2012/09/13/#ubuntu-devel.txt

[00:00] <cjwatson> Varies from country to country I believe
[00:00]  * xnox was reading WP about all of that as well
[00:01]  * xnox adds a TODO item to create a wiki page for cjwatson, if scott has one with out of date picture so can cjwatson ;-) 
[00:04] <cjwatson> I think you should beware WP:COI.  I wouldn't encourage anyone I worked with closely to write an article about me
[00:04] <SpamapS> cjwatson: the packages cache generation is really really fast, don't think it has much fruit for optimization
[00:04] <slangasek> since you wouldn't encourage it, there's no conflict of interest ;-)
[00:04] <SpamapS> user	0m0.056s
[00:05] <SpamapS> cjwatson: 163926 packages in 0.056s .. ;)
[00:05] <cjwatson> And of course WP:BIO
[00:05] <cjwatson> SpamapS: Huh, so why do I notice occasional ridiculous slowdown
[00:06] <cjwatson> Maybe the cache directory layout (such as it isn't) is poorly designed
[00:06] <SpamapS> cjwatson: the source cache does a sort.. testing that now
[00:07] <cjwatson> Mm, no custom sort functions though
[00:07] <cjwatson> s/functions/comparators/
[00:08] <SpamapS> yeah its blazing fast too
[00:10] <cjwatson> and yet
[00:10] <cjwatson> $ time HOME=/home/ubuntu-archive/www-data madison-lite --nocache man-db
[00:10] <cjwatson> real    0m52.245s
[00:10] <cjwatson> that's fairly ridiculous
[00:10] <SpamapS> hm
[00:10] <cjwatson> (also no output, there's evidently a correctness problem as well :-/)
[00:12] <cjwatson> well, I should be doing a couple of other things before going to bed.  let me know if you come up with anything
[00:13] <SpamapS> cjwatson: will do :)
=== cpg is now known as cpg|away
[00:23] <cjwatson> lifeless: I've reproduced your GRUB regression in a horrendously hacked-up VM, so I should be able to sort it out from here
[00:24] <lifeless> cjwatson: \o/
[00:24] <cjwatson> lifeless: Ended up simulating all the dmraid stuff with manual dmsetup commands, which isn't good enough to allow dmraid -ay to work but it is enough to brainwash GRUB
[00:24] <lifeless> cjwatson: let me know when you think its good and I'll get you a positive confirmation :)
[00:24] <cjwatson> Righto
[00:39] <cjwatson> lifeless: All right, fix (hopefully) uploaded - I'll check back tomorrow morning and let you know if it's built
[00:39] <lifeless> okeydoky
[00:41] <cjwatson> Fix was dropping a bit of the patch :)
=== spm is now known as stevemci
=== stevemci is now known as spm
=== spm is now known as stevemci
=== stevemci is now known as spm
=== cpg|away is now known as cpg
[06:13] <queency> hello all i'm using ubuntu9/04 can someone direct me to install power save to my machine
=== cpg is now known as cpg|away
=== cpg|away is now known as cpg
[06:17] <tjaalton> queency: that release was EOL'd almost two years ago..
[06:18] <queency> tjaalton: yes so ? why everybody keeps bugging me to upgrade ? everything works fine!
[06:19] <tjaalton> queency: don't expect to get any sort of support for it
[06:19] <tjaalton> like, security updates
[06:19] <tjaalton> or guidance
[06:21] <queency> tjaalton: security i'm behind a firewall i.e i don't afraid , guidance whats wrong with small tips here and then ?
[06:22] <tjaalton> queency: power management is built in. if your machine doesn't have it, it could be better in a more recent release. so my free tip is: please upgrade
[06:22] <queency> tjaalton: btw i don't run services in my private lab ubuntu 9.04 and i'm monitor my net steadily : i just don't familier with all the task yet !
[06:22] <queency> tjaalton: tnx but no tnx
[06:23] <queency> tjaalton: is the power save manipulated by the kernel ?
[06:24] <tjaalton> queency: i've nothing to add
[06:25] <queency> tjaalton:ok tnx for your time , i'm sorry but i like my system to be intact and upgrading however solve things but brake others.
[06:25] <tjaalton> use an lts release like 10.04/12.04
[06:26] <queency> tjaalton: even my monitor driver was patched by me manually from 2 years ago : upgrading will screw it up !
[06:26] <tjaalton> sure about it?
[06:28] <queency> tjaalton: don't want to check if i right
[06:29] <tjaalton> there are livecd's for that
[06:33] <queency> tjaalton: i don't say live cd won't work i just say i don't know what happened with upgrading my old system ; not installing a new one
[06:37] <RAOF> Is it still possible to upgrade from 9.04?
[06:39] <tjaalton> dunno
[06:39] <ajmitch> perhaps with dist-upgrade & adding old-releases.u.c into sources.list
[06:39] <tjaalton> yeah
[06:40] <RAOF> Anyone feel like sponsoring a colord upload to sid?
[06:43] <didrocks> slangasek: tjaalton: hey, you have a WI on https://blueprints.launchpad.net/ubuntu/+spec/desktop-q-xorg-lts-updates to review the rename script, do you think you will get a chance to deal with it?
[06:44] <tjaalton> didrocks: sure
[06:44] <tjaalton> now that it seems to be working fine, might just mark it DONE
[06:44] <RAOF> pitti: feel like sponsoring a colord upload to sid?
[06:45] <didrocks> tjaalton: great! :)
[06:45] <didrocks> thank
[06:45] <didrocks> thanks*
[06:46] <dholbach> good morning
[06:53] <pitti> Good morning
[06:54] <pitti> SpamapS: I guess dpm uploaded them; we do that in regular intervals, and yes, they should just be waved through (there's a whole testing process behind that)
[06:55] <pitti> RAOF: yep, can do
[06:55] <RAOF> pitti: Ta. It's in collab-maint git; you're looking for the debian/0.1.21-2 tag.
=== JanC is now known as Guest57134
=== christoffer is now known as Guest92863
[07:25] <darkxst> anyone know how to install a mainline kernel into a livecd chroot?
[07:26] <pitti> SpamapS, infinity: hm, I've got mail from dpm that he didn't upload these either
[07:27] <pitti> infinity: do you remember, were these proper uploads or copies from a PPA?
[07:33] <jibel> pitti, I enabled quantal-proposed for autopkgtest. There are now 2 sets of tests: one with packages in the release pocket, the other with proposed
[07:33] <jibel> tests in proposed are triggered when a newer version of a package is uploaded to proposed
[07:33] <pitti> jibel: niiice!
[07:33] <jibel> or a newer version of a dependency of a package in release is uploaded to proposed
[07:33] <jibel> or a newer version of a dependency is uploaded to release and the package is in proposed
[07:34] <jibel> pitti, sound good ? or I missed other rules ?
[07:34] <pitti> jibel: does that include build dependencies?
[07:34] <pitti> (this also applies to -release)
[07:35] <pitti> jibel: I don't see those yet, is that because there just are no packages in -proposed at the moment?
[07:35] <jibel> pitti, no, binary deps, should I include build deps ?
[07:35] <jibel> pitti, right, packages autodiscovered have been added to the private instance and will be published on first run
[07:36] <pitti> jibel: hm, do we have tests with "build-needed"? for those I think we should
[07:36] <infinity> pitti: They were direct uploads.
[07:36] <pitti> infinity: thanks; strange then, it seems nobody did those then
[07:36] <infinity> pitti: That's a bit disconcerting.  Surely, SOMEONE must have. :P
[07:37] <pitti> infinity: theres' a cronjob on macquarie which auto-uploads to the PPA for stable releases, and to the archive for the devel release; that hasn't changed in years, though
[07:37] <pitti> I'll take a closer look
[07:44] <jibel> pitti, there is no test with a restriction "build-needed" for the moment, but that's a good point. I'll add support to run on build deps change if packages have such restriction.
=== tkamppeter_ is now known as tkamppeter
=== henrix_ is now known as henrix
=== dholbach_ is now known as dholbach
=== mcclurmc_away is now known as mcclurmc
=== rose7676 is now known as roseoff
[09:59] <roseoff> hello !
[09:59] <roseoff> sudo apt-get install fglrx fglrx-amdcccle
[10:00] <roseoff> The following packages have unmet dependencies:
[10:00] <roseoff>  fglrx : Depends: xorg-video-abi-11 but it is not installable
[10:00] <roseoff> E: Unable to correct problems, you have held broken packages.
[10:00] <roseoff> ?
[10:00] <roseoff> HELP !
[10:00] <roseoff> ubuntu 12.10
[10:00] <mlankhorst> yes
[10:01] <roseoff> mlankhorst,
=== cpg is now known as cpg|away
[10:09] <pitti> roseoff: not much you can do -- just wait until a working driver is available, and use the free one until then
[10:09] <mlankhorst> pitti: I was hoping he would try to google the error :(
[10:40] <pitti> jibel: thanks for the aptdaemon autopkgtest-ification!
[10:40] <pitti> jibel: did you already check this in a minimal VM, or want me to? (I have my test VM here)
[10:41] <jibel> pitti, yw, I used it to write a small doc that explains how to add dep8 support to a package
[10:42] <jibel> pitti, I tested it in a minimal VM from the source tree
[10:43] <pitti> jibel: cool, thanks; I'll upload this now
[10:43] <jibel> pitti, 68 tests passed, 4 skipped, there are dbus message to investigate
=== doko_ is now known as doko
[10:48] <Laney> http://paste.debian.net/189671/ how do I make apt want to remove ubuntu-default-settings and install ubuntu-settings instead of holding back ubuntu-desktop?
[10:48] <Laney> u-s B/R (versioned appropriately) on u-d-s
[10:49] <tumbleweed> you need a transitional package
[10:49] <Laney> that is exactly what I want to avoid
[10:49] <tumbleweed> you can't
[10:49] <tumbleweed> http://wiki.debian.org/Renaming_a_Package btw
[10:49] <Laney> that package name is already being used as a virtual package
[10:49] <Laney> I know
[10:49] <xnox> Laney: breaks/replaces/provides + higher version number?!
[10:50] <xnox> Laney: it's the case of "multiple mta's, yet only one mta at the same time"
[10:50] <tumbleweed> provides are versionless (you should remember that from NM)
[10:50] <cjwatson> Well, no, it shouldn't be like the MTA case
[10:50] <cjwatson> This might improve once ubuntu-default-settings is no longer in the archive
[10:51] <Laney> you think?
[10:51] <cjwatson> They're exactly balanced in score - I think one no longer being available would tip the balance
[10:52] <cjwatson> I think
[10:54] <cjwatson> It should lose the one-point "installed package that isn't obsolete" bonus
[10:54] <Laney> yeah, that seems to be correct (hacked it out of the list manually to test)
[10:54] <Laney> who know? (you)
[10:55] <tumbleweed> Laney: you could mention this on that wiki page for future people to know
[10:55] <cjwatson> apt-pkg/algorithms.cc if you want to stare aimlessly
[11:31] <BenC> Did someone do a bunch of give-backs or is powerpc really doing that awesome this morning on ftbfs page?
[11:32] <xnox> BenC: i heard doko & infinity were planning give-backs before archive rebuild =/ or something like that
[11:33] <cjwatson> I noticed the output of a load of give-backs in my inbox
[11:34] <cjwatson> The hint is probably that powerpc is 33 hours behind :-/
[11:34] <doko> yes, builds were given back yesterday evening
[11:35] <ogra_> did anyone ever try to use qemu-ppc-static for this ?
[11:35] <ogra_> i bet its faster :)
=== dpb_ is now known as Guest14449
=== broder_ is now known as broder
=== Pici` is now known as Pici
=== zumbi is now known as Guest56315
=== MacSlow is now known as MacSlow|lunch
[12:00] <cjwatson> lifeless: https://launchpad.net/~cjwatson/+archive/grub/+packages has an update ready for you to test now
=== _salem is now known as salem_
=== DktrKranz is now known as Guest14632
=== Phantomas1 is now known as Phantomas
=== Quintasan_ is now known as Quintasan
=== diplo_ is now known as diplo
=== micahg_ is now known as micahg
[12:56] <hallyn> smoser: well, could be.  like i say i thought i'd fixed those.  but more likely i missed it, and it's a regression by the templates adding the second space
[12:56] <smoser> ah.
[12:56] <smoser> well glad it will work. thank you.
=== s1aden is now known as sladen
=== cpgz is now known as cpg
=== cpg is now known as Guest70811
=== MacSlow|lunch is now known as MacSlow
[13:11] <hallyn> smoser: btw, i went ahead and changed --new to --name in lxc-clone, but it occurred to me last night that --new is what i really intended :)
[13:11] <hallyn> and nowhere is '--new' actually spelled out, i don't think
[13:11] <smoser> --name is better.
[13:11] <smoser> its consistent with everything else.
[13:12] <smoser> imo
[13:12] <smoser> but either one working would have been "good enough"
=== Phantomas1 is now known as Phantomas
=== hrww is now known as hrw
=== cking_ is now known as cking
=== henrix` is now known as henrix_
=== henrix_ is now known as henrix
=== ScottK2 is now known as ScottK
=== ikonia_ is now known as ikonia
[13:45] <xclaesse> I'm testing Quantal's installer in a VM. I see encrypted LVM option exists now, but only for automatic partitioning?
[13:45] <xclaesse> is it possible to manually define the size of the partitions inside the LVM ?
[13:46] <xnox> xclaesse: not yet.
[13:46] <xnox> xclaesse: you can do it post-install.
[13:46] <xnox> xclaesse: by resizing your root partition & resizing logical volume.
[13:47] <xclaesse> xnox, not yet means not for Quantal or will it still make it in time for Quantal ?
[13:47] <xnox> xclaesse: in a way it is a limitation of the debian-installer's partitioning software partman, which ultimately requires something to take up all remaining space
[13:47] <xnox> xclaesse: I am coding it, there is FeatureFreezeException for it, and it is targeted for Quantal.... so it should land soon.
[13:48] <xclaesse> nice !
[13:48] <xnox> xclaesse: taking the rest of the space for automatic partitioning that is. No such requirements for manual partitioning (kind of)
[13:48] <xclaesse> xnox, while I'm at it: will it be possible to re-install without ripping all LVM volumes?
=== iainfarrell1 is now known as iainfarrell
[13:48] <xclaesse> IIRC that was not possible with alternat installer
[13:49] <xclaesse> xnox, hmmm, automatic won't make /home in a different logical volume, will it?
[13:49] <xnox> xclaesse: it can, but we do not offer that.
[13:50] <xnox> xclaesse: it's easy to fix that up / migrate to separate home post-install.
[13:50] <xclaesse> yeah, so manually creating the volumes would still be nicer IMO
[13:50] <xnox> i know
[13:50] <xclaesse> I totally agree that's not something you want for "normal" users, but having the choice somewhere would still be good :)
[13:51] <xclaesse> xnox, so manual LVM is not planned for quantal?
[13:52] <xnox> xclaesse: i told you, i am working on it.
[13:52] <xnox> xclaesse: automatic will stay as it is, but manual crypt/lvm will land in one form or another
[13:53] <xclaesse> ah ok, miss understood
[13:53] <xclaesse> good :)
[14:00] <smoser> cjwatson where would i find TFM to read about what kernel parameters d-i reads?
=== barry` is now known as barry
=== wgrant_ is now known as wgrant
[14:12] <cjwatson> smoser: https://help.ubuntu.com/12.04/installation-guide/
=== JanC_ is now known as JanC
[14:13] <ScottK> pitti: In case no one else told you, you got mentioned here: http://www.linuxuser.co.uk/interviews/whats-going-on-with-gnome/
[14:15] <smoser> cjwatson, so can you confirm for me that 'release=<something>' is ignored?
[14:16] <smoser> or rather is not expected to do anything ? i have something that is writing it, but it didn't seem like it should have to, and i dont see it at (https://help.ubuntu.com/12.04/installation-guide/i386/boot-parms.html)
[14:24] <cjwatson> smoser: I've never heard of that
[14:25] <smoser> cjwatson, thanks. it was probably copied along.
[14:31] <slangasek> didrocks: hmm, I think the script in question has by now iterated to the point that my review would no longer add anything
=== fginther` is now known as fginther
[14:32] <didrocks> slangasek: ok, put it in DONE and drop it please :)
[14:32]  * didrocks was in blueprint cleanage day
=== Guest14632 is now known as DktrKranz
[14:36] <SpamapS> cjwatson: I think you may be right that the caching needs optimization in madison-lite
[14:36] <SpamapS> %Time ExclSec CumulS #Calls sec/call Csec/c  Name
[14:36] <SpamapS>  83.2   0.990  0.990     13   0.0762 0.0761  main::cache_list_file
[14:36] <SpamapS> thats without a cache
[14:36] <SpamapS> and with a cache..
[14:37] <SpamapS>  80.3   0.160  0.160     13   0.0123 0.0123  main::search_cache
[14:38] <SpamapS> tho.. main::cache_list_file is basically "the whole thing" .. so need to break it up to get more from dprof
[14:43] <slangasek> didrocks: ok, done :)
[14:44] <didrocks> thanks slangasek :)
=== pgraner` is now known as pgraner
[15:02] <didrocks> ogra_: did the XDG_RUNTIME_DIR landed?
[15:03] <slangasek> didrocks: no... working on it still
[15:03] <didrocks> slangasek: ok, ETA is beta2 freeze I guess?
[15:03] <slangasek> didrocks: i.e., I'm working on it, not ogra
[15:03] <slangasek> didrocks: yes
[15:03] <didrocks> slangasek: ah ok, I saw that in the release meeting report, hence the question :)
[15:03] <didrocks> slangasek: thanks ;)
[15:03] <slangasek> yeah... :)
[15:05] <tjaalton> let's try here: anyone here with an intel i915/i945 machine willing to bisect mesa? I'll build the packages
[15:06] <pitti> ScottK: argh, with the one thing I don't really want to do (CK :) )
[15:07] <pitti> ScottK: thanks for the pointer
[15:07] <ScottK> So now you know you just need to line up some BSD hackers to help you out.
=== Quintasan_ is now known as Quintasan
=== dendro-afk is now known as dendrobates
[15:26] <smoser> cjwatson, around ? i'm looking at changes i believe were yours to /etc/init/iscsi-network-interface.conf under https://bugs.launchpad.net/ubuntu/+source/partman-iscsi/+bug/457767
[15:26] <ubottu> Launchpad bug 457767 in partman-iscsi (Ubuntu Karmic) "karmic: iSCSI root: boot hangs on starting iscsid" [High,Fix released]
[15:26] <smoser> quantal has regressed and is no longer writing /run/initramfs/open-iscsi.interface . but my iscsi root seemed to not suffer any negative from that
[15:27] <smoser> ie, i would have thought that that would cause my network interface to bounce and probably generally screw things up.
[15:28] <xnox> stgraber: ^^^^
[15:30] <cjwatson> smoser: honestly, it's been so long since I touched that that I don't think I'll be of any more help than reading the code from scratch
[15:30] <cjwatson> you might have got lucky I guess; regardless, we should be making sure that the interface we're doing iSCSI root over is never torn down ...
[15:31] <smoser> cjwatson, ok. i'll get that code added back in.
[15:32] <stgraber> smoser: did you see my comment in #ubuntu-server yesterday about just checking for /run/net-* from the upstart job rather than having another piece of code in initramfs?
[15:32] <smoser> stgraber, i did not.
[15:33] <smoser> and i considered that
[15:33] <smoser> but at the moment there could be 3 /run/net-* entries
[15:33] <smoser> (or 'N' of them)
[15:33] <smoser> and you wouldnt know which exact interface was the root
[15:34] <smoser> the initramfs [potentially] has that knowledge and was storing it for later consumption
[15:36] <Daviey> smoser: erm.. we add ipappend 2, which adds BOOTIF=01-mac-address of booting device to kernel cmd line
[15:36] <Daviey> so we know the boot device
[15:37] <smoser> Daviey, yeah, but that isn't known in all cases.
[15:37] <Daviey> smoser: sure?
[15:37] <smoser> and BOOTIF= is actually not necessarily the device that provides the link to the iscsi root
[15:38] <smoser> ip=:::::eth1 BOOTIF=01-mac-adddress
[15:38] <smoser> would bring up eth1 as dhcp
[15:38] <slangasek> so there's certainly enough information in the system that you can walk it to discover the correct device post-initramfs
[15:38] <slangasek> but surely it's simpler to just have the initramfs bit record it
[15:38] <smoser> right.
[15:38] <smoser> that is what i was about to argue.
[15:39] <smoser> just opened bug 1050480
[15:39] <ubottu> Launchpad bug 1050480 in open-iscsi (Ubuntu) "iscsi-network-interface.conf broken after upstream merge" [Undecided,New] https://launchpad.net/bugs/1050480
[15:39] <smoser> anyone else want to think about resolv.conf in this case?
[15:40] <smoser> i'm thinking i need to add code to iscsi-network-interface.conf that reads /run/net-*.conf and adds the dns servers
[15:40] <smoser> at lesat for resolvconf.
[15:40] <slangasek> smoser: I thought that what we discussed on this previously was sound
[15:40] <smoser> but in non-resolvconf world, that would have been updated by dhclient and its not going to get done.
[15:40] <slangasek> there is no non-resolvconf world
=== lan3y is now known as Laney
[15:41] <slangasek> it's unsupported, EOD
[15:41] <smoser> is that acceptable ?
[15:41] <slangasek> yes, ubuntu-minimal depends on resolvconf
[15:41] <smoser> ok then. much easier.
[15:41] <smoser> thanks.
[15:44] <slangasek> smoser: to be clear, a driving factor for putting resolvconf into ubuntu-minimal was to no longer have to worry about the 50 different ways to manage dns ;)
[15:44] <slangasek> so yeah, anybody who goes out of their way to disable resolvconf is on their ownw
[15:44] <smoser> thank you
[16:07] <smoser> someone tell me what i'm missing
[16:07] <smoser> http://paste.ubuntu.com/1202917/
[16:07] <smoser> lines 23-25. why is that path taken if /run/initramfs/open-iscsi.interface did not exist
[16:08] <smoser> it seems to me that will call /etc/network/if-up.d/upstart with IFACE=''
=== deryck is now known as deryck[lunch]
[16:14] <stgraber> smoser: looks like that part should be under the other diff
[16:14] <stgraber> s/diff/if/
[16:14] <smoser> thats what i thought
[16:15] <smoser> thanks.
[16:20] <slangasek> smoser: looks like a bug for this to not be inside the previous if block
[16:20] <slangasek> right, what stgraber said :)
[16:26] <smoser> hm..
[16:26] <smoser> does that job have access to INTERFACE ?
[16:26] <smoser> that is provided to the 'network-interface' job that it is 'start on' and 'stop on'
[16:26] <smoser> but will it have access to that?
[16:35] <smoser> shoot. now i'm reading that same pastebin (http://paste.ubuntu.com/1202917/)
[16:36] <smoser> wont the post-stop run the first time a network-interface is taken down?
[16:36] <smoser> i guess i can't think of a failure path there, but it'd be better if it only removed that on down of the specific $iface
[16:44] <mterry> dpm, do all Ubuntu translations specify utf8 encoding in the po file?
[16:46] <dpm> mterry, I believe yes
[16:47] <mterry> dpm, cool
[16:58] <doko> jamespage, do we have the no-jars-in-package fun in quantal too?
[16:59] <jamespage> doko, http://paste.ubuntu.com/1203022/
[16:59] <jamespage> yep
[16:59] <jamespage> more potential rather than actual at this point in time just like bytecode compat
[17:00] <doko> can we revert this change before a test rebuild?
=== LeonhardEuler is now known as KarlMarx
[17:00] <jamespage> yep
[17:02] <xnox> cjwatson: is there a good test to see if the livecd booted in efi or bios mode?
[17:02] <smoser> slangasek, or stgraber do you have thoughts on the above?
[17:03] <smoser> it really seems like the current code basically telling ifupdown that the iscsi interface is down as soon as *any* network interface goes down.
[17:03] <cjwatson> xnox: [ -d /sys/firmware/efi ]
[17:04] <xnox> cjwatson: thanks
[17:05] <xnox> cjwatson: any tips on migrating existing system to efi ? i do have separate /boot do I need to convert it to "efi" partition?
[17:05] <xnox> aka split it into two?
[17:05] <dobey> Is there any way to tell lintian to ignore an error/warning from a binary package, via some extra field in the control perhaps?
[17:06] <cjwatson> xnox: /boot and /efi are different categories of things
[17:06] <cjwatson> You must have an EFI System Partition
[17:06] <xnox> ok. will make one.
[17:07] <cjwatson> I don't have documentation on doing a conversion to hand
[17:07] <infinity>        For binary packages, Lintian looks for overrides in a file named
[17:07] <infinity>        usr/share/lintian/overrides/<package> inside the binary package, where <package> is the
[17:07] <infinity>        name of the binary package.
[17:07] <infinity> dobey: ^
[17:07] <xnox> cjwatson: ok.
[17:07] <dobey> infinity: ah, interesting. what doc is that in?
=== mcclurmc is now known as mcclurmc_away
[17:10] <infinity> dobey: The lintian manpage.
[17:10] <dobey> ah it's in the man page. but the syntax of course isn't. luckily first link i clicked on from duckduckgo seems to have it :)
[17:10] <dobey> infinity: thanks
[17:11] <fbond> mdz: Hope this is an okay place to contact you. Saw https://launchpad.net/cfontz but there is no code there. Any place I can find it?
=== Trewas666 is now known as Trewas
[17:20] <SpamapS> cjwatson: playing a bit more w/ madison-lite here and there... its hard to see how slow it is until you actually put all of the archive files in... each Packages.gz processes "fast".. but there are .. well.. quite a few. ;)
=== deryck[lunch] is now known as deryck
[17:45] <bdmurray> cjwatson: what networking drivers would be on the mini cd?  bug 1047092
[17:45] <ubottu> Launchpad bug 1047092 in linux (Ubuntu) "Mini.iso doesn't load wireless drivers" [Medium,Confirmed] https://launchpad.net/bugs/1047092
=== Guest70811 is now known as cpg
=== henrix is now known as henrix_
[18:17] <xnox> cjwatson: grub2 from ppa works fine in bios mode with lvm/luks. my laptop refuses to boot in uefi mode =(
[18:23] <infinity> cjwatson: grub2 no workie on powerpc.
[18:25] <SpamapS> cjwatson: just going through the cache files is still very expensive unfortunately
[18:25] <SpamapS> cjwatson: mostly because there are so many to go through
[18:28] <SpamapS>   1295992   0.57777   7.09000   525: my ($key, $value, $is_all) = split;
[18:29] <SpamapS> cjwatson: as you suspected, sqlite's indexes will probably blow the doors off that. )
[18:30] <lifeless> SpamapS: ?
[18:30] <SpamapS> lifeless: I'm taking a few moments to speed up rmadison
[18:30] <lifeless> SpamapS: oh cool
[18:31] <SpamapS> which I use quite heavily during SRU processing, and which is incredibly slow
[18:31] <lifeless> SpamapS: move it into LP ?
[18:31] <SpamapS> lifeless: thats what I thought actually :)
[18:31] <SpamapS> lifeless: but cjwatson is wary of the performance of versions+packages+sources in the API
[18:31] <lifeless> you'd want a new API entry point
[18:32] <lifeless> no freaking way you want to use the object programming model, it would die from round trips.
[18:32] <lifeless> -but-
[18:32] <SpamapS> Its asking "this random string, what packages/sources and versions and architectures are the highest published?"
[18:32] <lifeless> LP should be able to answer it in about 5 queries, bundle it into json and ship you a single result.
[18:32] <lifeless> where string can be a source package name or binary package name, right ?
[18:33] <SpamapS> lifeless: the other reason is, sometimes its nice to have a simple thing parsing the actual Packages/Sources lists
[18:33] <SpamapS> lifeless: right
[18:34]  * xnox was expecting to fetch binary publishing history & source publishing history where status is published for active series + some string formatting
[18:34] <xnox> to make an lpmadison
[18:35] <lifeless> so yeah, internally thats 1) identify the archives, 2) the series 3) find any matching binary package names, 4) find any published binaries for those names and get the source package names - add to the inputs, 5) find any source package names 6) find the highest published version per series. Done.
[18:35] <infinity> lifeless: Even if it all could be done quickly in LP, the results are still "wrong", in that people aren't interested in LP's publishing state, but in what's actually on-disk.
[18:35] <lifeless> That should be well under 100ms
[18:35] <lifeless> infinity: LP knows whats on-disk.
[18:35] <infinity> It sure doesn't.
[18:35] <infinity> It's about 30 minutes early on that.
[18:36] <lifeless> depending on which mirror they use, it can be hours early.
[18:36] <infinity> (Which is fine, but not helpful for madison use)
[18:36] <infinity> Also, your above is missing the source->binary mapping, including out-of-date binaries.
[18:36] <infinity> It gets more complicated.
[18:37] <xnox> but for SRU you want what's lp view of things is, even if that's what's only going to happen in the future 30 minutes.....
[18:37] <lifeless> ok
[18:38] <infinity> xnox: For SRU processing, you don't even want published, you also want accepted and unapproved, which isn't really what madison's for.
[18:38] <xnox> true
[18:38] <infinity> (In other words, you want to know everything that's about to happen, not everything that has happened)
[18:40] <SpamapS> true
[18:41] <SpamapS> for SRU I would like LP's view
[18:43] <infinity> Extending rmadison to have a "show me what LP thinks is published" and a "show me what's in accepted/new/unapproved too" option would be neat, but speeding up the current "show me what's on-disk" view is a nice start.
=== dendrobates is now known as dendro-afk
=== haakonn is now known as Guest67148
[19:53] <bryceh> @pilot in
=== udevbot changed the topic of #ubuntu-devel to: Quantal Quetzal development | Archive: Open | Dev' of Ubuntu (not support or app devel) | build failures -> http://qa.ubuntuwire.com/ftbfs/  | #ubuntu for support and general discussion for hardy -> precise | #ubuntu-app-devel for app development on Ubuntu http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: bryceh
[19:54] <mdz> fbond: I would be happy to send you the code but you left :-(
=== dendro-afk is now known as dendrobates
[20:02] <tedg> slangasek, Which Chinese is "Simplified"?  I've got zh_CN, zh_HK and zh_TW.  Though, none of them show the bug :-9
[20:02] <tedg> :-(
[20:05] <Daviey> bryceh: err, 'weston'.. Odd choice of sponsorships.
[20:06] <stgraber> tedg: zh_CN
[20:07] <Daviey> bryceh: The debian/changelog suggests it's a really useless upload.. but infact misses out that it drops a patch?
[20:07] <Daviey> bryceh: http://pb.daviey.com/FNu4/
[20:07] <tedg> stgraber, Thanks!
[20:08] <bryceh> Daviey, hmm
[20:11] <bryceh> no reference to the patch in the changelog
[20:11] <bryceh> nor does the patch exist in the git tree
[20:14] <Daviey> bryceh: and xgrep?  These seem like uploads not worth the buildd time.
[20:15] <Daviey> why didn't dholbach upload that himself?
[20:15] <bryceh> Daviey, dude, just trying to work through the pilot queue, why so critical?  topic says the archive's open
[20:16] <Daviey> bryceh: Sure, ok.
[20:16] <Daviey> Sorry if i seemed to pick, wasn't meaning to.
[20:20] <SpamapS> doko: any word on bug 1048710 ? New upload in progress?
[20:20] <ubottu> Launchpad bug 1048710 in python3.2 (Ubuntu) "Regression in argparse for Python 2.7, 3.2 and 3.3" [Critical,Confirmed] https://launchpad.net/bugs/1048710
[20:21] <bryceh> tjaalton, looks like the git.diff came from your 0.95.0-0ubuntu1 upload; was that just build tree cruft or is that supposed to be there?
[20:21] <doko> SpamapS, Fri, sorry, was distracted by other issues
[20:22] <SpamapS> doko: sounds good, thanks.
[20:26] <tjaalton> bryceh: yeah, source format 3.0 playing tricks
[20:26] <tjaalton> bryceh: also, those changes should go to debian
[20:26] <bryceh> tjaalton, yeah, and I told the author so.  But isn't that why we put this junk on alioth to begin with rather than in bzr?  ;-)
[20:27] <tjaalton> bryceh: yes, but in debian branches and then merge ;)
[20:28] <tjaalton> not a big deal, will get sorted out at some point
=== ajmitch_ is now known as ajmitch
[20:29] <slangasek> tedg: simplified chinese is the one used in the PRC (so zh_CN)
[20:30] <bryceh> tjaalton, *shrug*  there was already a 0ubuntu1 version and the changes are minor, they can wait
[20:30] <slangasek> tedg: Taiwan uses the traditional because they were too busy fighting a war on their rear guard to worry about changing their alphabet
=== tyhicks` is now known as tyhicks
[20:30] <tjaalton> bryceh: exactly
[20:32] <tedg> slangasek, Heh
[20:32] <tedg> slangasek, Thanks!
[20:32] <tedg> Still seemingly can't change my time format, not sure why.
[20:33]  * bryceh waves to tedg
[20:33] <tedg> Howdy bryceh!
[20:34] <slangasek> tedg: what bug are we talking about, anyway? :)
[20:35] <tedg> slangasek, bug 1028866 -- trying to recreate.
[20:35] <ubottu> Launchpad bug 1028866 in indicator-datetime (Ubuntu Quantal) "the date is not displayed correctly in the "date and time setting" window" [High,Confirmed] https://launchpad.net/bugs/1028866
[20:35] <tedg> slangasek, Not sure why LC_TIME=zh_CN LANGUAGE=zh_CN gnome-control-center isn't working.  Get a locale not supported error.
[20:36] <tedg> Ah, there we go.  Needed a ".UTF-8"
[20:36] <tedg> Hmm, bug looks fixed :-)
[20:37] <tedg> The box really should resize for the last character though.
[20:38] <slangasek> tedg: fixed> מזל טוב
=== salem_ is now known as _salem
[20:50] <c_clapham> Hi, I want to get involved with helping to develop ubuntu but don't know where to start. Can anyone point me in the right direction? Thanks.
[20:51] <bryceh> c_clapham, depends on what you are interested in, and what you hope to get out of it, mind elaborating on your interests?
[20:55] <smoser> anyone else have old school (ie 2005) fonts after upgrade today?
[20:57] <dobey> infinity: can you go ahead and reject ubuntuone-client-data please? (sorry)
[21:00] <infinity> dobey: I would if I'd gotten around to uploading it for you.
[21:00] <infinity> dobey: Since I didn't, you're set!
[21:01] <dobey> infinity: ah, ok. then don't upload it :)
[21:01] <infinity> dobey: Done.  I'm good at doing things that amount to not doing anything.
[21:01] <infinity> dobey: I might even not do it twice.
[21:01] <dobey> heh
[21:05] <dobey> infinity: thanks for nothin'! :)
[21:18] <smoser> slangasek, are you going to upload mountall ?
[21:19] <smoser> i have a fix for cloud-init 'start networking' (replacing it with the emit net-device-added) but it doens't fix my problem entirely
[21:19] <smoser> https://launchpadlibrarian.net/115772437/cloud-init_0.7.0%7Ebzr644-0ubuntu1_0.7.0%7Ebzr644-0ubuntu2%7Eppa0.diff.gz
=== sbeattie is now known as retro-grouch
[21:30] <slangasek> smoser: not this week, since I'm not in a position to deal with any regressions and haven't yet tested it to my own satisfaction
[21:30] <slangasek> smoser: but yes, I am going to upload
=== retro-grouch is now known as sbeattie
[21:45] <SpamapS> smoser: my 30 second skim over that tells me its good :)
=== charles_ is now known as charles
[22:29] <bryceh> meh.
[22:29] <bryceh> @pilot out
=== udevbot changed the topic of #ubuntu-devel to: Quantal Quetzal development | Archive: Open | Dev' of Ubuntu (not support or app devel) | build failures -> http://qa.ubuntuwire.com/ftbfs/  | #ubuntu for support and general discussion for hardy -> precise | #ubuntu-app-devel for app development on Ubuntu http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots:
=== dendrobates is now known as dendro-afk
=== m4n1sh_ is now known as m4n1sh
=== cinerama_ is now known as cinerama
[22:37] <micahg> bryceh: requestsync -h when sponsoring
[22:37] <micahg> gah
[22:38] <micahg> requestsync -s
[22:38] <micahg> gah
[22:38] <micahg> I mean syncpackage -s
[22:40] <bryceh> micahg, ok, didn't realize that was required.  It can't figure that out?  It did prompt me to login to LP.
[22:40] <micahg> bryceh: it shows the sync requester on the -changes list
[22:40] <lamont> this wonderful thing where unity decides that I want a full screen window because I moved it and it bumped into the top... how do I make it stop trying to help me that way?
[22:41] <bryceh> micahg, ok.  Is that really all that important?
[22:41] <micahg> bryceh: it shows up on the requester's +synchronised-packages page and will show up when we look for previous uploads if someone applies for dev membership
[22:42] <micahg> it makes it like any other sponsored upload basically
[22:44] <micahg> bryceh: there's also a -b flag for closing the sync bug
[22:45] <bryceh> ah, I was looking for that
[22:45] <micahg> (which happens to include whether or not it's sponsored)
[22:45] <bryceh> *sigh* @ piloting.  bleah.
[22:46]  * micahg hugs bryceh
[22:51] <Laney> you don't feel happy with it?
[22:52] <ogra_> lamont, that will immediately start working magically after you convinced it to have sane alt-tab behavior :P
[22:53] <lamont> ogra_: I don't believe you
[22:53] <ogra_> heh
[22:53]  * ogra_ doesnt think either is easily changeable 
[22:54] <lamont> ogra_: more seriously, in the way before time, I smashed unity-3d in the face about buttons in the title bar (moved them to the right), the moved to 2d because the rendering speed made things unusable, and well, I'd like to move the buttons back to the left now without reinstalling the machine... got any magic sause for me?
[22:54] <ogra_> lamont, unity --reset
[22:54] <ogra_> that should reset to defaults
[22:55] <lamont> ah, but I still want focus follows *()^& mouse, even if that isn't quite as strict as I'd like it to be
[22:57] <bryceh> Laney, was that Q for me?
[22:58] <Laney> bryceh: yeah
[22:58] <ogra_> lamont, https://launchpad.net/unsettings
[22:58] <ogra_> lamont, thats what i used to switch back to FFM
[22:59] <lamont> ta
[22:59] <ogra_> (and for changing font settings etc)
[22:59]  * lamont runs to fetch a kid
[22:59] <infinity> ogra_: unity --reset is removed in Quantal.
[22:59] <infinity> ogra_: But there's some gsettings invocation that will do vaguely the same thing.
[22:59] <ogra_> sadly the utterly unusable alt-tab behavior is not fixable at all :(
[23:00] <ogra_> i really wonder who thought its a good idea to make it app based vs window based
[23:01] <infinity> lamont: The edge->max/tile behaviour is in the "Grid" plugin in ccsm.
[23:01] <lamont> oh, ogra_ btw, unity --reset?  they dropped that
[23:01] <infinity> lamont: Edge -> Resize Actions.
[23:01] <ogra_> lamont, yeah, infinity said so too
[23:01] <lamont> infinity: thanks
[23:02] <lamont> anyway, afk
[23:02] <ogra_> if ther is an equivalent unity --reset should really trigger it
[23:02] <bryceh> Laney, ah, yeah just frustrating.  sometimes piloting gets a bit drudgery, lots of little details to tend to
[23:04] <bryceh> but... a nice break from fixing graphics I guess ;-)
[23:06] <micahg> Laney: you should try piloting some time :)
[23:08] <Laney> yeah, I probably should
=== dendro-afk is now known as dendrobates
[23:46] <smoser> hm...
[23:46] <smoser> so i have one thought about my cloud-init fix https://launchpadlibrarian.net/115772437/cloud-init_0.7.0%7Ebzr644-0ubuntu1_0.7.0%7Ebzr644-0ubuntu2%7Eppa0.diff.gz
[23:46] <smoser> if an entry is listed in /etc/network/interfaces
[23:47] <smoser> but is not present
[23:47] <smoser> then i'' emit a 'added' event for it.
[23:47] <Presonus_Probs> Hi roomies :) first time user here, sorry if im in the wrong channel
[23:47] <smoser> i could check for an entry in /sys/class/net/ but i'm not guaranteed for /sys to be mounted at that point.
[23:53] <Presonus_Probs> Has anyone had any experience with Presonus Mixing desks and UbuntuStudio? Im having a few problems
[23:53] <Presonus_Probs> already asked #UbuntuStudio, thought I'd throw it in here just incase
[23:57] <TheMuso`> Presonus_Probs: I suggest you ask in #ubuntustudio.
[23:57] <TheMuso`> oh sorry missed your second comment.
[23:57] <Presonus_Probs> Thanks TheMuso
[23:58] <Presonus_Probs> :) thought id chuck it in here just incase it was a common problem devs would know about
[23:59] <TheMuso`> I don't think many, if any people in here work with firewire audio gear. Have you asked in any of the Linux audio IRC channels?
[23:59] <TheMuso`> Presonus_Probs: Try #lad as well.