| === Ursinha is now known as Ursinha-afk | ||
| hallyn_ | lifeless: i was going to push netcf to saucy so you could test the memory leak, but i seem to nto have upload perms again (was sure it had been added to server set last time...) but you should be able to build your own from http://people.canonical.com/~serge/netcf-saucy.debdiff | 01:06 |
|---|---|---|
| lifeless | heh, I'm on raring anyhow; will have a poke | 01:06 |
| hallyn_ | lifeless: should be almost identical (but will pushd ebdiff for that in just a moment - of course that will need sru to get into the archive :) | 01:07 |
| infinity | hallyn_: I can sponsor that for you. | 01:14 |
| infinity | hallyn_: Or I can sponsor it to Debian and you can sync. | 01:15 |
| === Ursinha-afk is now known as Ursinha | ||
| hallyn_ | infinity: thanks. originally i was going to wait for ahs3 to sponsor to sid and sync from there, but then i was thinking it would be worth testing in saucy immediately. so i can go either way. | 01:22 |
| infinity | hallyn_: Your call. | 01:23 |
| hallyn_ | infinity: ok, pls do push to saucy | 01:24 |
| hallyn_ | meanwhile i should figure out why gnulib (under netcf/) wont' build on jessie | 01:24 |
| infinity | hallyn_: Uploaded. And where is this gnulib jessie build failure? | 01:26 |
| infinity | hallyn_: You mean specifically jessie, but not sid? | 01:26 |
| infinity | ... in which I discover that I still don't have jessie chroots, only squeeze, wheezy, sid, and experimental. Oops. | 01:27 |
| hallyn_ | infinity: yeah, only jessie. one sec, the failure is: | 01:33 |
| hallyn_ | http://paste.ubuntu.com/5983302/ | 01:33 |
| infinity | Erm, that really *only* fails on jessie? | 01:35 |
| infinity | Now I need to create a chroot just to satisfy my curiosity. :P | 01:35 |
| hallyn_ | yeah wheezy and sid were both happy. | 01:36 |
| === JZTech101 is now known as JZ|Zzz | ||
| hallyn_ | so it's complaining that gets is undefined when it's being marked insecure? weird. | 01:36 |
| hallyn_ | apparently cjwatson worked around a similar bug in grub2, commenting in changelog "Avoid assuming that gets is declared" | 02:04 |
| hallyn_ | (still waiting for devscripts to install so i can look more closely) | 02:04 |
| * hallyn_ sees if he can just steal gnulib_gets.patch | 02:06 | |
| hallyn_ | yup | 02:13 |
| infinity | hallyn_: So, uhm, it builds on jessie for me. | 02:29 |
| infinity | hallyn_: http://lucifer.0c3.net/~adconrad/netcf_0.2.3-3ubuntu1_amd64.build | 02:30 |
| lifeless | hallyn_: builds fine on raring | 02:31 |
| hallyn_ | infinity: well that's whack. i needed: http://people.canonical.com/~serge/netcf-jessie.debdiff | 02:31 |
| infinity | hallyn_: That patch certainly won't hurt, but curious that it Works For Me. :P | 02:32 |
| lifeless | hallyn_: running that build now, will see if it leaks overnight | 02:32 |
| hallyn_ | infinity: wait, that's the wrong version | 02:32 |
| hallyn_ | infinity: jessie has 1:0.2.0-5, you compiled netcf_0.2.3-3ubuntu1.dsc ? | 02:33 |
| hallyn_ | lifeless: great - thanks | 02:33 |
| infinity | hallyn_: Oh, fair enough. But who cares, then? When 2.3-3 migrates, you're good. | 02:33 |
| hallyn_ | here's hoping | 02:33 |
| infinity | hallyn_: I assume 2.3-3 has a newer gnulib that deals with glibc >= 2.16 correctly. | 02:33 |
| hallyn_ | infinity: heh, i admit i'm not clear on the package migration path ever since jessie was introduced | 02:34 |
| infinity | hallyn_: jessie = testing. How is that unclear? :) | 02:34 |
| infinity | hallyn_: squeeze = oldstable, wheezy = stable, jessie = testing, packages migrate from sid to testing. | 02:34 |
| hallyn_ | what is it waiting on for the sid version to get to jesie? | 02:34 |
| infinity | hallyn_: Looks like migrating it causes libvirt to become uninstallable. | 02:35 |
| infinity | hallyn_: Perhaps just waiting for the two to migrate together, I didn't look. | 02:35 |
| hallyn_ | ok | 02:35 |
| hallyn_ | so is it more unorthodox to push fixes into jessie then rather than wait for them to get through sid? | 02:36 |
| infinity | hallyn_: One generally doesn't upload to testing, except in a freeze. | 02:36 |
| infinity | hallyn_: That would be equivalent to an Ubuntu developer uploading to saucy (which we just don't allow, we redirect everything to saucy-proposed, which is our own "unstable") | 02:37 |
| infinity | hallyn_: That said, 74 days without migration is impressive, you might want to look at why it breaks libvirt. :P | 02:37 |
| hallyn_ | gotcha. so a waste of time is what that was :) | 02:38 |
| hallyn_ | yeah | 02:38 |
| hallyn_ | noted for later thsi week | 02:38 |
| hallyn_ | all right on that note i'm heading to bed - \o | 02:40 |
| lifeless | hallyn_: looks good to me | 02:48 |
| hallyn_ | lifeless: meaning memleak is gone? | 02:49 |
| lifeless | yes | 02:49 |
| hallyn_ | \o/ thanks :) | 02:49 |
| lifeless | no growth at all | 02:49 |
| hallyn_ | maybe i'll just start the sru right now then | 02:49 |
| lifeless | hallyn_: actually, it may still be leaking | 03:55 |
| lifeless | hallyn_: several hours got a 1M increase. | 03:55 |
| lifeless | hallyn_: but massively better... | 03:55 |
| === solid_li1 is now known as solid_liquid | ||
| === solid_liquid is now known as CodeGuru | ||
| === CodeGuru is now known as solidliq | ||
| === JZ|Zzz is now known as JZTech101 | ||
| === ivanka__ is now known as ivanka | ||
| dholbach | good morning | 07:10 |
| dholbach | didrocks, good work1 | 07:53 |
| dholbach | ! | 07:53 |
| didrocks | dholbach: thanks dude ;) | 07:53 |
| Cimi | knocte, pong | 07:58 |
| knocte | Cimi: hello, I have a couple of bugs where I would love to get your feedback, do you mind looking when you have a chance? the bug numbers are.. | 07:59 |
| Cimi | bugs of what? | 07:59 |
| knocte | Cimi: ubuntu-themes | 08:00 |
| knocte | https://bugs.launchpad.net/light-themes/+bug/945430 and https://bugs.launchpad.net/bugs/1211831 | 08:00 |
| ubottu | Launchpad bug 945430 in ubuntu-themes (Ubuntu) "Lists lack zebra-striping" [Low,Triaged] | 08:00 |
| ubottu | Launchpad bug 1211831 in ubuntu-themes (Ubuntu) "properties of state pseudo-classes are not applied successfully " [Undecided,New] | 08:00 |
| Cimi | knocte, I'm no longer working on them atm... | 08:00 |
| Cimi | that's why nothing changed in the last cycles.., | 08:01 |
| knocte | Cimi: by "them" you mean the ubuntu themes? who should I ping then? | 08:01 |
| Cimi | no idea, probably me :-) | 08:01 |
| Cimi | I should really recruit someone from community to help approving branches | 08:02 |
| knocte | Cimi: I see, well you can give me a nickname and I'll try to convince her/him :) | 08:03 |
| Cimi | OK :-) | 08:05 |
| === dholbach_ is now known as dholbach | ||
| === slomo__ is now known as slomo | ||
| === seb128_ is now known as seb128 | ||
| === tkamppeter_ is now known as tkamppeter | ||
| didrocks | ev: hey, small question on whoopsie. How did you handle the fact that you propose the settings per user and the daemon is a system one? Do you have specific dbus call and it's the daemon writing the settings? | 09:57 |
| === gusch is now known as gusch|lunch | ||
| ev | didrocks: there's a whoopsie-preferences dbus daemon | 09:58 |
| didrocks | ah, so a separate dbus daemon | 09:58 |
| ev | ja | 09:58 |
| didrocks | then you just write to config files? | 09:58 |
| ev | mostly | 09:58 |
| ev | it will also start/stop whoopsie | 09:58 |
| didrocks | interesting | 09:58 |
| didrocks | ok, thanks ev | 09:58 |
| ev | sure thing | 09:59 |
| rbasak | stgraber, hallyn_: around? I can't install mysql-server-5.5 in an lxc container, because it tries to load an apparmor profile. I understand the reason (thanks to http://s3hh.wordpress.com/2012/05/03/lxc-in-precise-and-beyond/) but I wonder if we can make the apparmor profile load in the container a noop in this case? | 10:05 |
| rbasak | stokachu: any news on bug 1121874 please? mysql-5.5 is stuck in saucy-proposed - I think because of a dep8 failure because of the same regression. | 10:12 |
| ubottu | bug 1121874 in mysql-5.5 (Ubuntu) "MySQL launch fails silently if < 4MB of disk space is available" [Medium,Triaged] https://launchpad.net/bugs/1121874 | 10:12 |
| rbasak | It's be nice to get saucy unbroken too, which we can do either by fixing this properly, or by reverting your last upload. | 10:13 |
| === MacSlow is now known as MacSlow|lunch | ||
| === gusch|lunch is now known as gusch | ||
| seb128 | does anyone knowing how image build/tasks work exactly could help us to figure out what's the issue with cups-filters/ghostscript-cups | 11:50 |
| hallyn_ | rbasak: A few points, | 11:50 |
| seb128 | the first one is replacing the second and all rdepends got updated to depends on cups-filter (>= 1.0.36) | ghostscript-cups | 11:50 |
| seb128 | but that seems to make CD builds unhappy for some reasons | 11:50 |
| hallyn_ | rbasak: 1. stacked apparmor profiles should be coming soon, which will allow the apparmor profile load to work | 11:50 |
| seb128 | tkamppeter, ogra_: ^ | 11:50 |
| hallyn_ | rbasak: 2. the solutionright now is supposed to be to run the container unconfined (lxc.aa_profile = unconfined), I guess | 11:51 |
| hallyn_ | rbasak: 3. it seems reasonable to me to have mysql postinst allow apparmor profile load to fail if it is in a container, but we'd have to ask security team how they felt about it... | 11:51 |
| hallyn_ | jjohansen: mdeslaur: ^ | 11:52 |
| jamespage | slangasek, hey - I started looking at merging samba from Debian but noticed the unreleased splits of the init scripts that you had done had been dropped in unstable - was that intentional? | 11:53 |
| mdeslaur | jdstrand: ^ | 11:53 |
| hallyn_ | lifeless: well, drat, but i'll run it under valgrind again in a few days and see what i can find | 11:54 |
| tkamppeter | cjwatson, hi | 11:55 |
| mdeslaur | jdstrand: don't we already have code to deal with packages loading in a container? | 11:55 |
| === MacSlow|lunch is now known as MacSlow | ||
| siretart | infinity: I've uploaded libav9 to unstable yesterday, iow, the transition has started now in debian | 12:06 |
| rbasak | hallyn_: thanks. I was thinking about something like having the apparmor package arrange for _all_ apparmor loads to fail if they are in containers. | 12:10 |
| rbasak | I mean silently noop, not fail. | 12:10 |
| rbasak | (or perhaps just a warning to stderr) | 12:11 |
| rbasak | Given that we know that lxc security is currently still somewhat limited, I don't think that this would cause a surprise compromise in any real use case. | 12:11 |
| hallyn_ | rbasak: but also, | 12:18 |
| hallyn_ | rbasak: given that the host is already protected from mysql by the container profile, the mysql profile isn't as crucial. | 12:19 |
| hallyn_ | still, stacked profiles ftw :) | 12:19 |
| rbasak | Right | 12:27 |
| rbasak | Unless the user's expectation is that the inside of the container is protected by the mysql profile too. | 12:28 |
| hallyn_ | rbasak: that expectation would be wrong as per the server guide :) They can also expect user namespace protection, that woudl be wrong too. ("sadly" on both :( ) | 12:30 |
| hallyn_ | lifeless: ok so 'lxc-create -B dir' on a btrfs DOES avoid the subvolume creation for me | 12:30 |
| hallyn_ | I'm loath to change default behavior, but at the same time we're nto at 1.0 yet, and I'm not sure where we would document this... | 12:31 |
| pete-woods | fginther: hi again! I have another project for enabling CI on (https://launchpad.net/indicator-secret-agent) - it should have working support for test and coverage reporting already | 12:52 |
| fginther | pete-woods, ack, I can get started on it in a few minutes | 12:52 |
| pete-woods | fginther: awesome! :) | 12:52 |
| ev | so scan-build is pretty cool | 13:02 |
| lifeless | hallyn_: ok, so I'll follow your lead here. | 13:07 |
| lifeless | hallyn_: the main feedback I can give you is that while its nice (if you want it) that it happens by default, its very surprising, given that the btrfs docs say subvolumes are == filesystems | 13:08 |
| lifeless | hallyn_: you don't expect lxc-create to make a new filesystem by default, after all. | 13:08 |
| hallyn_ | lifeless: yeah maybe it shoudl be teh default with -B btrfs but not when -B is not set | 13:10 |
| hallyn_ | meaning, auto-detection of btrfs should perhaps be dropped | 13:10 |
| hallyn_ | but i'd have to get input from smoser on that at this ponit. pretty sure last time he said he was ok with that, | 13:10 |
| hallyn_ | but i'm not sure if by now he's come to depend on it | 13:10 |
| * hallyn_ has to go for a drive, bbl | 13:10 | |
| === _salem is now known as salem_ | ||
| jdstrand | mdeslaur, hallyn_ , rbasak: /lib/init/apparmor-profile-load already has code to not load in a container | 13:32 |
| jdstrand | rbasak: are you using /lib/init/apparmor-profile-load or are you using the new apparmor upstart stanza? | 13:32 |
| jdstrand | s/apparmor upstart stanza/upstart apparmor stanza/ | 13:33 |
| tkamppeter | cjwatson, can you have a look at bug 1212239? It seems to be a problem of how the images get generated. | 13:45 |
| ubottu | bug 1212239 in cups-filters (Ubuntu) "cups-filters is reported to have unmet dependencies with the latest saucy server installations" [Undecided,New] https://launchpad.net/bugs/1212239 | 13:45 |
| smoser | lifeless, other than your understanding of some bit of documentation in a man page for btrfs, do you have some reason as to why creating subvolumes is bad ? | 13:52 |
| smoser | hallyn_, i think maybe we want something like '-B ANY', '-B NONE', '-B IF_POSSIBLE', '-B btrfs,lvm,none' | 13:54 |
| rbasak | jdstrand: that's odd. The upstart script uses /lib/init/apparmor-profile-load, so I'm not sure why that isn't working. Thanks - I'll look in detail again later. | 14:00 |
| jdstrand | rbasak: note that apparmor-profile-load does this: | 14:01 |
| jdstrand | [ -x /bin/running-in-container ] && /bin/running-in-container >/dev/null 2>&1 && exit 0 | 14:01 |
| jdstrand | rbasak: so maybe running-in-container doesn't exist or isn't working right? | 14:01 |
| rbasak | jdstrand: I'll check. Also, would dh-apparmor in debian/rules affect anything? | 14:02 |
| jdstrand | rbasak: it will put in postinst something to load the profile into the kernel | 14:06 |
| === blitzkrieg3 is now known as jmleddy | ||
| jmleddy | cjwatson: hi, we have some more information about the 4k/4k disk issue bug 1065281 | 14:19 |
| ubottu | bug 1065281 in OEM Priority Project quantal "Installer crashed when trying to partition 4k/4k sector hard disks" [High,In progress] https://launchpad.net/bugs/1065281 | 14:19 |
| jmleddy | we're finally able to get the machine into a state where it can boot | 14:19 |
| === and`_ is now known as and` | ||
| rbasak | Wow. That's an impressive bug! | 14:28 |
| hallyn_ | smoser: -B btrfs,lvm,dir I'm ok with. The others sound too magical. (and '-B dir' is -B NONE IIUC) | 14:32 |
| smoser | literal 'dir' ? | 14:32 |
| hallyn_ | yes | 14:32 |
| hallyn_ | dir for directory | 14:32 |
| smoser | hallyn_, that would be fine then. | 14:33 |
| smoser | the only thing i dont like is that i dont want ot have to know a list of what is availalbe. | 14:33 |
| smoser | thus 'ANY' | 14:33 |
| hallyn_ | smoser: if there were more possibilities that would make sense... the thing about lvm and zfs is that there are more parameters | 14:34 |
| smoser | ah. well then the comma delimited falls apart anyway | 14:34 |
| hallyn_ | we could simply assume that if a vg called lxc or zfs group called lxc exists, | 14:34 |
| hallyn_ | then it's "available". | 14:34 |
| smoser | i would do that. | 14:34 |
| hallyn_ | anyway I'm goign to opena bug where we can collect this info and make a decision | 14:34 |
| hallyn_ | smoser: thanks | 14:35 |
| rbasak | jdstrand: turns out that there were two errors, and the postinst dh-apparmor warning fooled me into thinking that it was an apparmor problem. | 14:35 |
| smoser | i sweare there was a feature in openstack now for termination protection | 14:36 |
| rbasak | jdstrand: dh-apparmor does "apparmor_parser -r -T -W "$APP_PROFILE" || true" which fails with "apparmor_parser: Unable to replace "/usr/sbin/mysqld". Permission denied; attempted to load a profile while confined?\nWarning failed to create cache: usr.sbin.mysqld" but of course doesn't cause the postinst to fail. | 14:36 |
| rbasak | jdstrand: thanks for your help. | 14:36 |
| jdstrand | rbasak: ah, good, glad I could help :) | 14:37 |
| hallyn_ | smoser: btw, not sure if lifeless has responded (doesn't look like it) - creating subvolumes can be bad because rsync -va --one-filesystem will not back up the subvolumes | 14:41 |
| hallyn_ | smoser: lifeless: bug 1212290 is to track this. | 14:43 |
| ubottu | bug 1212290 in lxc (Ubuntu) "update backing store defaults at lxc-create" [Low,Triaged] https://launchpad.net/bugs/1212290 | 14:43 |
| smoser | hallyn_, | 14:47 |
| smoser | it seems daily ppa doesnt have | 14:47 |
| smoser | source-precise-amd64 | 14:47 |
| smoser | /usr/share/lxc/templates/lxc-ubuntu-cloud: line 388: /usr/share/lxc/hooks/ubuntu-cloud-p | 14:47 |
| smoser | rep: No such file or directory | 14:47 |
| smoser | lxc-create: container creation template for source-precise-amd64 failed | 14:47 |
| hallyn_ | oh, feh. | 14:48 |
| hallyn_ | stgraber: can you add that to lxc.install? | 14:48 |
| hallyn_ | (I figured it did the whole hooks directory) | 14:48 |
| smoser | hallyn_, so did i. | 14:49 |
| smoser | and i never bothered 'make install' | 14:49 |
| smoser | :-( | 14:49 |
| smoser | sorry | 14:49 |
| hallyn_ | hm, it does | 14:49 |
| smoser | http://paste.ubuntu.com/5985238/ | 14:50 |
| hallyn_ | smoser: your patch didn't add it to hooks/Makefile.am | 14:50 |
| smoser | right | 14:50 |
| hallyn_ | lemme do that and runa new build | 14:50 |
| smoser | you said "i figured it did the whole hooks directory" | 14:50 |
| smoser | i figured it did that by default too. but it has a manual list. | 14:50 |
| smoser | our "it" was just different in this reguard. | 14:50 |
| hallyn_ | :) | 14:50 |
| smoser | sorry to hvae been a pain hallyn_ | 14:51 |
| smoser | hallyn_, i could make that have sane behavior if you are interested. | 14:52 |
| hallyn_ | smoser: np :) pushed to git, building pkg now... hopefully won't take too long | 14:52 |
| smoser | ie, for hooks to install everything not named 'Makefile.*' | 14:52 |
| hallyn_ | smoser: the makefile? sure | 14:52 |
| smoser | or maybe everything executable | 14:52 |
| hallyn_ | that'd probably be best | 14:53 |
| smoser | hallyn_, it seems non-obvious to me how i'd do it without a grep essentially | 14:55 |
| smoser | wildcard in gnu make does not allow for exclude | 14:55 |
| hallyn_ | can you do some $(filter) crap? | 14:56 |
| hallyn_ | (just thinking however debian/rules often detects arch) | 14:57 |
| hallyn_ | smoser: eh, it's probably not worth it. it's not something that will cause regressions | 14:58 |
| smoser | hallyn_, fwiw, this works | 15:02 |
| smoser | http://paste.ubuntu.com/5985282/ | 15:02 |
| smoser | but would include other stuff in that dir | 15:03 |
| hallyn_ | smoser: looks good to me. more stuff shouldn't be added that isn't meant to be shipped, but if it is we can just make a $exclude variable including makefile* | 15:04 |
| smoser | hallyn_, i'dm be more concerned about stuff in a local checkout | 15:06 |
| smoser | getting installed accidently | 15:06 |
| smoser | ie, foo.bk1 | 15:06 |
| hallyn_ | heh, you're not supposed to do that, you're supposed to buidl a package :) | 15:07 |
| hallyn_ | ok, let's leave it as is | 15:07 |
| pete-woods | didrocks: good afternoon! I have another project I'd like to do releases of (https://launchpad.net/indicator-secret-agent) - it already has jenkins autolanding activated | 15:40 |
| didrocks | cyphermox: I think that one will be for you ^ | 15:40 |
| didrocks | (please add to the spreadsheet) | 15:41 |
| didrocks | pete-woods: any integration tests? | 15:41 |
| pete-woods | didrocks: it's not integrated with the shell yet (I'm waiting on the new system dialogue stuff) - but it does have 97% unit test coverage | 15:41 |
| didrocks | great ;) | 15:42 |
| pete-woods | didrocks: as soon as there is actually some integration happening, I'd be more than happy to write integration tests, however it is I'm supposed to do that (autopilot?) | 15:43 |
| cyphermox | yeah | 15:44 |
| cyphermox | alright | 15:44 |
| === Ursinha-afk is now known as Ursinha | ||
| === gusch is now known as gusch|away | ||
| === JZTech101 is now known as JZ|Zzz | ||
| === gusch|away is now known as gusch | ||
| === jasoncwarner__ is now known as jasoncwarner | ||
| infinity | smoser: Your cloud-init SRU has a patch in there (future_utils.patch) with no reference or explanation in the changelog. | 18:23 |
| infinity | smoser: Rejecting on that basis, please either reference it so I know what it is, or drop it because it's cruft. :P | 18:24 |
| smoser | boo. | 18:24 |
| smoser | its just stuff from cloudinit/utils.py in newer versions of cloud-init. | 18:24 |
| infinity | smoser: Added to make the azure backport easier, perhaps? | 18:25 |
| smoser | yes. | 18:25 |
| infinity | smoser: I don't like guessing. :P | 18:25 |
| infinity | smoser: But okay, I can let that slide. | 18:26 |
| infinity | smoser: It's at least more readable than the azure patch... | 18:26 |
| smoser | i'm ok to re-upload if you'll look quickly | 18:26 |
| smoser | i kind of a gree that the patch there should have a dep-3 header | 18:27 |
| infinity | smoser: I'm looking at it all this afternoon/evening, after some of my walinuxagent concerns are addressed too. | 18:27 |
| smoser | so you want to just NAK that and i'll re-upload ? | 18:27 |
| infinity | smoser: Rejected. | 18:27 |
| ogra_ | infinity, did you notice the image build failures today ? | 18:28 |
| infinity | ogra_: I'm off sick (can't you tell from my inactivity on IRC), so no. :P | 18:29 |
| ogra_ | seems we have an issue with the old "tasks resolve deps differently to metapackages" .... | 18:29 |
| ogra_ | haha, ok, i wont bother you | 18:29 |
| ogra_ | its a cjwatson thing anyway i think | 18:29 |
| infinity | Erm, wait, which image(s) is this that you're talking about? | 18:30 |
| infinity | ogra_: Oh, the cups thing? | 18:30 |
| ogra_ | infinity, yeah | 18:31 |
| ogra_ | bug 1212239 | 18:31 |
| ubottu | bug 1212239 in cups-filters (Ubuntu) "cups-filters is reported to have unmet dependencies with the latest saucy server installations" [Undecided,New] https://launchpad.net/bugs/1212239 | 18:31 |
| smoser | infinity, so i just upload again with same version, right? | 18:31 |
| infinity | smoser: Yep. | 18:31 |
| ogra_ | seems the deps are fine, but the task resolves them backwards | 18:31 |
| ogra_ | (the "or" dep specifically) | 18:31 |
| ogra_ | i dont think we ever found a proper solution for that | 18:31 |
| === JZ|Zzz is now known as JZTech101 | ||
| smoser | is 'dch -D some-thing --release' broken for anyone else? | 18:37 |
| infinity | ogra_: It's because it recommends and conflicts. | 18:39 |
| ogra_ | oh | 18:39 |
| === Nisstyre-laptop is now known as Nisstyre | ||
| infinity | ogra_: At least, that's my best guess. And it's a bug either way. | 18:40 |
| ogra_ | yeah | 18:41 |
| * infinity fixes. | 18:41 | |
| === JZTech101 is now known as JZ|Zzz | ||
| esing | hi | 18:44 |
| esing | How do I install gummiboot on ubuntu | 18:44 |
| infinity | ogra_: Uploaded. Will test in a chroot to make sure this fixes it once it's built. | 18:45 |
| * ogra_ hugs infinity | 18:46 | |
| ogra_ | (dont infect me though) | 18:46 |
| smoser | infinity, i uploaded. http://paste.ubuntu.com/5986081/ is diff from last upload. | 18:55 |
| infinity | smoser: Shiny, thanks. Once I get utlemming's fixed walinuxthingee, I'll do them together. | 18:56 |
| infinity | ogra_: Right, this should definitely fix it. The problem was that cups-filters Recommending ghostscript-cups landed the ghostscript-cups package in the ubuntu-desktop task. So, when you went to install the task, it exploded. | 19:14 |
| infinity | ogra_: Can't test for sure until it's all migrated to the release pocket and gs-cups falls out of the task, but logically it makes sense. | 19:14 |
| ogra_ | yeah, it definitely does ... | 19:15 |
| === JZ|Zzz is now known as JZTech101 | ||
| infinity | Bah, and same fix needs to be applied to cups itself. | 19:16 |
| === dosaboy_1 is now known as dosaboy | ||
| infinity | Aaaand, the cups testsuite fails now. Awesome. | 19:39 |
| infinity | tkamppeter: HALP. CUPS sucks. | 19:39 |
| ogra_ | hahaha | 19:40 |
| infinity | FAIL: 35 error messages, expected 33. | 19:41 |
| infinity | So informative. | 19:41 |
| tkamppeter | infinity, it worked for me last time, what did you change? | 19:45 |
| infinity | tkamppeter: Nothing in the code. | 19:45 |
| infinity | tkamppeter: https://launchpad.net/ubuntu/+source/cups/1.6.3-1ubuntu1 | 19:45 |
| tkamppeter | infinity, and why are you rebuilding CUPS? | 19:45 |
| infinity | See above. | 19:45 |
| infinity | tkamppeter: As it turns out, though, my motivation for uploading doesn't change whether or not the testsuite works. :P | 19:48 |
| infinity | tkamppeter: Anyhow, image builds will continue to be broken until a cups that doesn't Recommend ghostscript-cups actually manages to build. I leave the testsuite breakage to you, since I suspect you know it much better than I do. | 19:49 |
| infinity | E [14/Aug/2013:13:50:34.039455 -0600] Filter "gstoraster" not found. | 19:53 |
| infinity | E [14/Aug/2013:13:50:34.039476 -0600] Filter "gstoraster" not found. | 19:53 |
| infinity | I'm assuming those are the lines to blame.. | 19:54 |
| tkamppeter | infinity, I see, CUPS needs to build-depend on cups-filters now where Ghostscript-cups has gone away ... | 19:58 |
| infinity | tkamppeter: It does. | 19:58 |
| infinity | tkamppeter: I have it installed here, make check still fails with that. | 19:58 |
| infinity | tkamppeter: http://paste.ubuntu.com/5986315/ | 20:00 |
| infinity | tkamppeter: So, something more weird is going on... | 20:00 |
| tkamppeter | infinity, the test suite of CUPS is really broken, it should not depend on cups-filters nor on ghostscript (which is needed by gstoraster which the test suite tries to call). | 20:08 |
| infinity | tkamppeter: Sure, but both cups-filters and ghostscript *are* build-deps, and are installed. | 20:12 |
| infinity | tkamppeter: And it's the server startup that's whining about gstoraster being missing. | 20:12 |
| infinity | Well, "missing". It's obviously there. :/ | 20:13 |
| tkamppeter | infinity, CUPS does a really bad thing, it links the filters contained in cups-filters one by one into the test filter directory, meaning that CUPS' test suite needs to be changed with every change of cups-filters making the "make check" of CUPS only work with a few versions of cups-filters. | 20:43 |
| infinity | tkamppeter: Ah, I *just* got there myself and was about to upload a fix. :) | 20:48 |
| tkamppeter | infinity, debian/patches/tests-use-cupsfilters.patch needs to get fixed, adding a link for gstoraster. | 20:48 |
| infinity | tkamppeter: Already one it. | 20:48 |
| infinity | s/one/on/ | 20:48 |
| infinity | tkamppeter: Test building, thanks for the extra set of eyes on that one. | 20:53 |
| tkamppeter | infinity, did the package build for you now? | 21:00 |
| infinity | tkamppeter: Oh, erk. There's also the IPP compliance tests failing. Fun. | 21:00 |
| infinity | tkamppeter: http://lucifer.0c3.net/~adconrad/cups-str-1.6-2013-08-14-adconrad.html | 21:02 |
| brendand | unity isn't showing even if i run it from a terminal | 21:19 |
| brendand | what could be missing? | 21:19 |
| === JZTech101 is now known as JZTech101_ | ||
| === JZTech101_ is now known as JZTech101 | ||
| === Nisstyre-laptop is now known as Nisstyre | ||
| infinity | tkamppeter: Yeah, those ipp failures have me a little bit stumped. Possibly a race between creating fake jobs and looking for them? | 21:26 |
| brendand | sorry to ask again, i had to reboot so may have missed a response. i've upgraded to saucy and unity won't seem to start. unity_support_test returns 0 though | 21:48 |
| slangasek | jamespage: "dropped in unstable"> how do you mean, "dropped"? It hasn't been landed yet at all - but that's something I'm working through this week at DebConf as part of the upstarting of the package | 22:06 |
| slangasek | jamespage: "dropped in unstable"> how do you mean, "dropped"? It hasn't been landed yet at all - but that's something I'm working through this week at DebConf as part of the upstarting of the package | 22:08 |
| === james_ is now known as Guest53848 | ||
| tkamppeter | infinity, I have done test builds of CUPS 1.6.3 and 1.7rc1 on my Saucy machine now and I get exactly the same IPP failures. Seems to be that something changed in Saucy which breaks the CUPS tests. | 22:21 |
| infinity | tkamppeter: Alright, I'm uploading an ugly hack for now to ignore the IPP failures, so this can build and images can be buildable again, but obviously this needs looking into. | 22:30 |
| === salem_ is now known as _salem | ||
| tkamppeter | infinity, thank you very much, and sorry for the inconvenience. The test suite of CUPS is really a problem. | 22:43 |
| blitzkrieg3 | ay | 23:28 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!