/srv/irclogs.ubuntu.com/2012/07/17/#ubuntu-devel.txt

=== PaulW2U_ is now known as G4MBY
ScottKNow here's a descriptive changelog entry for you: "Bugger off cdbs, may it burn in hell."00:23
mdeslaurlol00:24
\sha good one ;)00:24
=== cpg is now known as cpg|away
slangasekCloses: all the bugs, without a doubt00:25
=== cpg|away is now known as cpg
robert_ancellanyone know how to disable make check in dh7?00:40
ionI’d try “override_dh_auto_test:” first.00:41
slangasekyes00:42
slangasekif you want it disabled at the package level00:42
slangasekif you want to disable it temporarily for a local build, DEB_BUILD_OPTIONS=nocheck00:42
robert_ancellthanks00:42
slangasekbut why are you disabling tests? :)00:42
robert_ancellslangasek, libsecret tests aren't running for some reason when building with debuild, but do work fine when using local build00:43
slangasekmissing build dep?00:43
robert_ancellno, this is building locally in both cases (one using bzr-buildpacakge, one in git)00:43
robert_ancelland I think at least one test required $DISPLAY00:44
slangasekah00:44
slangasekideally 'xvfb-run make test' then00:45
slangaseker, check00:45
robert_ancellbut too many things are blocking on it now, so really need to get it working00:45
* slangasek nods00:45
* slangasek reads about libsecret. Why is upstream replacing gnome-keyring? :P00:45
robert_ancellit's a rewrite and it's no longer gnome specific00:46
robert_ancellsame authors00:46
slangasekhmm00:46
slangaseksame interfaces? :)00:46
robert_ancellnope00:46
slangasekbleh00:46
robert_ancellI think it's an opportunity to learn from past mistakes00:46
slangasekwell, I guess the set of revdeps is manageable anyway00:46
=== cpg is now known as cpg|away
RAOFYou might also need dbus-test-runner, which someone should get to Debian, too.00:47
robert_ancellbtw, does anyone know the correct way to make GIR files install into /usr/lib/girepository-1.0 when the libraries are multi-arch?  Or should gir files be multiarch too>00:47
robert_ancell?00:47
slangasekideally they should be in the multiarch dir00:48
slangaseksince they are arch-specific and you may want them for more than one arch at a time00:48
robert_ancellslangasek, does that work? There don't seem to be any other packages doing that00:49
slangasekrobert_ancell: there's precisely one package doing this currently: gir1.2-guestfs-1.000:49
slangasekdunno if it works00:49
slangasekmost don't because of the gir bits being from a different, non-multiarched source package from the C lib00:49
roaksoaxis quantal archive frozen or anything?01:26
ScottKroaksoax: Not officially.  We're trying to collect a stack of packages to accept all at once for scalability testing.  see cjwatson's mail on freezing to Ubuntu devel.02:06
micahgbenc: BTW, I uploaded the PowerPC fixes for lightning to our staging PPA, I'm waiting on the branch that quantal is using to upload that one02:50
BenCmicahg: I saw it built successfully, glad I could help02:51
micahgbenc: thank you, that means we get coverage across the board 6 weeks early02:51
micahgwell, except for the broken neon on arm*...02:51
=== cpg|away is now known as cpg
infinitymicahg: Which broken neon on arm?03:08
infinityScottK: We're not trying to get a stack of packages to do at once, just to do lots of different packages via the CLI tool to see if anything explodes or times out.03:09
micahginfinity: the thing you fixed in quantal03:09
micahgjust having it enabled at build time03:09
infinitymicahg: Well, sure, but I fixed it, so I was curious about the "still broken" bit.03:09
ScottKinfinity: If we aren't we should.  When I had trouble with the web UI, it was mostly with multiple accepts.  Very rarely was it a single package.03:10
infinity(Well, "fixed"... With a bit of a large hammer)03:10
micahginfinity: well, I meant in the stable releases it'll be broken unless we need a respin03:10
micahgor until th next train03:10
infinityScottK: If you have to slim down to a single package here and there, such is life.  At least that's trivial with the CLI (queue info | mangle output | xargs queue accept)03:10
infinityScottK: The concern is that there could be single packages that will end our world.03:11
micahgbut otherwise, we have all archs back (well, except for sparc/ia64, but who's running a lucid desktop on that these days anyways)03:11
ScottKinfinity: Yes, but it varies.  If you can accept them in large stacks, the odds you get screwed by one go down as the only times I ever had problems with specific individual packages were also times I was constrained by the number it would do before timeout.03:12
infinityScottK: *shrug*03:12
ScottKAnd doing them one at a time you never know if you tried one of the 'harder' ones or not.03:12
infinityScottK: I'm literally just blindly batch-accepting what's in unapproved occasionally.03:12
ScottKOK.03:12
* ScottK is trying to kill off Debian RC bugs tonight. Got 5 so far today.03:14
=== cpg is now known as cpg|away
BenCinfinity: How do orphaned packages get dropped from the archive?03:30
BenCe.g. kde-runtime no longer creates kde-runtime-dev, but it's still a valid package (although uninstallable)03:30
BenCat least one thing build-deps on it (plasma-mobile), but of course, it can't be built now (I'm planning on fixing the build-deps after I verify it)03:31
pittiGood morning03:36
infinityBenC: I've been waiting on the last merger of plasma-mobile to package up a new upstream snapshot that makes it sane and buildable again.03:40
infinityBenC: We drop NBS packages when nothing is left depending on them.03:40
BenCinfinity: Ok, then I'll leave plasma-mobile till then03:41
infinityBenC: Yeah, changing build-deps won't be enough, already been there. ;)03:41
infinityBenC: Have you met http://people.canonical.com/~ubuntu-archive/nbs.html ?03:42
BenCNo, but thanks03:43
BenCMore lengthy than I would have guessed03:43
infinityIt'll start shrinking rapidly, now that Debian autosyncs are off.03:43
* micahg wonders when we'll get our first rebuild03:46
infinitydoko wanted to do one late last week.  He must have forgotten in the excitement of kayaking and bodysurfing.03:46
micahggreat, I'm curious to see what havoc gcc-4.7 and the various new upstream imported have wreaked on the rest of the archive03:48
infinitygcc-4.7 shouldn't be too bad, in that rebuilds were done already and bugs filed.03:49
infinityBut I'm sure some where missed.03:49
infinityI think getting the NBS list down before a rebuild is sane anyway.03:49
infinitySince lots of stuff is FTBFS right now with broken or out-of-date build-deps.03:50
micahgwell, we have --as-needed by default :)03:50
micahgalthough, that's old news03:50
infinityI've not noticed any weird interactions between 4.7 and as-needed... Have you?03:50
infinitySeems to be business as usual.03:50
micahgno, not really, which means it's about time for me to go to sleep :)03:50
infinityToodles.03:50
=== cpg|away is now known as cpg
=== cpg is now known as cpg|away
pitticjwatson: just accepted two pacakges from quantal/unapproved using the local queue tool; great work on this!05:07
=== cpg|away is now known as cpg
=== cpg is now known as cpg|away
=== cpg|away is now known as cpg
infinityrbelem: Did anything ever come of the plan to update plasma-mobile in quantal with a newer version?06:59
=== philipballew__ is now known as philballew
cjwatsonScottK,infinity: Lots of packages at once will definitely not be a problem with the API client.  The timeout is applied per request, and acceptFromQueue is a method on individual PackageUpload objects, so the client always (has to) accept one at a time regardless of how many you give it on the command line.07:15
cjwatsonScottK: I know skaet occasionally reported problems with even accepting a single package using the web UI.07:16
cjwatsonBut it's true that lots at once greatly increased the odds.07:17
dholbachpitti, weird - now booting without 'quiet splash' seems to have worked *shrug*07:19
pittidholbach: yeah, it always works for me as well07:20
pittisomething weird with the plymouth integration of encrypted swap devices or so07:20
dholbachbut it got stuck 3 times before07:20
dholbachmaybe it's just intermittent07:20
bambeehi, http://paste.ubuntu.com/1096244/  ^^08:16
=== cpg is now known as cpg|away
seb128cjwatson, hey, do you know if that ubiquity issue is know,being worked? (that's one of the most reported issues on errors.ubuntu.com)?10:27
seb128https://errors.ubuntu.com/bucket/?id=/usr/lib/ubiquity/bin/ubiquity:ValueError:watch_debconf_fd_helper:process_input:wait:cleanup:preseed:%3Clambda%3E:command10:27
seb128cjwatson, I find some similar looking bugs in launchpad bug I'm not sure they are exact matches10:28
seb128cjwatson, like the title on bug #986804 matches but the bug has hardly any detail10:32
ubottuLaunchpad bug 986804 in ubiquity (Ubuntu) "ubiquity crashed with ValueError in command(): I/O operation on closed file" [Undecided,New] https://launchpad.net/bugs/98680410:32
=== jibel_ is now known as jibel
pittiseb128: do you know what's the yellow bits on http://people.canonical.com/~ubuntu-archive/pending-sru.html ?10:45
pittithat seems to be new, and not documented in the header10:45
Laneybdmurray explained in #-release10:46
Laneyit means "new comments since v-done" IIRC10:46
seb128pitti, I was wondering the same thing today10:47
pittioh, I see the commit10:47
seb128"  sru-report: as a first draft change the bug color to yellow if it's date_last_message is later than the publishing date of the package10:47
seb128"10:47
pitti  sru-report: as a first draft change the bug color to yellow if it's date_last_message is later than the publishing date of the package10:47
seb128snap :p10:47
seb128I like that they are sorted by time in proposed btw10:48
pittihm, why would that need date of last message? if someone flips it to v-done, it'd still be yellow?10:48
seb128pitti, no, v-done are green10:48
pittiah, so that means "a bug which should be inspected by the SRU team"10:48
pittiperhaps as a replacement for reading bug mail10:48
seb128pitti, it's to see which ones might need an update, i.e are still verification-needed but got comments10:48
seb128it might be that a commenter verified or reported a regression but didn't know to tg10:49
seb128tag10:49
seb128pitti, I guess so10:49
RiddellLaney: I'm getting a curious build error when compiling against libxml211:22
Riddellhttp://paste.kde.org/518930/11:22
RiddellLaney: if I rebuild libxml2 it goes away11:22
Riddelldo you recon it would work if I just rebuilt it in the archive?11:23
Laneyerm11:23
Laneydid zlib break abi?11:23
=== MacSlow is now known as MacSlow|lunch
RiddellLaney: it has had multiarch things going on11:26
Laneyhmm11:29
Laneylibxml2 was uploaded after zlib11:29
LaneyRiddell: is your chroot fully updated?11:34
RiddellLaney: should be but let me try again11:36
Laneyhttp://bugs.debian.org/cgi-bin/bugreport.cgi?bug=67329711:38
ubottuDebian bug 673297 in libxml2 "mipsel: /usr/lib/mipsel-linux-gnu/libxml2.so.2: undefined reference to `gzopen64@ZLIB_1.2.3.3'" [Critical,Open]11:38
RiddellLaney: that fixes it, thanks :)12:38
Laney;-)12:38
ogra_pitti, pingaling12:41
ogra_pitti, i'm testing ac100 images atm, while running "ubuntu-drivers list" gets me the proper return value (nvidiy-tegra), i dont get anything in the drivers gui12:41
pittiogra_: pongedidong12:41
ogra_any hint where i should look ?12:42
pittiogra_: we don't have a drivers guy yet, it's still in a branch of didrocks'/cyphermox'12:42
ogra_oh, so what i see it the old jockey12:42
pittiyes, that'll go away12:42
ogra_k, that explains also why it works on the pandaboard :)12:42
pittiogra_: "drivers guy" argh, "driver GUI"12:43
ogra_(we didnt have a jockey hook for ac100, but for omap4)12:43
ogra_pitti, thanks a lot12:43
pittiI'm currently working on the backend API, then didrocks will have it finished in no time, I'm sure12:43
pittiI hope we can land it for a312:43
ogra_yeah, yesterday ...12:43
ogra_he is did*ROCKS* after all :)12:43
pitti+112:44
didrocksogra_: pitti is doing all the hard work, don't trust this free ad ;)12:44
ogra_didrocks, verbally bluching wont help, you still rock ;)12:45
ogra_*blushing12:45
didrocksheh :-)12:45
pittididrocks: system_device_drivers() just pushed; feel free to use from git (just set PYTHONPATH) if you want to play with it already12:45
didrockspitti: \o/ thanks a lot!12:46
* didrocks pulls12:46
pittididrocks: I'll add a new mode to ubuntu-drivers ("by-device") so that you can use it in textual form12:46
didrockspitti: that will be really handy :)12:46
ogra_btw, since i just built a new machine here and added two graphics adapters ... if someone would be insane enough to add radeon and nvidia cards at the same time to the same machine, can we finally handle that case ?12:47
ogra_(iirc there were issues with the GL libs, i was wondering if multiarch saves us from that)12:48
pittiogra_: /usr/share/ubuntu-drivers-common/fake-devices-wrapper <command> (e. g. ubuntu-drivers or didrocks' GUI), and you'll have both of them :)12:48
ogra_well, would they also *run* ?12:48
pittibut yes, nvidia still replaces the system's GL libray; I woudl have thought that tseliot's alternatives handling might save us here, though12:48
ogra_ah, yeah12:49
pittiyou can't run them both at the same time, I'd guess12:49
pittibut at least switch between the two without changing packages12:49
ogra_yup12:49
didrockspitti: ah, http://paste.ubuntu.com/1096566/ using the wrapper12:52
pittididrocks: in case you pulled already, please pull -f12:52
pittididrocks: yup, that's what I just fixed, sorry about that12:53
didrocksno worry :)12:53
pittididrocks: fglrx isn't being tested in the test suite, bad me :/12:53
didrockspitti: you rewrote history, how dare you! (j/k)12:53
pittididrocks: but presumably I just did the very thing you tried12:53
didrocksheh, seems so :)12:53
=== _salem is now known as salem_
pittiyeah, I prefer consistent commits, as long as the previous push hasn't been too long ago12:53
didrockspitti: was just kidding :)12:53
* pitti hugs didrocks12:54
* didrocks hugs pitti. Thanks for your good work!12:54
didrockspitti: looks perfect to me, but yeah, some case in the wrapper where you create a fake "manual_install" one will be good. Will fake that into the my system answer for now12:58
=== MacSlow|lunch is now known as MacSlow
pittididrocks: any suggestions? http://paste.ubuntu.com/1096582/13:08
didrockspitti: you don't have an example from not comming from distro as far as I can see (but this is linked to the manual_install discussion we had I guess). Otherwise, the needed info seems there! Great :)13:12
didrockspitti: no ascii orders for attributes?13:12
pittiright, no support for the manual flag yet13:12
pittididrocks: sorry, ascii orders?13:12
didrockspitti: like "builtin distro free" instead of "distro free builtin"13:13
pittididrocks: it's a pretty arbitrary order, but I can change it around if you want; but it's just a debugging tool anyway13:14
didrockspitti: right, I was wondering if later we want to do some platform tests using it13:15
pittididrocks: oh, it's a predictable (i. e. stable) order13:15
pittiso you can string-match it if you want13:15
didrocksok, as long as we have examples on the order, that's fine :)13:16
brendandwhy is user /usr/bin/python still pointing to 2.7 in quantal? i thought that was fixed?13:19
tseliotogra_, pitti: you can do that as long as you use open drivers for both cards ;)13:19
pittibrendand: fixed how? "python" will always be python 2.x13:20
pittipython3 is 3.x13:20
ogra_tseliot, lol, well, then i wouldnt use ubuntu-drivers i guess13:20
brendandpitti - ok. maybe a different issue then. but apport-cli points to python so is not running on quantal13:21
pittibrendand: oh, bug13:22
brendandpitti, right. so the approach is that the applications need to change their shebang, rather than point /usr/bin/python at 3.2?13:23
pittiyes, as soon as they got ported to py313:23
pittibrendand: I made a note to fix the shebang of apport-cli13:23
ScottKstgraber: FYI, the partman-target piece of Bug #992241 is done (in -updates) and I copied the pending Ubiquity SRU as well, so the path is clear to do the Ubiquity part of that one.13:26
ubottuLaunchpad bug 992241 in ubiquity (Ubuntu Precise) "Upgrading using the live cd wipes /var/lib/mythtv/*" [High,Triaged] https://launchpad.net/bugs/99224113:26
=== zyga is now known as zyga-afk
stgraberScottK: thanks. I'll prepare an ubiquity upload today.13:39
bdrungdholbach: sponsors-queue crashes on my system with: lazr.restfulclient.errors.ServerError: HTTP Error 503: Service Unavailable13:47
bdrungdholbach: does it work on your system?13:47
dholbachlet me see13:48
dholbachbdrung, still running13:52
dholbachbdrung, it works13:56
LambdaDuskhi, I am trying to develop some graphics app, and somehow I get errors when trying to open a window with GLFW and GLUT... GLFW tells me nothing, but GLUT says something about a framebuffer missing13:57
bdrungdholbach: it's reproducable here: http://paste.debian.net/179474/14:03
dholbachbdrung, I won't have the time to look into it, I'm afraid14:03
DebolazLambdaDusk: It seems #ubuntu-app-devel might be a better place to ask.14:29
=== zyga-afk is now known as zyga
BenCLaney: Any idea when ghc-7.4.2 is coming?15:13
Laney17/07 12:28:02 <erikde> oh, ok. will do.15:13
Laney17/07 12:28:37 <erikde> maybe not today, bit busy. should be over the next couple of days.15:13
BenCIt appears to fix a test suite failure in cryptocipher on ppc15:13
BenCOk, thanks15:14
xnoxany idea or bug # when theming aka square radio buttons will be fixed?15:30
xnoxseb128: or somebody else? ^^^15:30
seb128xnox, no idea, I didn't even know they were broken15:31
seb128kenvandine, ^15:31
xnoxseb128: are you using quantal? =)))15:31
kenvandinethey are a little ugly, not sure if there is a bug filed for it15:31
seb128xnox, no, I'm on the 12.04.1 team15:31
mterrysladen, you have some unpushed changes in ubuntu-mono.  Is trunk  OK to push to quantal?  (I have a imagemagick4->5 transition that I wanted to push)15:31
kenvandinethe theme needs quite a bit of love, which hopefully will happen real soon15:32
kenvandinexnox, wouldn't hurt to file a bug15:32
* ogra_ has seen this being discussed here several times15:32
ogra_would be surprising if there wasnt a bug already15:32
xnoxkenvandine: it looks as if, it's the fallout from theming brake upstream15:33
kenvandinexnox, yeah15:33
xnoxogra_: kenvandine: it's just I want to subscribe to know when it will be fixed "for real"15:33
kenvandinethere are quite a few things broken because of upstream changes and our theme not keeping up15:33
stgraberyeah, highvoltage mentioned it once and mpt tracked it down to a gtk regression IIRC, I seem to remember an LP bug being mentioned15:33
mptiz gtk boog15:34
xnoxstgraber: was it something like "invalid network-manager & ubiquity, fix-release light-themes?"15:34
xnoxmpt: lolz =)15:34
xnoxmpt: do you have a bug number on a square post-it note at your desk by any chance? =)15:34
mptxnox, if anyone knows, it's Cimi15:34
mptI just asked him and he waved his hand jedi-like and said "everything will be fixed"15:35
highvoltagexnox ftw [ funny ]15:35
xnoxbug 101549715:35
ubottuLaunchpad bug 1015497 in gtk+3.0 (Ubuntu Quantal) "Cannot select gtk dropdown list value in multiple applications: ubiquity, landscape client" [High,Triaged] https://launchpad.net/bugs/101549715:35
xnoxis the closest I can find15:36
smoserkenvandine, can you sort out https://bugs.launchpad.net/ubuntu/+source/light-themes/+bug/76216715:37
ubottuLaunchpad bug 762167 in light-themes (Ubuntu) "missing dependency on gtk2-engines-pixbuf" [Low,Triaged]15:37
sladenmterry: to the best of my knowledge, yes15:37
xnoxbug 102409915:37
ubottuLaunchpad bug 1024099 in light-themes (Ubuntu) "Widgets look square" [Undecided,In progress] https://launchpad.net/bugs/102409915:37
mterrysladen, cool, thanks15:37
sladenmterry: try it and see, and if not, poke me again and I'll do some more investigation about what it/was15:37
kenvandinesmoser, again...15:38
mterry:)15:38
kenvandinesmoser, yes.,.. i will15:39
smoserkenvandine, i dont think you ever uploaded. i pinged you once after you uploaded without the fix, but i'm sure it just got lost in the hussle.15:39
smoserits partially my fault for not just doing an upload, but pushing to lp:ubuntu/ branch15:39
kenvandinesmoser, no i am sure i did upload it15:39
smoserhm..15:40
smoserah.15:40
smoserthen the bug is just probalby in the wrong state.15:40
kenvandinebut the packaging branch got moved around a couple times15:40
smoser(i see it just got moved from fix released to triaged)15:40
kenvandineso it might have gotten accidentally reverted15:40
kenvandinesmoser, oh... we fixed it by not needing it15:42
kenvandine    - Remove the need for gtk2-engines-pixbuf in the gtk2 theme,15:42
kenvandine      saves some CD space15:42
kenvandinesmoser, so if it is still actually needed, it is a bug in the theme15:47
smoserkenvandine, ok. good enough. then i think the users of that bug are generally just confused.15:48
smoserso do you believe this to be fixed in precise?15:48
kenvandineyes15:48
kenvandinethat version was in april15:48
smoserif so, can you please comment on that in the bug.15:48
kenvandinei just did15:48
smosergreat. thank you sir.15:48
kenvandinenp15:48
smosersorry for the fire drill15:48
kenvandineno worries15:48
kenvandine:)15:48
kenvandineif it is still printing the warnings then the theme needs fixing15:49
=== dendrobates is now known as dendro-afk
keesjodh: any plans to add seccomp-bpf to upstart? :)15:56
zulcan an archive admin review python-cliff, python-django-compressor, python-django-appconf, and python-tablib please16:06
=== beuno is now known as beuno-lunch
jodhkees: no plans this cycle. I wasn't even aware of this feature: although the kernel supports it, the include files and man pages don't appear to cover it yet.16:14
=== beuno-lunch is now known as beuno
keesjodh: it's in 12.04 via a backport. it's all in 3.5.16:17
=== jbicha is now known as Guest50245
maxLimitGood evening16:53
nanderssondidrocks, hi, I am having a look at your project session-migration. Is there  a way to try that out in Ubuntu 12.04?17:05
nanderssondidrocks, I guess I could compile it and package it?17:06
didrocksnandersson: just retake the same packaging, it should work17:06
nanderssondidrocks, Ok, thanks!17:07
didrocksyw :)17:07
maxLimiti tried to import totem source code to eclipse but i get an error message:17:09
maxLimitmake: *** No rule to make target `all'.  Stop.17:09
maxLimitthe message appears after building in eclipse17:09
=== shadeslayer_ is now known as evilshadeslayer
=== yofel_ is now known as yofel
=== Quintasan_ is now known as Quintasan
hippiehackerI'm off and on, I sent you an email 8)17:28
highvoltageogra_: hey ogra_, so you're officially ogra underscore now? :)17:32
ogra_highvoltage, i'm ogra underscore since hmm, three years now ...17:37
* highvoltage takes long to adjust to these things17:37
ogra_since i started using a proxy and got to lazy to fight with nickserv17:37
infinityogra_: Three year is some serious lazy.17:49
infinityogra_: Maybe you should just fix it? :)17:49
alazare619how do you add a ppa to a chroot build phase with livecd-rootfs?17:51
ogra_infinity, i will if i switch to a better proxy :)17:52
barry@pilot in17:54
=== udevbot changed the topic of #ubuntu-devel to: Quantal Quetzal A2 released! | Archive: open | Dev' of Ubuntu (not support or app devel) | build failures -> http://bit.ly/HaWdtw | #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: barry
=== cpg|away is now known as cpg
=== Guest50245 is now known as jbicha
=== jbicha is now known as Guest72435
=== Guest72435 is now known as jbicha_
=== kentb-out is now known as kentb
scientesindirect linking isn't working for me with multiarch cross-building http://paste.debian.net/179495/18:22
jtaylorscientes: kytokabinet is underlinked, either link it correctly or disable as-needed18:23
scientesoh ok18:23
scienteswhy does it work natively then?18:24
jtayloris kyotocabinet also not linked against lzma natively?18:24
scientesit is18:24
scientesi mean, it is linked against lzma18:24
scientesalthough ldd doesn't work cause the binary is foreign18:25
jtayloryou need to do the same for the cross build18:25
infinityThat's easily fixable.18:25
scientesbut i copied it to a armel host, and ldd worked18:25
scientesyeah it seemed like something that the multi-arch stuf should handle18:25
scientesinfinity, what is easily fixable?18:25
infinityscientes: The ldd thing is on my TODO.  I had hoped to sort it for wheezy (and maybe I'll still try to squeeze it in)18:26
scientesdo i just need to link against everything, even that which is indirect?18:26
scientesok, and when ldd gets fixed it should work?18:26
infinityscientes: As for your linking issue, it shouldn't relate to the ldd thing, no.18:26
jtaylorscientes: you also need to disable as-needed18:26
infinityscientes: Could just be that the cross toolchain is lacking sane -Ls?18:26
scientesinfinity, is that in LDFLAGS, cause if tou see the paste i tried adding those18:27
infinityscientes: I see a shocking lack of whitespace in -D_KC_APPLIBS="\"-L/usr/lib/arm-linux-gnueabi18:29
infinityIn fact, I'm not sure what that's meant to do. :P18:29
infinityOh, I see.  That's a big long list with the -lfoo as well.18:30
infinityI'm misreading that bit.18:30
scientesthat first one works18:30
infinityMisreading it as something that makes sense to pass to a compiler.18:30
infinityWhich that doesn't.18:30
scientesits when its indirectly linked it doesn't18:30
jtaylorthats because it isn't, kyto is not linked against lzma18:31
scientesjtaylor, but it IS18:31
jtayloroh18:31
scientesi copied to a a armel host and checked with ldd18:31
infinityscientes: Does the exact same command line also work on armel?18:32
jtaylorI misunderstood you before then18:32
scientesinfinity, good question18:32
jtaylorldd -r shows no undefined references?18:32
scienteshttp://paste.debian.net/179516/18:33
scientesinfinity, yes it does work on arm, and creates the kcutiltest binary, which works18:34
infinityCurious.18:36
infinityI'm inclined to blame the crosss toolchain, but not entirely sure what to blame it for doing off the top of my head.18:36
scienteswith g++-4.7 too18:36
scientesi had this problem with the emdebian squeeze g++-4.418:37
scientesthen compiled it according to instructions here: http://gsoc.sitedethib.com/posts/apt-get_install_gcc-4.7-arm-linux-gnueabihf/18:37
scientesand had the exact same problem18:37
infinityAt least it's consistent. ;)18:37
scientesand i subbed armel and eabi, for armhf stuff18:37
infinityIs this a source package I can test?  I suppose I could just do a minimal testcase, but that sounds like effort.18:38
scientesif you do DEB_BUILD_OPTIONS=nocheck the kyotocabinet from unstable/quntal will work18:40
scientesbut it doesn't even get that far, so it doesn't really matter18:40
jtaylorhave you tried without as-needed?18:40
scientesSteve Langasek on #multiarch on OFTC says its a known bug with ld18:49
infinityMmkay.18:49
slangasekhttps://wiki.linaro.org/Platform/DevPlatform/CrossCompile/UsingMultiArch, search on "DEB_LDFLAGS_APPEND"18:50
slangasekin particular, -Wl,-rpath-link=/usr/lib/arm-linux-gnueabi:/lib/arm-linux-gnueabi:/usr/lib is used to work around this wrong default in the cross-binutils package18:50
infinityThis seems easily fixable in the cross toolchain...18:51
slangaseknothing ever seems easily fixable in the cross toolchain, AFAICT :)18:51
smoserSpamapS, around?19:28
smoser /etc/init/cloud-init-local.conf is: start on mounted MOUNTPOINT=/19:29
smoseris that guaranteed to have /run or /dev/shm? access?19:29
SpamapSno19:29
SpamapSsmoser: virtual-filesystems would guarantee that, I think.19:30
SpamapSbut does not block19:30
smoservirtual-filesystems definitely would guarantee that.19:30
SpamapSI haven't looked in a while when MOUNTPOINT=/ is emitted19:30
smoseri think its not guaranteed.19:31
=== AlanChicken is now known as AlanBell
=== zyga is now known as zyga-afk
=== jbicha_ is now known as jbicha
ScottKWould someone who understands something about Java look at Bug 937710?  It appears that the SRU we just copied to -updates might not have fully resolved the bug.20:10
ubottuLaunchpad bug 937710 in visualvm (Ubuntu Precise) "[SRU] VisualVM does not start with openjdk 7" [Undecided,Fix released] https://launchpad.net/bugs/93771020:10
=== lfaraone_ is now known as lfaraone
dylan-mThanks for the detailed code review, barry! I'll be able to jump into that later this week, I hope :) Mpt filed a bug report about the height of the updates list earlier today: https://bugs.launchpad.net/ubuntu/+source/update-manager/+bug/102567421:26
ubottuLaunchpad bug 1025674 in update-manager (Ubuntu) "List of updates is much too short by default" [Undecided,New]21:26
dylan-mFixing it should be pretty painless. Just needs different defaults in dconf to start with. (Though some fancier design-related updates might lead to good places, too).21:26
barrydylan-m: sure thing!  thanks very much for the branch.  since there's a separate bug on the details height, don't feel obligated to fix that in your branch.  i.e. it could be fixed in a separate branch, but if the fix is easy and you want to include it in your updated branch, feel free to do so!  please do link your branch to that bug if you do though.21:28
barrydylan-m: and feel free to ping me if you have any questions or just disagree with me. :)21:28
=== salem_ is now known as _salem
YokoZarIs there a way to get the slightly older firefox 13 source packages off launchpad anywhere?  I want to put them in a PPA22:50
scientesYokoZar, debian has esr releases in wheezy, you could just rebuild those22:51
scienteswhich is firefox 1022:51
scientesotherwise you shouldn't run older versions of firefox, except for regression testing22:51
scientes*bisecting22:52
YokoZarscientes: right, that's what I want to do22:52
micahgYokoZar: why ,that's not recommended22:52
scientesYokoZar, just use mozregressions22:52
scientesits super awesome and downloads the daily builds from ftp.mozilla.org22:52
YokoZarmicahg: I have an environment where we have to do a good chunk of manual work to support each new firefox release22:52
scientesand they have them back for years22:52
micahgYokoZar: you want the ESR then22:53
YokoZarmicahg: the proper solution to this is to grab packages from our own repository and update it when firefox updates22:53
scientes^^^^^^^^^^^^^+1 for esr22:53
cjwatsonYokoZar: aside from whether it's a good idea or not, https://launchpad.net/ubuntu/+source/<source package name>/+publishinghistory for any source package has everything22:53
scientesYokoZar, precise runs firefox esr22:53
YokoZarcjwatson: ahh yes I was hoping for that :)22:53
scientesAFAIK22:53
micahgscientes: hrm? we just pushed out 1422:53
scientesoh yah22:53
scienteswhoops22:53
cjwatsonthough it might time out for a big package like firefox.  you can use https://launchpad.net/ubuntu/+source/<source package name>/<source package version> directly22:53
scientesmicahg, you should also package esr22:53
cjwatsonwell s/big/old/ I guess22:54
scientesoh wow, you push out ff14 to even lucid22:54
micahgYokoZar: as long as you're updating your own packaging to the latest release, or staying on the ESR, you're fine, otherwise, you're open to security issues (and even the ESR might have some, but hopefully not critical ones)22:54
micahgscientes: have you read http://www.chriscoulson.me.uk/blog/?p=11122:55
scientesmicahg, so that is the real reaon you do not build libxul as a shared library anymore?22:56
YokoZarmicahg: Yes, I know, keeping up to date is a good thing :D  But the immediate impact is it will break some very custom-specific scripts here until I can forward-port the changes22:56
scientesYokoZar, that is why you should run esr22:56
micahgYokoZar: we have builds available in PPAs for you to adjust your stuff in advance22:56
scientesmicahg, i run ff nightly ;), if you put a sym link to it from ~/bin/firefox then even the .desktop files work22:56
YokoZarscientes: I will migrate at the next ESR, at the moment I need some 12+ specific features22:58
micahgYokoZar: 12 weeks notice here: https://launchpad.net/~ubuntu-mozilla-daily/+archive/firefox-aurora, 6 here https://launchpad.net/~mozillateam/+archive/firefox-next/+packages if you grab the build after the migration (which happens shortly after release)22:59
StevenKRAOF: 'ubuntu-bug nvidia' no worky23:00
YokoZarNoted micahg :)23:01
RAOFStevenK: That's because there's no “nvidia” package; “ubuntu-bug nvidia-current” works, right?23:05
StevenKRAOF: Right, that does.23:06
StevenKRAOF: I've restarted X and removed the module, so I'm not sure filing a bug now is the right thing to do. :-)23:11
RAOFHeh.23:11
=== apachelogger_ is now known as apachelogger__

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