/srv/irclogs.ubuntu.com/2008/03/01/#ubuntu-devel.txt

jdongpersia: right, certainly blindly setting MAKEFLAGS=-j2 will cause FTBFS'es that wouldn't ordinarily happen in the buildds. I was wondering if there was a more intelligent flag for that00:02
jdongI recall someone saying that the buildd's do intelligently use multiple cores00:02
jdongs/cores/cpus/00:02
persiaMaybe by building multiple packages simultaneously in separate environments?00:02
persiaSome packages seem to try to calculate the number of cores in debian/rules build, and call make with -j.00:03
=== kitterma is now known as ScottK2
jdongpersia: alright, given that it doesn't seem like there's a clear-cut utopian answer, I'll stick with default build options and exercise patience :D00:05
persiajdong: Look to the future.  There've been several threads on this in debian-devel, and I think support ought hit policy sometime in the 3.7.x series.  After that, it's just a matter of testing and adjusting all the source packages, and you'll have an answer :)00:07
jdongpersia: awesome. Just how again do we test packages for makefile race conditions? :D00:07
jdong(very carefully)00:08
persiajdong: Last test I saw was a complete archive rebuild of unstable with a build system that preferred parallel builds.  debian/rules seems to be a significant offender for issues running in parallel.00:09
jdongpersia: ah, not at all surprising... hence why I was worried about globally setting MAKEFLAGS00:10
persiajdong: Check the debian ML archives.  I think Aurelien Jarno had something that parallelised everything except debian/rules at one point.00:11
jdongpersia: will do00:11
=== Martinp24 is now known as Martinp23
=== mdomsch_ is now known as mdomsch
=== vorian is now known as vorian_
sorenbdmurray: Around?03:00
jdong!ping03:03
ubotuping yourself ;-) really the diodes all down my left side are sore03:03
sorenslangasek: Still around? https://bugs.edge.launchpad.net/ubuntu/+source/kvm/+bug/196868 is still untouched, and I'm running out of kittens.04:16
ubotuLaunchpad bug 196868 in kvm "[ffe] Upgrade kvm from vers. 60 to 62" [Undecided,New]04:16
bdmurraysoren: am now04:53
sorenbdmurray: Whee.04:58
sorenbdmurray: I was about to look through the pile of hibernation related bugs to see if I might have fixed some of them by accident. Kees said that you might have all of them in your head, and that beats the crap out of Launchpad's search features.04:59
sorenbdmurray: So... Do you remember any bugs about machines seeming to hibernate properly, but failing to resume either always or just once in a while or something.05:00
soren?05:00
jdongsoren: I think that globs just about every suspend2disk failure case05:04
bdmurraysoren: gah, I feel like a failure since I don't have them all in my head05:05
sorenEr... Let me be a tiny bit more specific.05:06
sorenIt hibernates properly, but when it's supposed to start resuming, it just boots as though you hadn't hibernated.05:06
sorenjdong: better? :)05:06
=== steve__ is now known as vorian
jdongsoren: much :D05:07
bdmurrayThis doesn't have to do with the resume partition being wrong though?05:07
jdongthat certainly could cause those symptoms05:09
jdongIIRC there were some uuids-dont-resolve kind of bugs that cause these symptoms05:09
bdmurrayright, but soren indicated he fixed *something*.  I was wondering what that something was.05:09
sorenYeah. I just fixed one of those cases and instead of trawling through the boat load of bugs I just wanted to feed off of your brain.05:10
bdmurraywhich case did you fix?05:10
sorenWhat I fixed was a race in the resume script in the initramfs that would check for the swap partition's existence just a tiny bit before it actually showed up.05:10
sorenWell.. "fixed" is a bit strong :)05:12
jdong- sleep 205:12
jdong+ sleep 1005:12
bdmurrayheh05:12
jdonglike that?05:12
sorenAlmost, but not quite :)05:12
sorenIt didn't use to sleep at all, so I added a loop that checks for 5 seconds and if it hasn't showed up by then, it bails.05:13
jdonghow does an every-5-second check not use sleep? :)05:13
jdongbut more seriously, is there some API call in the init script "library" that performs this kind of thing?05:14
jdongit's useful in more than one place05:14
sorenOh, it does use sleep. It's just not stupidly extending an already existing sleep. :)05:14
sorenThere's not, but there should be.05:15
bdmurraysoren: bug 93039?05:16
ubotuLaunchpad bug 93039 in initramfs-tools "Regression: resume from disk (hibernate) fails sometimes" [Undecided,Confirmed] https://launchpad.net/bugs/9303905:16
sorenbdmurray: That's the one.05:20
sorenbdmurray: Awesome. Thanks.05:20
bdmurraysoren: no problem05:20
* superm1 wanders in and glares around at the present core-devs. free kittens for someone who is willing to sponsor a small debdiff :)05:21
fabbionesoren: just look at how you wait for root :)05:21
fabbionesuperm1: where is the debdiff?05:22
superm1bug 18745905:22
ubotuLaunchpad bug 187459 in cdrkit "genisoimage creates bad iso images if joliet is used - error: "Unexpected joliet directory length" appears" [High,Confirmed] https://launchpad.net/bugs/18745905:22
fabbionesuperm1: btw.. i am having issue with the latest mythtv dvd internal player. it can't play all dvd's while both xine and mplayer can do on the same05:22
sorenfabbione: Yeah, I got a bit of inspiration from that :)05:23
superm1fabbione, you have a bug filed on it?05:23
fabbionesuperm1: no, i noticed only about 8 hours ago.. before going to sleep05:23
superm1ah05:23
fabbionesuperm1: and i didn't track down the differences in the iso yet.. some play some don't05:23
fabbionesuperm1: so a bug would be very much incomplete05:23
superm1ah05:25
fabbioneoh boy.....05:26
fabbionedpatch for a one liner?05:26
fabbioneno05:26
fabbionethat's just insane05:26
superm1there would be an easier way?05:26
superm1it will be dropped as soon as there is a new upstream version05:26
superm1i figured this keeps it easy to merge with debian05:27
superm1and just drop it when necessary05:27
fabbionezcat cdrkit_1.1.6-1ubuntu5.diff.gz |lsdiff05:27
fabbioneit already shows modified files directly in the tree05:27
fabbioneso either you switch all the diffs to dpatch and do a clean jon05:27
fabbionejob05:27
fabbioneor just patch it and suck the diff in the diff.gz05:28
fabbione2 different patch systems at the same time are _BAD_05:28
* superm1 shrugs05:28
superm1okay05:28
* superm1 doesn't like it when people don't change stuff outside debian/ with proper patch systems05:29
fabbionei don't like when they are incosistent :)05:29
superm1okay i'll just put it right in the .diff.gz then05:30
superm1give me a sec05:30
superm1fabbione, okay i updated that patch on the bug05:34
fabbionebetter...05:35
fabbioneok i will sponsor it, but you get to fix the crap if it breaks05:35
superm1sounds fair :)05:36
fabbionebecause i have never seen that error before05:36
superm1it is fscking up our mythbuntu disks05:36
superm1so that's the only reason the bug became so apparent05:36
fabbionesuperm1: i guess you mean the disks created by mythubuntu disks.. it's not related to my problem tho..05:36
superm1i mean our live cds.  they are not browseable on macs or windows, which with the next alpha we're doing - we're adding a nice little windows menu to install the "Mythtv player" for windows and a few little things05:38
fabbionesuperm1: uploaded05:38
superm1i'll keep an eye out for new bugs related to this bug though.  if something goes wrong i'll take care of it. Thanks fabbione :)05:39
fabbioneno problem05:39
=== freeflyi1g is now known as freeflying
slangaseksoren: mmh, sorry, I did get it taken care of while it was still today for /me/, I didn't expect you'd be up this late still looking for it07:15
=== doko_ is now known as doko
Hobbseecalc: ah09:00
HobbseeKeybuk: pumpernickle's proxy appaers to suck.09:00
Hobbsee(it's happened before)09:00
superm1slangasek, could you merge another branch related to our alternate cd building? http://bazaar.launchpad.net/~mythbuntu/debian-cd/mythbuntu-debiancd has the .pcx/.rle image files so we get the right artwork09:11
hunger_tI can't update anymore since yesterday. What is wrong?09:15
=== hunger_t is now known as hunger
superm1slangasek, also added to there our first seeds that should hopefully pick the right tasks off the bat.09:29
Hobbseehunger: you broke it?09:31
Hobbseehunger: open office?09:31
hungerHobbsee: Nope, I get 404 errors when trying to download anything but the package lists.09:31
hungerfrom http://archive.ubuntu.com/ that is.09:32
Hobbseeah09:32
crevettehunger: yeah I had that since yesterday too09:32
crevettes/had/have/09:33
hungerFrom the apt-get output I gess that the filenames it is trying to download are missing .deb extension or something.09:34
hunger"Err http://archive.ubuntu.com hardy/main update-manager 1:0.87.10\n  404 Not Found"09:35
superm1hunger, happened to me too, but it was transient09:35
superm1an apt-get update and then trying again appeared to fix it at least for me09:35
hungersuperm1: I had it a couple of times yesterday and am seeing it consistently the whole morning already.09:35
mdkecould someone paste me a copy of a default hardy sources.list (does it vary for the server edition?)09:46
Hobbseemdke: no09:46
Hobbseemdke: (surely, you should know better than to even ask such a thing.  next, you'll be asking what the repos for kubuntu are)09:46
mdkeHobbsee: no to both questions? ;)09:46
Hobbseemdke: to the latter :)09:47
=== tritium is now known as HDStrikeOut
Hobbseemdke: http://ubuntuforums.org/archive/index.php/t-308907.html do?09:50
Hobbseethen sed09:50
mdkehas the layout not changed since edgy?09:51
Hobbseehm, maybe.09:51
* Hobbsee doesn't remember09:51
mdkeI was wondering if universe/multiverse is on the same line as main/restricted or if the wording of the comments has changed, etc09:51
=== kuechengarten is now known as thekorn_
mdkei'll download the source package09:54
mdkedamn, nothing very helpful there; it's created automatically09:57
mdkewhich i suppose might mean it's different for a different edition :)09:58
hungerAh, now the download works once more.10:02
hungerWell, mostly...10:12
Hobbseegrumble.  why does hardy, a LTS release, *still* not pick up vertical scrolling for touchpads?10:56
Hobbseeand i can't even figure which part in particular does that now10:57
lifelessHobbsee: works for me10:57
Hobbseelifeless: which model?10:58
Hobbseedear tracker, if i've told you to pause indexing, then why are you still doing it?10:59
NafalloHobbsee: I think you want to talk to trackerd :-)11:00
lifelessHobbsee: kill trackerd kthxbye.11:00
lifelessdell d43011:00
Hobbseeyummy.  gnome-app-install curls up and dies if something else already has the apt lock11:00
Hobbseelifeless: yeah, well, i'm about to do that :)11:00
Hobbseelifeless: i was giong to see if it behaved *vaguely* well before doing so, though11:01
HobbseeNafallo: yeah, something about kill and purge...11:01
Nafallo:-)11:01
Hobbseelifeless: strange.11:02
Hobbseeevand: can i lodge a complaint, please?11:05
Hobbseeevand: there's absolutely no indication on the initial dialog on the cd that you can install ubuntu from the first option - it now appears that you can either a) play without touching anything permanently, or b) install ubuntu, but not both.11:05
Hobbseenice wallpaper.11:09
Le-Chuck_ITAHi there, what is the freeze for universe bug fixes?11:20
Le-Chuck_ITAis that betafreeze?11:20
HobbseeLe-Chuck_ITA: same as main.11:21
HobbseeLe-Chuck_ITA: see the meeting minutes for the latest info11:22
Hobbsee(which i need to update on teh wiki)11:22
Hobbseejdong: oh crackmaster, please get transmission done and uploaded.11:22
Le-Chuck_ITAHobbsee: same as main is betafreeze for bugfixes, right?11:22
HobbseeLe-Chuck_ITA: i think so11:22
Le-Chuck_ITAThanks. Any developer acquainted with xinput here?11:22
Le-Chuck_ITAthere are big troubles for tablets in hardy11:23
Le-Chuck_ITAand I don't want to see it shipped as is :(11:23
Hobbseeprobably not on a weekend11:23
Le-Chuck_ITAwe are in the weekend already?11:23
Le-Chuck_ITA:)11:24
NafalloSat Mar  1 11:24:05 GMT 200811:24
Le-Chuck_ITAI will wait monday then11:24
* laga has spent one hour more than Nafallo in the weekend. yay.11:24
Nafallolaga: GMT == UTC, go figure ;-)11:24
laga:)11:24
HobbseeAmaranth: is https://bugs.edge.launchpad.net/ubuntu/+source/compiz/+bug/140913 fixed in hardy now?11:33
ubotuLaunchpad bug 140913 in compiz "session save/restore does not work" [High,Triaged]11:33
Hobbseejdong: xvidcap, ditto.11:36
Le-Chuck_ITAI found that the most important bug related to wacom tablets has already been reported three days ago by Tom Jaeger, together with a fix (using GIT version of linux wacom). I can do nothing but alert developers here and on ubuntu bugs on Bug #195953, please take a look and prioritise if you think it's worth. Tablet support is slowly improving in linux and breaking driver support in a LTS would be a shame11:43
ubotuLaunchpad bug 195953 in wacom-tools "Tablet input resolution tied to display resolution" [Medium,Confirmed] https://launchpad.net/bugs/19595311:43
Le-Chuck_ITAalready done on ubuntu-bugs11:44
Le-Chuck_ITAsorry for noise :)11:44
HobbseeLe-Chuck_ITA: can you milestone it please?11:44
Le-Chuck_ITAHobbsee: yes sure - a couple of months away from launchpad and I forgot completely how to work with bugs :)11:45
Hobbseehehe11:45
affluxwill that funny tracker thing be enabled per default to index *everything* I have on my disk in hardy?12:11
james_wHi all. Does anyone know what package provides the boot menu on the live cd?14:02
james_wI think it may be a syslinux menu, but I can't find what configures syslinux.14:02
james_wThis is to find a package for bug 19721414:03
ubotuLaunchpad bug 197214 in ubiquity "Ubuntu installation menu bug" [Undecided,Incomplete] https://launchpad.net/bugs/19721414:03
james_wor rather confirm the package.14:03
mastrohi there.. i had an idea for improving the liveCD feature but i don't know some thing... my idea is this ( http://brainstorm.ubuntu.com/idea/2031/ ) it's simple: create a way to save session of a live cd and let you use the liveCD as if it's an installed Ubuntu... as i know the LiveCD work on a virtual filesystem loaded into RAM.. the idea should be to "expand" this filesystem to split this filesystem to be half o14:11
mastron RAM and half on some other support.. like if there's more ram... than to save new thing you can do a "binary diff" and tar it somewhere... when you load the liveCD again you need a way to load this binary diff and apply it... but i don't know where it should be loaded and how you can friendly let the user choose from which support to load it...14:11
persiamastro: I remember reading something about someone having done something like that using a USB key.  You might want to try searching for such a project, and joining that team to seek implementation.14:13
mastropersia, really? i search on google.. thanks14:13
james_wIt's wanted to have an easy way to save data on a USB stick or similar when using the live cd, but I don't know if it is planned to have full session save/restore.14:18
mastrojames_w, yep... i think it's not an easy task to achieve! But the main useful thing would be: "possibility to try restricted drivers like nvidia or ati drivers directly from liveCD"14:25
nosrednaekimif this the correct line for dapper-proposed?14:31
nosrednaekimdeb http://us.archive.ubuntu.com/ubuntu/ dapper-proposed main14:32
james_wmastro: isn't that already possible?14:32
james_wnosrednaekim: I believe so. Does it not work?14:32
nosrednaekimjames_w» doesn't seem to have the "upgrade-manager" package in it14:33
=== toresbe_ is now known as toresbe
james_wnosrednaekim: the package is called "update-manager"14:37
nosrednaekimoo! :)14:37
nosrednaekimI feel dumb now... thanks14:39
mastro<james_w> mastro: isn't that already possible? «« i don't think so.. not easily14:40
mastroit require to reboot14:40
james_wah, ok.14:41
=== asac_ is now known as asac
sorenslangasek: Ah, ok. I didn't know your day extended all the way to midnight :) I'm in Boston, so I'm not on my usual hours.15:29
affluxdoko: could you please have a quick look at bug 197160? It's a crash in python-central 0.5.60ubuntu5 which was previously marked as fix released in bug 19633516:01
ubotuBug 197160 on http://launchpad.net/bugs/197160 is private16:01
ubotuBug 196335 on http://launchpad.net/bugs/196335 is private16:01
lifelessthom: so what did you not like about the air?16:32
jdongHobbsee: *hugs* thanks! Now off to write another one for Nexuiz ;-)16:47
Nafallojdong: MIRs? ;-)16:50
jdongNafallo: ROFL no, though nexuiz-by-default would be quite interesting ;-)16:50
jdongFFe16:50
Nafallo:-)16:51
=== allee_ is now known as allee
evandHobbsee, suggestions for an alternative menu item title are welcome.17:05
evandI honestly don't recall what the current options are and I don't have vmware set up at the moment.17:05
* Keybuk tries a new trick17:19
Keybukchange the test cases *before* changing the code17:19
ScottKdoko: I agree that the change I made for Bug #179668 is a work around.  I shifted the duplicates around and left Bug #138189 open to document that libythonize0 is dlopening libpython2.5.so and not .so.1.  So I'm going to put the one back to fix released since the underlying issue is documented in another bug.17:26
ubotuLaunchpad bug 179668 in pykdeextensions "User management in systemsettings cannot find libpython2.5.so" [High,Confirmed] https://launchpad.net/bugs/17966817:26
ubotuLaunchpad bug 138189 in pykdeextensions "application tries to dlopen /usr/lib/libpython2.5.so (only found in the -dev package) " [Medium,Confirmed] https://launchpad.net/bugs/13818917:26
dokoScottK: ahh, ok17:26
superm1ScottK, promoting those to recommends, they won't be coming by default though will they?  I thought there was a decision at UDS-Boston to automatically install recommends for hardy17:37
ScottKsuperm1: Whether it does or not doesn't affect the right answer.  Last I noticed it's not implemented, but I may be wrong.  If it's recommeds, users can still remove it.17:38
ScottKGotta run.17:38
superm1ok17:38
=== j_ack_ is now known as j_ack
=== luka74 is now known as Lure
__Neo__Hi, is there a variant of Ubuntu that uses official (or mirrors of) Debian repositories either stable or later?18:20
Nafallonot that I know of.18:20
laga__Neo__: i think that's called "debian"..18:21
__Neo__DreamLinux does this for example18:21
CarlFKheh18:21
__Neo__Essentially real Debian testing with extra things added and customized for the Desktop user - whilst maintaining compatibility with the debian repository18:21
__Neo__I presume the answers no then?18:24
Amaranth__Neo__: The answer is no18:28
__Neo__I'll go back to DreamLinux - thanks anyway18:31
pecisk_hi people, question - do automatical daily live cd building system includes up-to-date translations from Launchpad, like debian-installer and it's help module?19:48
evandstgraber, can you make evand a developer on brainstorm, so I can post developer comments?19:56
superm1i've seen at least one spec on there that is already implemented in hardy too.  if the ability to nuke such things as "complete" is available, that should be done (i'm referring to the ability to cancel fsck on bootup)19:58
Keybuksuperm1: not to mention "Ubuntu should recognize hardware changes"20:06
superm1yeah that too :)20:06
Keybukcomplete things shouldn't be nuked though20:07
Keybukthey should be moved to a "ideas that we listened to" board20:07
Keybuklike the one in TESCO20:07
superm1yeah, well who admins the site?20:08
superm1stgraber, ?20:08
Nafallosuperm1: I think so20:10
nandevand: The role is not yet here, we will add it at the next update, probably monday.20:38
evandnand, oh, I thought the developer comment section that's already up was the change.  Nevermind then.20:43
nanda lot of changes are incoming :)20:44
=== allee_ is now known as allee
=== lando_ is now known as lando
vignattihi20:59
vignattiwho is the tool responsible for mount /proc on ubuntu?20:59
vignattiI tried to comment fstab line to not mount this but it doesn't work21:00
crevettevignatti: why don't you want /proc ?21:02
vignatticrevette: just to kill an xserver bug. Some guy is trying to initialize Xorg without /proc and it's segfaulting21:03
vignatticrevette: for sure Xorg must a least exist nicely :)21:03
vignattiany hints?21:05
crevetteI don't how /proc is mounted21:05
crevetteumount /proc doesn't work I assume21:05
nandcrevette: it should21:06
vignattinop21:06
vignattiit returns EBUSY21:06
geser/etc/init.d/mountkernfs.sh mounts /proc21:06
vignattieven if I pass --force21:06
nandat least, I can mount/umount on a small chrooted system. If processes use files in /proc, it's another story...21:07
vignattigeser: humm, interesting21:08
* vignatti booting21:08
vignattihumm, just commenting out the line which mounts /proc in mountkernfs.sh didn't worked :(21:15
=== Onyxyte is now known as Quaver2
slangaseksuperm1: hrm, this appears to require me figuring out how the debian-cd repo is configured21:43
superm1slangasek, yeah evand wasn't sure either, so he said to talk to you :)21:43
slangaseksoren: ah yes, I forgot about the sprint, doh21:44
slangasekScottK: huh, didn't this pythonize issue come up before?22:26
ScottKslangasek: It did.  You fixed it (work around) right before Gutsy release.22:27
ScottKslangasek: Then mvo reverted your work around in December with no rationale documented.22:27
slangasekah22:28
ScottKSo I spent several hours last night beating my head against the wall trying to figure out the real fix, traced it (I think) to libtool and gave up and reuploaded the workaround.22:28
slangasekheh22:28
ScottKIt had enough dupes that I really thought it ought not be left broken for Alpha 6.22:30
* slangasek nods22:32
slangasekso 138189 is the bug designated for tracking getting libpythonize fixed to use the soname?22:34
ScottKThat's the idea.22:34
ScottKIt actually correctly describes that problem and not the impacts of it on other packages.22:34
* slangasek takes responsibility for that one22:38
slangaseksuperm1: mythbuntu debian-cd changes merged22:39
superm1slangasek, thanks.  hopefully tonight's build turns out more fruitful :)22:47
Keybuk17124 scott     18   0  522m  86m  17m S   11  4.3 139:56.82 deskbar-applet23:08
Keybuk*sigh*23:08
ion_Heh23:12
jwest--hi23:15

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