 in isc-dhcp (Ubuntu Oneiric) "apparmor profile for sbin.dhclient3 not compatible with wicd" [Undecided,New] https://launchpad.net/bugs/588635 [16:14] stokachu: for that second bug, can you update the bug report with the required SRU tags? (rational, test case, regression potential)? [16:14] stokachu: https://wiki.ubuntu.com/StableReleaseUpdates#Procedure [16:16] stokachu: I have the SRU ready for upload here, so as soon as the bug is updated I'll upload it [16:43] one question: does quantal have the same toolchain like unstable right now? e.g. does it have gcc 4.7 default? [16:46] stgraber: yea ill get that updated shortly for you [16:49] ari-tczew: could you please check your pm when you get a moment [16:51] I need to upload an SRU change for 12.04 that has the same package version as the latest package in quantal -- should I add ~precise to the SRU package version, er sumpin'? [16:53] bregma: I would probably recommend appending ~ubuntu12.04.1 [16:53] or something like that [16:54] (assuming that the delta from current precise to this meets the SRU guidelines) === deryck is now known as deryck[lunch] [17:36] jcastro: yo, I don't see a tenant or account ID on the page you linked from HPCloud. Just "Access Key ID" [17:36] do I need to do anything else? [17:37] refresh the wiki page [17:37] I just added it [17:37] "Note that the Tenet ID doesn't show up on that page until you've clicked on "Activate Now: for one of the availability zones in the Dashboard." [17:37] hah, you did /just/ add it indeed. [17:37] thanks. [17:39] * Laney wonders how to make that choice [17:41] Laney: anything you can add from the "I started with nothing on hp cloud" to the instructions would be appreciated [17:41] a bunch of us were already on it so it's difficult to figure out what starting from zero is like [17:42] sure [17:42] the account ID is on a different page, so I'll add a link to that [17:42] <3 [17:43] stgraber: re: SRU 588635 is done [17:43] * xnox can't find my USD credit card [17:43] for hp cloud thing =) [17:45] I just used my UK one [17:46] stokachu: thanks [17:46] stgraber: anytime [17:46] stgraber: now im off to wrestle autofs :D [17:46] have fun ;) [17:51] stokachu: i have been merging autofs today, what's up? [17:51] or is it for stable release? =) [17:51] xnox: hey, im working an issue where automount is hanging on reloading 11k mounts [17:51] :X [17:52] yea its for lucid atm [17:52] ... enjoy =) [17:52] haha taking the hands off approach wish i was you :D [17:56] 11k mounts?! [18:01] penguin42: thats just a rough estimate i haven't actually looked at the mount table yet [18:03] * penguin42 has plenty of sympathy for the automounter in that case [18:04] * stokachu agrees === deryck[lunch] is now known as deryck [18:54] doko: pong [19:01] * xnox Ladies and Gentlemen, Minions and Overlords 2 hours left until FeatureDefinitionFreeze [19:11] StevenK: I just did my first package removal. Feels good. [19:11] nice [19:15] ScottK: I got a feeling =) that tonight gonna be a real good night =) [19:15] ;-) [19:16] ScottK: can you remove stale binaries of bitcoin on powerpcc only? [19:17] bdmurray: hey, can you remove one of the requested reviews on https://code.launchpad.net/~brian-murray/lptools/grab-descriptions/+merge/106469 ? then i can review it and get it landed (i added lptools to a tarmac instance to land the branches) [19:17] although it will be dealt as part of powerpc port cruft [19:20] dobey: either one of them? [19:21] bdmurray: yeah [19:23] dobey: there is only a possibility to reassign not remove. should I resubmit then? [19:23] hmm, weird. nah === jalcine is now known as Jacky [19:38] xnox: I can. Please file a bug explaining what you want removed and why and subscribe (not assign) ubuntu-archive. [20:08] Ok. [20:08] ScottK: ok, thanks. [20:35] why does sbuild fail with "E: Can't determine architecture of chroot:" these days, on precise? [20:35] only thing i've chagned is the kernel (3.4 from quantal) === salem_ is now known as _salem [20:37] pae kernel? [20:37] amd64 [20:37] hmm. no clue. pass. [20:38] and when you boot the old kernel it works again? [20:39] i'll try, but it's too unstable to use [20:41] geser: exactly [20:43] if my perl foo is good enough this error happens if "dpkg --print-architecture" fails [20:43] I'll try with a backport kernel, this one was built on quantal [20:44] dpkg --print-architecture worked fine though [20:49] stgraber: could i ask a favor of you to bump this build: https://launchpad.net/~barry/+archive/python/+build/3539392 [20:49] barry: done [20:50] stgraber: thanks! [21:22] all right. my struct has (more than) two fields, 'char *name' and 'char *error_string'. when i set 'c->error_string = NULL;', the previously correctly set c->name becomes "". Does this ring any bells with anyone? [21:22] I can't reproduce it with a minimal testcases [21:25] you sure that setting error_string destroys your name? [21:25] i print it out right before, and right after [21:26] well, clearly i'm corrupting my stack somewhere [21:26] i'll look tomorrow, need to clear my head. thanks [21:40] Is there a set group of Ubuntu Ruby folks like there is for Python or do we just inherit from Debian? [21:42] what do you mean by "set group"? I'm not aware there's a set group of Ubuntu Python folks either [21:43] slangasek: What I mean is when I think of python packages I tend to think of folks like scottk and doko [21:43] even though it's informal [21:45] ah, well, the ruby packages seem to be in sync with Debian at the moment, so I don't know [21:49] No. There isn't. [22:02] stgraber: one more favor ;) https://launchpad.net/~barry/+archive/python/+build/3539477 [22:03] barry: done [22:03] stgraber: thx! === jono is now known as Guest23141 [23:42] cjwatson, how does d-i locate the installation cdrom? in particular when it's actually a usb flash drive? it works fine for me on real hardware, but when I boot a qemu vm from my flash stick, d-i says it can't find the cdrom... the desktop image works fine though, it's just a problem with d-i [23:56] roaksoax: you seem to be forgetting to pass -v when creating the .changes files