[01:30] <slangasek> lucas_: so I know your relationship with ruby packaging runs hot and cold... :) do you have any interest in fixing bug #749099, or know someone who might?  libcairo-ruby fails to build because it bypasses pkg-config and tries to parse .pc files directly
[01:50] <infinity> slangasek: Ick.
[01:52] <slangasek> infinity: is that the sound of you volunteering to fix it? :)
[01:53] <infinity> slangasek: It's the sound of me vomiting a little at people who feel the need to re-invent pkg-config.
[01:54] <ScottK> If you're going to be around Ruby much, you'll be needed a large barf bag.
[01:54] <slangasek> infinity: I should share the list of all the things I've had to fix in build systems for multiarch :)
[01:54] <slangasek> php still wins, hands down
[01:55] <slangasek> but you probably could already have guessed that :)
[01:55] <ScottK> yeah.  that was a given.
[01:55] <infinity> slangasek: With our history with it, I can't imagine it surprised you, at least
[01:55] <slangasek> no, but I realized I'd repressed just how bad its autoconf usage was
[01:55] <infinity> Heh.
[01:59] <ScottK> cjwatson: It looks to me like something ate your dh-autoreconf backport https://launchpad.net/ubuntu/+source/dh-autoreconf/+publishinghistory
[01:59] <ScottK> re Bug #755661
[01:59] <ScottK> (I don't see it in New either)
[02:08] <psusi> say, umm... don't we have an auto generated list of packages that ftbs?  so why file bug reports on them as well?
[02:09] <slangasek> because a list of packages that ftbfs doesn't let you claim/assign the issues, triage them, or file them against different packages that are the ultimate source of the breakage
[02:12] <cjwatson> ScottK: no, I just forgot to flush it
[02:12] <ScottK> cjwatson: That's a relief.
[02:16] <cjwatson> ScottK: apologies for the duplicate NEW mails, lag-induced errors
[02:16] <ScottK> No problem.  Thanks for taking care of it.
[02:17] <cjwatson> accepted now
[02:20] <ScottK> Great.  I'll take care of binary New once it hits.
[02:21] <cjwatson> thanks
[03:43] <psusi> I have always explicitly specified -r or -h, but if you don't specify, shutdown seems to drop to runlevel s.  Is this normal?  the man page does not say what happens if you don't use any switches
[03:43]  * psusi is getting the feeling this is normal and this bug report should be converted into a request to improve the man page
[05:44] <abhinav-> cjwatson: Hi, one more question. Is "man -K" available only on Ubuntu ? I tried it on Opensuse , it wasn't available , although opensuse is also using man-db.
[06:03] <ohsix> doesn't that just call apropos, or vice versa
[06:03] <abhinav-> man -k calls apropos, in Ubuntu man -K does a full text search over man pages :)
[06:04] <hyperair> ooh nice. i never kenw that.
[06:04] <ohsix> you can grep the debdiff on the package page to see if it's added as a patch; that'll literally answer your question, but i don't personally know
[06:05] <ohsix> i see --global-apropos on lots of online manual listings
[06:06] <abhinav-> what does --global-apropos do ?
[06:07] <hyperair> maybe they all use ubuntu ;-)
[06:07] <hyperair> abhinav-: it's the longname for -K
[06:07] <abhinav-> oh
[06:07] <hyperair> it's in man man
[06:07] <ohsix> hyperair: looks like it could be :D
[06:07] <hyperair> heheh
[06:08] <abhinav-> :)
[06:08] <ohsix> hyperair: quick look at man page i know is modified for ubuntu doesn't have those modifications though
[06:08] <hyperair> mm what about debian?
[06:08] <hyperair> debian has -K as well.
[06:09] <ohsix> debian is awesome; they could be getting the man pages from there as well
[06:35] <lucas_> slangasek: [libcairo-ruby] I'll take a look
[06:35] <SpamapS> hrm.. autofs is frustrating.. no way to inform it that it should retry all failed mounts
[06:38] <ohsix> SpamapS: what are you using it for?
[06:38] <SpamapS> personally, nothing
[06:38] <slangasek> lucas_: ok :)
[06:38] <SpamapS> but trying to pinpoint when to start it during the boot
[06:39] <ohsix> its a bit of a speciality for a desktop thing when there's udisks
[06:39] <SpamapS> ohsix: basically in a server with static interfaces, 'started networking' is the exact right place to start autofs. On machines w/ dhcp or transient network connections.. this results in an autofs that takes 60 seconds to notice that it can mount an NFS share that it couldn't before. :-P
[06:40] <ohsix> i see
[06:41] <SpamapS> The kernel caches negative mount attempts and provides no way to flush that cache. :-P
[06:41] <SpamapS> just times them out
[06:41] <ohsix> theres a fuse filesystem that can do the same, but run arbitrary commands too
[06:42] <SpamapS> well regardless of that.. the current start on is totally broken.. 'and mounting TYPE=nfs' .. not sure what zul was thinking there. :-P
[06:43] <ohsix> lots of stuff is broken/weird with nfs :D
[06:43] <SpamapS> well autofs has lots of uses w/o nfs.
[06:43] <ohsix> right
[06:43] <SpamapS> and mounting is only emitted by mountall.. so not sure how it relates to autofs
[06:44] <ohsix> just saying theres a lot of weirdness to accept with nfs; how autofs interacts with it is just another one
[06:45] <SpamapS> meh.. autofs doesn't care what its mounting really.
[06:50] <ohsix> well mounting nfs manually is weird too; so i don't see how it'd drop out of the equation
[06:55] <SpamapS> Yeah it just depends on how you want to use nfs
[06:55] <lifeless> SpamapS: thats an oxymoron, no ?
[06:59] <SpamapS> true.. nobody *wants* to use NFS
[06:59] <SpamapS> lifeless: so, are you convinced yet that Cassandra is t3h suck? Or do I need to put Solandra/Lucandra on my oneiric-alpha-1 todo?
[07:01] <StevenK__> SpamapS: Bah, I use NFS here.
[07:02] <Guest27069> Services, bite my shiny metal ...
[07:04] <lifeless> SpamapS: I got good feedback from mdennis
[07:04] <lifeless> SpamapS: I thought I forwarded his 'can I'd love to come to UDS' mail to you/
[07:04] <lifeless> s/can/and/
[07:05] <SpamapS> lifeless: You may have. I'm just watching each patch release totally destroy at least a few peoples' installs and thinking they really don't understand what a stable release is yet.
[07:05] <lifeless> SpamapS: ouch
[07:06] <lifeless> SpamapS: that part I have no experience with yet
[07:06] <lifeless> SpamapS: have you played with lp:oopsrespository ?
[07:06] <lifeless> SpamapS: thats what I need solandra for
[07:06] <lifeless> s/need/think it might be nicer than solr
[07:06] <SpamapS> 0.7.1 and 0.7.2 were particularly ugly.. 0.7.3 gives you the tools to clean up the mess of those releases.. but then 0.7.4 broke anybody who went from < 0.7.3 to 0.7.4 ...
[07:07] <lifeless> SpamapS: though, TBH, if its still got the insane xml file for config
[07:07] <SpamapS> I haven't looked at all
[07:07] <lifeless> SpamapS: thats a bit uncool
[07:07] <SpamapS> It does seem like the best option for a hyper-scalable text search engine.
[07:09] <SpamapS> I have to wonder if there is something similar for Hbase
[07:30] <SpamapS> hmm.. I'm trying to upload a new version of autofs5 but dput is erroring out with "550 Changes file must be signed with a valid GPG signature: Verification failed 3 times: ["(7, 9, u'No public key')", "(7, 9, u'No public key')", "(7, 9, u'No public key')"] : Permission denied.
[07:32] <geser> SpamapS: you used the right gpg key (the one LP knows about)?
[07:33] <SpamapS> geser: indeed, F4BCB38E
[07:33] <SpamapS> have used it to upload before
[07:33] <SpamapS> wondering if my recent python problems might be the culprit
[07:35] <SpamapS> no.. something else.. dupload has the same problem
[07:35] <geser> if gpg --verify works on the .changes files, then I've no idea and try in #launchpad
[07:35] <geser> it's not dput generating the error message but the ftp server on LP
[07:37] <geser> alternatively you can also try to upload via sftp
[07:37] <SpamapS> yeah I was thinking maybe dput was mangling the file
[07:42] <SpamapS> hrm looks like it worked the first time despite the error
[08:27] <lifeless> so we check the signature during the ftp session now
[08:27] <lifeless> which is pretty cool
[08:27] <lifeless> Unless others are having the same issue
[08:28] <slangasek> unless you started checking within the past 2 hours, I'm not
[08:28] <lifeless> nope
[08:28] <lifeless> a week or so now
[08:28] <lifeless> we were hoping it would stop the mysterious 'I uploaded and no error and no email back' situation
[08:28] <lifeless> bigjools blogged aboot it
[09:29] <tjaalton> slangasek: actually, 1.4.2 was the first time there _was_ an ubuntu branch, it had been a local one until then.. I just should've checked the archive too before uploading
[09:29] <slangasek> tjaalton: oh, in that case yes, I blame you! ;)
[09:29] <tjaalton> darn :)
[09:30] <tjaalton> can't keep my mouth shut
[09:30] <slangasek> ;)
[09:35] <tjaalton> there's still xserver 1.10.1rc2 and mesa 7.10.2 the team would like to upload before the freeze, libx11 was just a warmup :)
[09:54] <cjwatson> abhinav-: it's not Ubuntu-specific; I implemented it in man-db upstream.  openSUSE are probably just a few versions behind and need somebody to prod them to come up to date.
[09:56] <cjwatson> abhinav-: The other GPLed man implementation, which used to be in Fedora (though Fedora switched to man-db recently) actually had -K first, although the implementation wasn't as good so it was several times slower to run.
[09:58] <abhinav-> cjwatson: oh ok. thanks. I was googling to learn more about man-db. I suppose man-db handles internationalization properly as opposed to the old man implementation ?
[09:58] <cjwatson> yes
[09:59] <cjwatson> when distros have switched to it, I think that's been the main reason for them
[10:00] <abhinav-> cjwatson: yes, I saw mailing list discussions, all of them had trouble with internationalization
[10:02] <cjwatson> yes, firstly it still uses catgets so it has no way to handle program translations in an encoding-safe way, and secondly it hasn't had enough work done on all the stuff you need to do to handle different encodings of manual pages themselves
[10:02] <cjwatson> I fixed all of that stuff in man-db years ago, so it's more than reached the point where distros might as well just switch if they're having problems
[10:03]  * cjwatson wonders if Gentoo will ever get around to actually doing that, having apparently decided to do so
[10:03] <abhinav-> :)
[10:04] <abhinav-> the only point I can mention in my proposal for not using man-db is that it is under gpl :)
[10:06] <cjwatson> yes, I believe in free software actually staying that way. :)
[10:07] <abhinav-> that actually gives me the opportunity to get this project :-D .
[10:20] <click170> I'm maintaining the TicGit-ng program, a fork of the original now unmaintained TicGit, but I've run into a problem. Both the deprecated TicGit gem and the TicGit-ng gem use the same bin file(s), so when installing the latter without the former, you get told that you need to install TicGit in order to use that bin file
[10:21] <click170> How do I resolve the problem on Ubuntu where trying to call a bin file with TicGit-ng installed results in a notice telling you to install the TicGit gem?
[10:22] <lifeless> sounds like your script isn't installed properly
[10:22] <lifeless> and you are triggering command-not-found behaviour
[10:23] <click170> Well, if I install the TicGit-ng gem on Debian, it works without problems. Do I need to do something more to satisfy a requisite for Ubuntu?
[10:25] <click170> Am I in the wrong channel? Should I be in the rubygems channel instead?
[10:26] <lifeless> I don't know :)
[10:26] <lifeless> what does 'which <your command name>' report
[10:27] <click170> 'which ticgitweb' has no results, but the file exists in '/var/lib/gems/1.8/bin/ticgitweb'.  This is a fresh install with git-core ruby and rubygems installed, and then TicGit-ng installed via gems.
[10:28] <lifeless> so, is /var/lib/gems/1.8/bin on your path?
[10:28] <click170> Apparently not. Sounds like a bug in Rubygems...
[10:47] <click170> lifeless: Thanks.  I've gone to answers.launchpad.net to ask for more information there and to find out if this is a bug in rubygems or not.
[12:14] <seiflotfy> any1 here using natty on an intel chipset
[12:14] <seiflotfy> core i5
[12:14] <seiflotfy> i cant  use my external monitor
[12:14] <seiflotfy> it keeps flickering
[12:15] <c2tarun> seiflotfy: what is your laptops model?
[12:19] <bradm> I'm having huge problems with my laptop due to bug 727620, its random if it'll boot or not
[13:37] <pitti> chrisccoulson: binutils> awesome, as it's in the archive now, do you think we can land this by tomorrow freeze?
[14:19] <chrisccoulson> pitti - yeah, can do. i've got some other globalmenu-extension bug fixes to land with firefox too though (most are fix committed already, but there is one which i haven't finished investigating yet)
[14:20] <chrisccoulson> i want to be sure i've not introduced any crashers before i upload though (so i'm running it today)
[14:20] <chrisccoulson> when is the cut-off tomorrow?
[14:37] <pitti> chrisccoulson: we freeze at 0900 UTC
[14:39] <chrisccoulson> pitti - oh, i guess i'll have to get it all fixed tonight ;)
[14:58] <pitti> chrisccoulson: Sundays are overrated :)
[14:58] <chrisccoulson> heh :)
[14:58] <chrisccoulson> jo works on sunday`s, which means i have my daughter here to help me
[14:58] <pitti> chrisccoulson: but in exchange you can slack tomorrow
[14:59] <chrisccoulson> and she likes using my laptop ;)
[14:59] <chrisccoulson> she has figured out that when she bashes the keypad, things happen on the screen
[14:59] <chrisccoulson> **keyboard
[14:59] <pitti> now this just needs a tad of coordination :)
[14:59] <chrisccoulson> lol
[15:00] <chrisccoulson> she's even tried drawing her own image in the screen with a felt tipped pen
[15:01] <chrisccoulson> i'm not sure she has figured out yet how things are normally rendered on to the screen ;)
[15:03] <pitti> chrisccoulson: oh, do you get that off the screen again?
[15:03] <chrisccoulson> pitti - she hasn't actually got as far as drawing on the screen yet. i stop her before she manages to do it
[15:03] <chrisccoulson> but she's got pretty close!
[15:03] <chrisccoulson> she's already drawn on the sofa ;)
[15:04] <chrisccoulson> but the cats do more damage to our sofa than her anyway ;)
[18:22] <Laney> anyone available to process the gio-sharp sync? want to rebuild banshee against it before the freeze
[19:32] <pitti> Laney: just doing
[19:32] <pitti> I did all outstanding syncs
[19:34] <Laney> cheers pitti
[19:38] <ricotz> pitti, hi
[19:39] <ricotz> docky needs to be built when gio-sharp is ready
[19:39] <pitti> ricotz: doesn't it have a versioned build-dep?
[19:39] <ricotz> docky copies the unstable lib to its private directory
[19:39] <ricotz> gio-sharp is considered unstable
[19:40] <pitti> then we'll need a -build1 upload, sorry
[19:40] <ricotz> ok
[19:40] <ricotz> no worries
[19:41] <Laney> i should have s/accepted/published/ in my comment
[19:41] <Laney> sorry
[19:42] <ricotz> pitti, but of course thank you for accepting it
[19:58] <lifeless> doko: hi
[20:43] <micahg> should I add tasks for the rdepends to bug 745544
[20:46] <ScottK> micahg: Sounds reasonable.
[21:26] <SpamapS> hrm.. twice now my box has panicced because building drizzle ate up all the RAM...
[21:26] <lifeless> SpamapS: what -j param are you passing?
[21:26] <lifeless> SpamapS: and -panic- wtf
[21:27] <SpamapS> lifeless: I know.. I'm getting an "OOM and no processes left to kill"
[21:27] <ScottK> I've hear lamont is particularly fond of -j 24 on armel.
[21:27] <SpamapS> lifeless: I'm not sure what -j it passes..
[21:27] <ScottK> hear/heard
[21:27] <SpamapS> I'm pretty much at "defaults"
[21:27] <lifeless> SpamapS: package or upstream build
[21:27] <SpamapS> lifeless: package
[21:27] <lifeless> whats your deb concurrency option set to
[21:28] <SpamapS> lifeless: I've never done anything to change it, so probably 1
[21:29]  * SpamapS is almost religious about using defaults unless he fully understands the setting being changed. :-P
[21:30] <SpamapS> plus w/ a dual core honestly using -j2 just means sacrificing being able to use the box. :-P
[21:34] <lifeless> SpamapS: heh
[21:34] <lifeless> SpamapS: my test build box is a (2 year old) i7
[21:34] <lifeless> SpamapS: 8 cores 4eva
[21:37] <SpamapS> I have one box
[21:37] <SpamapS> a laptop
[21:37] <SpamapS> everything else is "in the cloud" :)
[21:38] <SpamapS> Ok so I think this is related to the chroot upstart support
[21:38] <SpamapS> its happening when installing dbus
[21:38] <lifeless> -win-
[21:38] <lifeless> SpamapS: how old is your laptop that its only 2 core?
[21:39] <SpamapS> macbook pro 15 5,1
[21:39] <SpamapS>     1 root      20   0  962m 941m 1336 R  100 23.9   0:18.53 init
[21:39] <SpamapS> DOOHHH!!!
[21:39] <SpamapS> oh boy and its unstoppable
[21:39] <SpamapS>     1 root      20   0  962m 941m 1336 R  100 23.9   0:18.53 init
[21:39] <SpamapS> oops
[21:39] <SpamapS> 2.5G now
[21:40] <SpamapS> sync
[21:49] <SpamapS> haha my 'sync' came out in irc
[21:49] <SpamapS> box was a bit unresponsive by then.. :-P
[21:52] <SpamapS> ahh.. so dbus is special in that it sends SIGUSR1 to init...
[22:20] <DnaX> I've requested a FeatureFreeze Exception for Gweled package: bug #756877
[22:21] <DnaX> can anyone in the release team look this?
[23:04] <stgraber> cjwatson: sorry for the late response ;) I can confirm a race condition in plymouth-upstart-bridge. Out of 5 boots in my VM I only got one boot.log containing the upstart boot messages.
[23:05] <stgraber> it's testing with a desktop VM on my laptop, so it takes less than 5 seconds to boot the whole desktop. I'm guessing it's less visible on a "normal" system :)
[23:06] <stgraber> I'll see if I can "easily" fix that so I can then easily test bug 752393
[23:07] <stgraber> as I don't really want to reboot 5 times everytime I want to test something ;)
[23:07] <stgraber> I could also find a slower machine (as in, non-SSD and with less than 8GB of RAM), that should workaround it ;)
[23:08] <cjwatson> stgraber: could just be that dbus starts so shortly before gdm that there's no time for p-u-b to do anything
[23:09] <cjwatson> it's possible for both gdm and p-u-b to start immediately after dbus, after all
[23:09] <cjwatson> p-u-b is more useful on server systems
[23:11] <cjwatson> but you could also just shove 'pre-start exec sleep 10' in /etc/init/gdm.conf, or something like that
[23:12] <stgraber> cjwatson: well, my choice of using ubuntu desktop was to make the boot process slower ;) which kind of worked but only 1 out of 5 times ... server was booting even faster so that I'd never get a boot.log containing upstart entries :(
[23:14] <stgraber> I'm not sure the sleep 10 would work because on Friday I tried removing plymouth-stop.conf entirely, which means keeping a running plymouth after boot. But still didn't get anything in boot.log. I'm guessing the issue is really that p-u-b is starting "too late" and everything is already done
[23:19] <cjwatson> stgraber: roll on http://alban.apinc.org/blog/2011/02/16/introducing-multicast-unix-sockets/ or similar, and then we'd be able to talk to upstart without needing to wait for dbus-daemon to start
[23:21] <stgraber> that'd be nice indeed