/srv/irclogs.ubuntu.com/2010/05/20/#ubuntu-devel.txt

=== dendrobates is now known as dendro-afk
=== sconklin is now known as sconklin-gone
nigelbpitti: this is the trouble with apport I was talking about yesterday malone bug 58310900:58
ubottuLaunchpad bug 583109 in Apport "attach_gconf collects default settings instead of non-default" [Undecided,New] https://launchpad.net/bugs/58310900:58
=== beek_ is now known as kentb
=== jjohansen is now known as jj-afk
=== dendro-afk is now known as dendrobates
=== dendrobates is now known as dendro-afk
bluefoxicyso, besides a license that basically says "Do whatever the fuck you want to," there is also a license that requires you to be dead to use the software.03:14
bluefoxicyI swear you just can't make this shit up.03:14
Gbluefoxicy: what requires you to be dead to use?03:22
=== maco is now known as maco2
=== maco2 is now known as maco
m4tis there a way, either through session or system dbus, to get mouse/kb idle for an x session?04:55
m4ti downloaded like 600MB of xmlto dependencies to compile gnome-screensaver docs, but the listed GetSessionIdleTime doesn't seem to exist04:55
=== lifeless_ is now known as lifeless
m4tmaybe i want consolekit's GetSystemIdleSinceHint?04:58
dholbachgood morning07:09
ddecatordidrocks: ping07:24
=== SandGorgon_ is now known as SandGorgon
=== dejai is now known as bwright
=== radoe_ is now known as radoe
didrocksddecator: yes?08:07
ddecatordidrocks: hey, i was just wondering if the apport hook mentioned in bug 579548 is something you guys would really like, and if you'd be willing to have me write one for you08:09
ubottuLaunchpad bug 579548 in clutter-1.0 (Ubuntu) "Add apport hook for lspci" [Undecided,New] https://launchpad.net/bugs/57954808:09
didrocksddecator: yeah, the hardware stuff is the first step. TBH, I'm not sure about additional info that can be relevant, maybe glxinfo will be good?08:11
didrocksddecator: if you can do it, that would be awesome :)08:11
ddecatordidrocks: yah, i'm just starting to learn how to write hooks, but that one seems simple enough. if you can just let me know what info exactly you would like it to grab, i can work on getting it setup tomorrow and attach it to the report for you guys to review :)08:12
didrocksddecator: sweet, you can maybe also have some look at clutter forums to see what's people are asking for too. That can be a good hint about what to include in the report08:13
=== bwright is now known as dejai
ddecatordidrocks: sure thing, i'll look into that tomorrow (2:14am here right now, haha)08:14
didrocksddecator: take some rest ^^ In any case, I won't really have time to review it before next week :)08:15
ddecatordidrocks: good deal, thanks for getting back to me :)08:16
didrocksddecator: you're welcome. Thanks for tackling that! Just ping me when you need some help08:17
pittiGood morning08:18
ddecatormorning pitti08:18
pittijdstrand: thanks08:18
pittikees: ok, can do; I thought about globally resetting it tomorrow, but I can do those individually as well08:18
RAOFddecator, didrocks: If you need any help you probably want to look at the X apport hooks, which already collect that information.08:21
ddecatorRAOF: thanks for the tip08:22
mb4_morning09:26
mb4_can i ask questions about packaging here?09:26
pittikees: updated db, charts should update in 30 mins09:40
pittimb4_: yes09:43
=== StEdward is now known as saispo
=== smb is now known as smb-afk
juliankmvo: Do you have any plans to move to update-manager 0.200 for maverick? Having to mostly independently developed series of update-manager does not look like a good thing.11:21
* cjwatson wraps his head around the dpkg merge11:43
mvojuliank: it was on the agenda for last cycle already, but time was/is a issue12:08
dholbachjuliank: I just had a look over it, it seems like the version in Debian needs to still merge a lot12:08
juliankmvo: OK.12:11
juliankdholbach: Yeah, I know that there are large differences in some parts.12:11
mvojuliank: btw, did you notice that I merged the "unbranded-icons" branch into trunk? that is probably useful for the debian packages of s-c12:12
juliankmvo: I currently use the standard package management icons (system-software-install) and I may stay with 2.0.X for squeeze to have the same as lucid, unless someone asks me to ship the same release as maverick.12:15
mvojuliank: ok, sounds good. trunk/ is in a lot of churn currently, so staying with 2.0.x is a good choice12:15
mvoturnk is also pretty exciting, but that is a different matter :)12:16
juliankmvo: I will merge the 2.1 stuff into a debian-next branch once you released it.12:16
mvook12:16
juliankmvo: You forgot to merge the changes for debian/control, BTW; including the one to only conflict with gnome-app-install (<< 1) [without the version, it breaks the upgrade path]. I also reformatted a bit there, since tabs and spaces where mixed in trunk.12:18
juliank(And without merge information, merging it back becomes increasingly complicated)12:19
mvojuliank: ok, I fix that12:19
jdstrandpitti: fyi, still waiting on sparc and powerpc (est start time, 3 and 4 hours respectively)12:22
=== cking is now known as cking-afk
=== smb-afk is now known as smb
mvojuliank: I added a unbranded desktop file now too, so with the icons and the desktop file its just the .ui file that needs some work12:34
=== cking-afk is now known as cking
juliankmvo: OK.12:45
=== jelmer is now known as Guest38382
julianktseliot: Is there a branch for x-kit somewhere?12:46
mvojuliank: if you cherry pick r789,790 then that should make the branding easier, I moved code into the Distro class where it belongs12:50
=== thegodfather is now known as fabbione
dmartHas anyone been having trouble with NFS servers running on lucid?13:03
dmartI'm finding that when a client executes fcntl(... F_SETLKW ...) in an NFS chroot, it hangs for minutes at a time13:03
dmartnormal file assess works OK though, and lockd processes are present on client and server13:04
dmartThe server lockd is stuck in disk sleep state the whole time, but since it's a kernel process this is possibly normal.13:04
=== Guest38382 is now known as jelmer
pittijdstrand: please let me know if you need build score bumping, etc.13:09
jdstrandpitti: I think it would probably be a good idea (at least for hardy). redhat released today. if we plan on it being in -proposed for a week, then it doesn't matter. if we want to release today, it would probably be good13:11
jdstrandpitti: I am testing the i386 and amd64 binaries this morning13:11
dmartslangasek: Have you seen rpc.statd failing to launch on lucid?  It looks like this may be the cause of my problem (see scrollback)13:12
tseliotjuliank: it's here: https://launchpad.net/xorgparser why are asking?13:16
julianktseliot: Debian packaging. The branch contains a debian/ directory for 0.4.2, but code from 0.4.2.2 though.13:18
tseliotjuliank: right, it shouldn't contain a debian directory. The tarball doesn't contain it13:20
julianktseliot: The package is native, so from my point of view something went really wrong then.13:21
tseliotjuliank: ?13:21
tseliotjuliank: what's wrong with having a native package? x-kit is not in debian13:23
julianktseliot: The package is a native package with one tarball and the version 0.4.2.2; although it should be 0.4.2.2-1 and using a .diff.gz on the upstream tarball (which despite your statement above, also contains the debian/ directory)13:23
juliankor 0.4.2.2-0ubuntu113:23
tseliotjuliank: only 0.4.2.2 contains the debian directory and it was mistake13:24
tseliotthe other tarballs don't have it13:25
julianktseliot: Well, but the package should then still be not native, since you have two different tarballs with the same version. A native package is only for the case where the upstream tarball is equal to the package; that's not the case here.13:26
juliankIf you do upstream releases without a debian/ directory, your package must not be native.13:26
juliankThe tarballs even complete the .bzr directory!!!13:28
julianks/complete/contains/13:28
tseliotyes, I've noticed that (in the launchpad page)13:28
jdstrandpitti: not sure if you rescored yet, but hardy is currently scheduled to 1-3 hours, so maybe not worth it13:29
tseliotjuliank: but yes, good point, I can replace the upstream tarball and correct the version13:34
julianktseliot: Never replace an existing tarball, it will break the assumptions of your users.13:35
tseliotjuliank: the one in the launchpad page13:35
julianktseliot: Your upstream users.13:35
julianktseliot: You don't work with Python very much, right? (My points being: non-PEP8 function names (myFunc instead of my_func), comparisons to None using == instead of 'is', classes not derived from object ['class MyClass:' should be 'class MyClass(object):')13:39
tseliotjuliank: I wrote that code some time ago, now I have different habits. I used to adopt the QT coding style but now it's no longer the case.13:41
tseliotand I don't want to break the api13:42
ScottKIt's not rare to see PyQt code that follows Qt conventions instead of Pythonic ones.13:43
juliankScottK: We're talking about simple Python code, not Qt related.13:44
ScottKAh, sorry.  Poor assumption on my part.13:44
tseliotjuliank: may I ask the purpose of your code review?13:46
tseliotjust curious13:46
julianktseliot: Well, I just want to understand it, to consider how to go on with packaging jockey for Debian (which needs python-xkit). I can either package it standalone as in Ubuntu or merge X-Kit into jockey and not expose the API at all.13:47
juliankIt's just the question of whether I want to support it generally or only for jockey-internal use.13:49
tseliotjuliank: ok13:50
dmartKeybuk: hi there, do you have a moment for a quick upstart question?14:00
dmartKeybuk: never mind, looks my issue is already tracked here: https://bugs.launchpad.net/ubuntu/+source/nfs-utils/+bug/58194114:03
ubottuUbuntu bug 581941 in nfs-utils (Ubuntu) "statd does not start automatically when needed nor can be forced to start on boot" [Undecided,Incomplete]14:03
=== vanhoof_ is now known as vanhoof
nigelbpitti: got a min?14:17
pittinigelb: what's up?14:17
nigelbremember I was talking to you about some trouble with nondefault gconf values?14:17
nigelbhere's the bug, bug 58310914:17
ubottuLaunchpad bug 583109 in Apport "attach_gconf collects default settings instead of non-default" [Medium,In progress] https://launchpad.net/bugs/58310914:17
pittinigelb: yep, saw it an assigned to me14:18
nigelbpitti: ok, so you're faster ;)14:18
nigelbshall we proceed as usal with the hook and let this bug be fixed in apport?14:18
nigelbi.e. not wait for this to be fixed for the hook14:19
=== barry` is now known as barry_
=== ]reed[ is now known as [reed]
apwdoes anyone know if we have any 'first boot' processing which occurs after a karmic to lucid upgrade but not after a scratch install of lucid ?14:27
pittinigelb: yes, I think so; it's independent from the hook14:27
nigelbpitti: ok, tks :)14:28
=== barry` is now known as barry_
=== barry_ is now known as barry
cjwatsonapw: not unless individual packages implement it14:34
cjwatsonapw: (postinst scripts can tell whether they're being upgraded or fresh-installed)14:34
juliankLaunchpad always times out ....14:35
smoserpersia, or cjwatson or dmb member.14:40
smosertrying to add my feed to planet ubuntu14:40
smoseri can't write to lp:~planet-ubuntu/config/main planet-ubuntu14:40
cjwatsonthat's not a dmb thing ...14:40
smoserwell, dmb approved me14:40
cjwatsonwell, wait, were you given per-package upload rights?14:40
cjwatsonI may have forgotten to add you to ~ubuntu-dev14:40
=== dendro-afk is now known as dendrobates
smoseryeah, i was granted ppu14:41
smoseri searched that list and didn't find myself, but didn't know if i would have been in a sub group14:41
cjwatsonright, I'll sort that out now, thanks for the note14:41
smoserthank you cjwatson14:41
cjwatsonsmoser: should work for you now14:43
smosercjwatson, it does. thanks.14:44
nigelbstefanlsd: ping15:03
=== sconklin-gone is now known as sconklin
slangasekdmart: there are various race conditions with statd startup in lucid, currently.  Do you have /var on a separate partition?15:16
dmartslangasek: I don't think /var is on a separate partition, but is may be experiencing the race problems you refer to: see https://bugs.launchpad.net/ubuntu/+source/nfs-utils/+bug/58194115:19
ubottuUbuntu bug 581941 in nfs-utils (Ubuntu) "statd does not start automatically when needed nor can be forced to start on boot" [Undecided,Incomplete]15:19
=== SpamapS_ is now known as SpamapS
slangasekdmart: right - in that case that's the known fallout from the preceding portmap startup change; discussed in bug #525154, though that's now something of a catch-all bug for statd startup issues15:21
ubottuLaunchpad bug 525154 in nfs-utils (Ubuntu) "mountall for /var races with rpc.statd" [Undecided,Triaged] https://launchpad.net/bugs/52515415:21
dmartI guess I can work round this one manually for now15:23
=== plars-away is now known as plars
slangasekdmart: setting statd to 'start on local-filesystems' probably fixes it for you, if you're not on nfsroot15:24
slangasekI've just been slow to make that change in lucid because the code is touchy and I haven't had a solid block of time to think all the way through it15:25
dmartOK, I'll try it.  The box in question is a server; my main problem is huge timeout delays when clients do locking calls15:25
slangasekwell, /separately/, there's a race condition with nfs-kernel-server startup because it's not converted to upstart yet15:27
dokocjwatson: about the cpuname change: I think it's the correct fix. the triplet is used to be passed to configure, and it should be the correct one.  it easier to grep the rules files for wrong conditionals.15:27
cjwatsondoko: is there any way to get this merged to Debian in advance of Debian switching gcc's defaults?15:28
cjwatsone.g. can we add that cputable line as a separate line?15:28
cjwatson(in dpkg)15:28
dokobut: currently we are using the triplet names as multiarch names (as for multilib). maybe this really should be changed to architecture names, maybe the linux names extended to something like i386-linux15:29
dokocjwatson: afaik this a table lookup, with the arch as index. don't know how this should work15:30
cjwatsonwell, I guess I can play with it15:31
cjwatsonI care because this is one of the two remaining unforwarded changes in dpkg - tantalisingly close to being able to sync it15:31
=== JontheEchidna is now known as another_jonathan
hungerIs maverick going to use gcc 4.5 or 4.4? What was decided at UDS?15:47
AquinaI read nothing about that in UWN and slo no statement about this in the video casts so far.15:51
AquinaI gess 4.5 is possible however.15:51
hungerAquina: Thanks. I only found a mail from pitti (pre-UDS) stating that the default toolchain is still set to 4.4 at that time since gcc version was a UDS decission.15:52
=== deryck is now known as deryck[lunch]
cjwatsonright.  new dpkg heading for maverick.  hold on to your hats.15:59
* ogra grabs the rope16:02
ionAny hope dpkg-multiarch will make it into maverick? Anyone working on it?16:02
=== mathiaz_ is now known as mathiaz
cjwatsonion: slangasek has been making noises about working on it this cycle16:04
ionNice16:04
cody-somervilleugh, getting kubuntu blueprint changes for some reason16:15
Riddellprobably because kubuntu-dev is the assignee and ubuntu-core-dev is a member of that16:18
tremoluxrickspencer3: I have more data for the bug report btw16:20
rickspencer3nice16:20
rickspencer3tremolux, I'd like us to know what is going on so that if we see that problem again, folks know what's up and how to work around it16:21
tremoluxrickspencer3: unfortunately, it's a weird error16:21
tremoluxrickspencer3: yes, we need a report, I'll write it and add what I found out16:21
rickspencer3something about auth or keys or something?16:22
tremoluxrickspencer3: yes, something seemed wrong there16:22
tremoluxrickspencer3: plz try to find out what happened to the ppa if you can16:22
rickspencer3he deleted it16:22
tremoluxrickspencer3: oh, bummer16:22
rickspencer3he said he added it a long time ago, and the keys were misbehaving or something16:22
rickspencer3he's in #quickly if you want to chat to him directly16:23
tremoluxrickspencer3: that sounds like what the problem then, yeah16:23
rickspencer3tremolux, but I would like software-center to handle the problem if it's at all likely to happen again to others16:23
tremoluxrickspencer3: agreed16:24
rickspencer3of course if it's a rare circumstance, prioritize appropriately16:24
tremoluxrickspencer3: should be very rare, yep16:24
rickspencer3in that case, meh16:25
rickspencer3it's not like there was data loss or crashes, etc...16:25
tremoluxrickspencer3: I'll log it, it's important to keep track of16:26
rickspencer3k16:26
rickspencer3thanks16:26
=== jj-afk is now known as jjohansen
=== deryck[lunch] is now known as deryck
=== Chex_ is now known as Chex
nucc1hi i'm a novice programmer, i'm trying to track down an Evolution bug, can anyone help me interpret this error message? its only 3 lines. http://pastie.org/96960317:13
=== yofel_ is now known as yofel
cjwatsonthe first two lines are probably irrelevant, and the third means that the program attempted an invalid memory access17:16
cjwatsonhttp://en.wikipedia.org/wiki/Segmentation_fault17:16
cjwatsonyou'll need to use further tools to find out why: evolution may have its own debugging options to emit more verbose output, or else use tools such as strace, gdb, or valgrind17:17
nucc1thanks, i'll look into using a debugger. evolution doesn't seem to have any option for getting more verbose debug output17:21
=== beuno is now known as beuno-lunch
cjwatsonnucc1: #ubuntu-desktop may have more detailed experience here17:25
nucc1ok. trying to use gdb right now17:27
tekknokratHi, I need a workaround for /usr/share/python/python.mk for backporting trac-0.11.7 for hardy17:28
tekknokratIs this the right place to ask?17:29
Drakesonwould someone please remove the extra space between "]" and "<" (in "...] </default>") from line 121 of /usr/share/gconf/schemas/control-center.schemas, in package capplets-data ?  every upgrade keeps nagging about not being able to parse that.17:29
AquinaI would check the package in Launchpad and report it there.17:30
nucc1cjwatson, gdb says the segfault comes from libgtkhtml-editor17:31
Drakesoncjwatson: upgrade evolution17:31
Drakesonoops!17:31
Drakesonnucc1: upgrade the evolution package.17:31
nucc1Drakeson, i think im uptodate. trying to check if i somehow ended up with an odd version of libgtkhtml17:32
Drakesonnucc1: what is the version of evolution that you have? older than 2.30.1.2-1ubuntu2 ?17:34
nucc1i'm using 2.28.3, ubuntu lucid default17:35
Drakesonare you on maverick?17:37
ScottKMost people won't be at this point.17:38
ogra_cmpcone would hope so17:38
Drakesonsorry, that was meant for nucc117:39
Drakesonnucc1: are you on maverick?17:39
Drakesonnucc1: are you on maverick?17:39
nucc1i'm on lucid. trying to trace random evolution crashes.17:40
nucc1looks like i had libgtkhtml-editor 3.30 as opposed to 3.29 that ships with lucid17:41
Drakesona few days ago evolution got broken (gtkhtml-editor got upgraded, but evolution was blocked because evolution-exchange was holding it back). thought that might be your problem ...17:42
nucc1thanks, i must have entered the wrong channel. i was thinking that 'ubuntu-devel' means 'ubuntu developers'17:43
nucc1now i realise it means 'development version'17:43
ogra_cmpcwell the development version i swhat the ubuntu developers develop :)17:44
ogra_cmpcso you are actually right here ...17:45
ogra_cmpcfor your particular evo prob #ubuntu-desktop would be the better channel though17:45
nucc1ogra_cmpc, but that usually means they are working on the current development version, whereas i'm interested in something to do with current stable :)17:45
nucc1yea, i've been told. just stayed here cos of the inertia17:45
ogra_cmpcwe're also intrested in breakage in the recently released version in case its severe17:46
ogra_cmpcif you find a bad enough bug in the stable relese there is always the possibility for a stable release update to fix it17:47
nucc1ah, yes, i'll stay on it. if i find it's worth escalating, i'll find out how and do it17:50
=== oubiwann_ is now known as oubiwann
=== another_jonathan is now known as JontheEchidna
=== RoAk is now known as RoAkSoAxB
=== beuno-lunch is now known as beuno
Sarvattdarn, llvm-dev isn't installable in maverick at the moment :( depends on oprofile which isn't installable because - Depends: binutils (< 2.20.2) but 2.20.51.20100518-1ubuntu1 is to be installed19:33
cjwatsondoko: ^- can Sarvatt's problem be fixed by a simple rebuild?19:46
ccheneypitti: i noticed some users can't seem to get apport-collect to work for openoffice.org it is complaining the package 'openoffice.org' isn't installed but that is just the source package name (and a generally non-installed binary meta package)20:15
pitticcheney: right, it works against binary packages; otherwise it cannot collect information about the binary package, since it doesn't know which one20:15
ccheneypitti: it appears for some unknown reason to me that the binary package hint for openoffice.org packages is usually the openoffice.org meta package instead of something like openoffice.org-common or -core that is always installed, do you know how that ends up getting added to the bug report?20:17
ccheneypitti: and i am assuming the line "Binary package hint: openoffice.org" is what apport-collect looks at to see what to lookup?20:20
pitticcheney: I actually don't know what this "binary package hint" is or where it comes from; apport doesn't use it20:20
ccheneypitti: oh ok, so how does it determine the binary package name when launchpad only knows about source package names?20:21
ccheneypitti: is there some file to modify to do the proper mappings?20:21
pitticcheney: you mean for retracing? It's in the "Package:" field of the apport data20:21
pittii. e. the one you report the bug against20:22
ccheneypitti: er no when someone files a bad lp bug and i tell them run eg: apport-collect bugnumber20:22
ccheneypitti: for several people so far it has returned that they don't have the openoffice.org (meta) package installed, which is generally never installed20:22
pitticcheney: ah, for this we need an apport hook which also covers source packages20:23
pitticcheney: it tries to collect information about the binary package of the same name, but if that doesn't exists (or rather, isn't installed), it's simply not collected20:23
pitticcheney: then it only collects the info from the source_<srcpackagename>.py hook20:23
ccheneypitti: if i don't want to do anything special other than get the normal apport reported info what would i do to fix it for that?20:24
pitticcheney: for apport-collect? or for bug filing?20:26
ccheneypitti: for apport-collect i suppose, the bug filing probably should be fixed to but isn't as urgent20:26
pitticcheney: hm, does OO.o have a package hook? I don't have one here20:28
ccheneypitti: no it doesn't have a package hook, i was just wanting them to add the standard apport info that all packages get via apport-collect to an old bug filed without apport20:28
ccheneypitti: there isn't too much that i would need in a hook, maybe what video driver they are using, but the standard info is usually all i ever need20:29
pitticcheney: you could ask people to run apport-collect -p openoffice.org-draw 12345, for example20:29
pittii. e. against the application that they actually have a problem with20:30
ccheneyah ok so the -p is still useful, i thought it was deprecated and not actually used anymore20:30
pittifor apport-collect it's still useful20:30
pittiit's not necessary any more for ubuntu-bug20:31
ccheneyoh ok20:31
=== sconklin is now known as sconklin-away
ccheneypitti: thanks for the help20:32
pittiyou're welcome20:32
hendryi'm deploying a commericial binary package on Ubuntu and I'm wondering how I should create a shortcut on the user's desktop. Any advice here?20:45
juliankhendry: According to the topic, you should ask in #ubuntu-app-devel. But anyway, you should not create stuff on the desktop.20:57
juliankWhich is impossible on installation time anyway, since installation is run as root; and not as the normal user who wants to run the application20:58
=== apachelogger_ is now known as darthvader
Sarvattcjwatson: changing the oprofile Build-Depends: to binutils (>= 2.20.51.20100518-1ubuntu1) and rebuilding certainly seems to work so far as to letting llvm-dev install at least :D21:15
=== darthvader is now known as apachelogger
=== rgreening_ is now known as rgreening
=== blueyed_ is now known as blueyed
=== mok0_ is now known as mok0
=== sconklin-away is now known as sconklin

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!