/srv/irclogs.ubuntu.com/2010/11/22/#ubuntu-devel.txt

StevenKRAOF: O hai! Do you tim to throw some debugging options at me for a DPMS issue I'm having?00:46
StevenKs/\(tim\)/\1e/00:46
RAOFI have any number of Tims to throw.  And maybe even a little time :)00:47
RAOFWhat's the problem?00:47
StevenKRAOF: I have "Put display to sleep when inactive for: 0:10" in Power Management, it just ... doesn't00:48
RAOFHeh.00:48
RAOFIs something inhibiting the display sleep?00:48
StevenKI'll come back to the machine after a few hours, and the monitor is still helpfully showing the screensaver00:48
RAOFDoes “xset dpms force off ; sleep 5; xset dpms force on” turn off the monitor for 5 seconds?00:49
StevenKNope, it blanks, and then comes right back00:50
StevenKAlthough, it seemed to work the second time ...00:50
RAOFHeh.00:51
RAOFI'd suspect something higher in the stack than X, generally.00:52
RAOFParticularly if xset works :)00:52
=== sanchaz is now known as sanchaz-away
=== dendro-afk is now known as dendrobates
=== TerminX_ is now known as TerminX
=== amitk is now known as amitk-afk
dholbachgood morning!07:04
didrocksgood morning07:53
pittiGood morning08:02
pittibuxy: as ebroder says, we primarily need to install less; compressing .debs more won't help much08:03
=== almaisan-away is now known as al-maisan
=== hanska is now known as dapal
=== doko_ is now known as doko
=== rraphink is now known as raphink
pittijibel: do you think you can play with the test case in bug 672964, so that someone else than just me tests this? I know that it doesn't help to fix an already broken system, but I'm fairly sure the patch will prevent breaking it in the first place08:53
ubottuLaunchpad bug 672964 in udev (Ubuntu Lucid) "2.6.32-26 unbootable: does not find root file system" [Critical,Fix committed] https://launchpad.net/bugs/67296408:53
jibelpitti, I'm looking at it.09:00
pittijibel: merci; not that urgent, though09:01
mvojussi: could you please add to the channel topic that I'm "patch-pilot" of the day ? (ref is https://wiki.ubuntu.com/UbuntuDevelopment/CodeReviews#Patch%20Pilots)09:58
jussimvo: this channel is not +t - you are able to do so yourself :=)09:59
jussimvo: though I think we will have the bot to help out soon09:59
pittimvo: happy piloting!10:00
pitti"launch clearance granted"10:00
seb128mvo, oh nice, that's starting today ;-)10:01
* mvo puts his pilot hat on and starts the engine10:01
mvojussi: aha, thanks10:01
dholbachcan somebody approve my mail to u-d-a - it's also about the patch pilot programme10:01
dholbachplease :)10:01
pittidholbach: will do10:01
* dholbach hugs pitti10:02
pittiflushed10:02
* pitti hugs dholback10:02
* dholbach will blog it too10:02
=== mvo changed the topic of #ubuntu-devel to: Archive: Open | maverick-proposed is now unfrozen | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs; Friendly Patch Pilot of the day: mvo
dholbachhaha10:02
* dholbach hugs mvo10:02
=== cjwatson changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Friendly Patch Pilot of the day: mvo
cjwatson(I think we can drop the maverick-proposed unfrozen item now)10:03
* mvo nods10:04
brycehdholbach, is patch pilot only looking at debdiffs?10:06
dholbachbryceh, see the mail I just sent to u-d-a10:07
dholbachI'll give you the link to my blog post in a sec :)10:07
dholbachhang on10:07
brycehdholbach,  I just did10:07
brycehdholbach, but I'm unclear10:07
dholbachhttp://daniel.holba.ch/blog/?p=82010:07
dholbachbryceh, http://reports.qa.ubuntu.com/reports/sponsoring/ is a good start10:08
om26ermvo,  can you sponser this fix https://code.launchpad.net/~om26er/ubuntu/maverick/gwibber/gwibber-fix-674894/+merge/4089910:09
dholbachbryceh, is your question about "what you're supposed to do" or "which queue to look at"?10:09
brycehdholbach, no, my question is, if a bug report has a patch that is not in debdiff form, is it acceptable or unacceptable to subscribe ubuntu-sponsors?10:10
om26erthe fix is taken from gwibber trunk10:10
ogra_acdidrocks, hrm, you messed up ubuntu-netbook-default-settings10:10
dholbachbryceh, if you reviewed the patch and it's sound and you want to upload it and it's only a matter of adding a changelog entry, it's great if you just go ahead and do it10:11
didrocksogra_ac: well, on armel… that's why I asked you to test it before on an armel build :)10:11
dholbachwhat I usually do is say "this is what I did to get it in an uploadable state"10:11
ogra_acdidrocks, you should also never use DEB_HOST_ARCH, that breaks cross building and makes lool unhappy ;)10:11
jibelpitti, re 672964, I cant reproduce the original issue with udev 151-12.2 in lucid-updates and the steps described in the test cases , is there any other prerequisite ?10:11
didrocksogra_ac: oh really? what should be used?10:11
didrocks(saw that in other packages too)10:12
ogra_acdidrocks, not armel related, seems that dh7 cant handle the code in rules the way you used it10:12
pittijibel: hm, I only tried on maverick, and there it reproduced well10:12
didrocksogra_ac: but you know I like make lool unhappy :)10:12
ogra_acdidrocks, ifeq ($(DEB_BUILD_ARCH),armel) would be better for cross builds10:12
ogra_acthat takes the target arch instead of the build machine arch10:12
pittijibel: there shouldn't; if the initramfs gets built correctly while udev is unpacked, but not configured yet (i. e. /sbin/udevadm is diverted), the bug wouldn't happen10:12
pittijibel: but I don't see what would stop that from happening in lucid10:13
jibelpitti, I'm trying in maverick.10:13
didrocksogra_ac: ok, updating to that then (same for ubuntu-netbook)10:13
ogra_ack10:13
didrocksogra_ac: then, those two packages are yours :)10:13
mvoom26er: sure, I have a look now10:14
ogra_aci_m still trying to figure out what makes it fail with "missing separator"10:14
didrocksogra_ac: maybe you want to do the change and test with your hw rather?10:14
ogra_acits not HW related10:14
ogra_acbut i cant see a typo either10:15
didrocksogra_ac: it's building locally…10:15
didrocksand it built everywhere !armel10:15
geserogra_ac: whitespace issue: space vs tab before the dh_install10:16
=== amitk is now known as amitk-afk
ogra_acwell, it doesnt go the script path you created if DEB_HOST_ARCH isnt set to armel10:16
jussits2: you going t explain for all the lovely folks here how this works?10:16
looldidrocks, ogra_ac: It's hte other way around; you want to use DEB_HOST_ARCH, not _BUILD_ARCH10:17
=== jelmer is now known as Guest24864
loolthe arch on which you build it on shouldn't affect the results; only the target arch on which you expect to run it (host)10:18
ts2jussi: one sec, yeah :)10:18
loolgeser found the real problem it seems10:19
ogra_acdidrocks, http://paste.ubuntu.com/535147/ that way it shoudl work10:19
didrocksyeah, I should sometimes have an automatic rules to not strip tabs when opening debian/rules in vim10:20
ogra_acgeser, oh10:20
didrocksreplacing by a tab and uploading then10:20
didrocksthanks geser, lool, ogra_ac10:20
didrocksand so nothing to change for the ubuntu-netbook metapackage10:20
mvoom26er: thanks a lot for this branch. it looks fine, the only missing bit is a "TEST CASE" in the bug #674894 so that the SRU verification can be done, it should be pretty trivial in this case :) one more question - did you talk to kenvandine about the upload? I just want to ensure that he does not plan to upload this as part of some bigger patchset to maverick-proposed10:20
ubottuLaunchpad bug 674894 in gwibber (Ubuntu Maverick) "gwibber prepends "is" to messages sent to facebook" [Undecided,New] https://launchpad.net/bugs/67489410:20
loologra_ac: No it wont; this will not define override_dh_install on != armel, so it will run dh_install by default10:20
ogra_aclool, yeah, right10:21
ogra_aclool, thanks for both10:21
looldidrocks: It would be nicer to ifneq ($(DEB_HOST_ARCH),armel) override_dh_install: endif and not have any dh_install10:21
didrockslool: oh, directly wrapping the target? ok, can do that :)10:21
om26ermvo, there is nothing coming in maverick-proposed from ken since the last -proposed (big bug fix release) was just gone into -updates10:22
seb128mvo, om26er: you should let kenvandine's do the gwibber sponsoring10:22
loolAlternatively, you could use -X or -N; e.g. override_dh_install: dh_install $(if ...), but I find that harder to read10:22
seb128mvo, om26er: he's active on it and might have a reason for not having uploaded yet, at least check with him before upload10:23
didrockslool: yeah, that was my first idea but I didn't like it when reading10:23
didrocks# Skipping dh_install - empty override <- nice dh7 warned about it :)10:23
om26erseb128, ok sure will do that today10:24
mvoseb128, om26er: ok, thanks. I will comment in the branch but wait for kenvandine for his final word then10:24
mvoif he is busy I'm still happy to sponsor it10:24
om26ermvo, ok this https://code.launchpad.net/~om26er/ubuntu/maverick/gexiv2/gexiv2-fix-636161/+merge/41398 ;)10:26
om26erits recommended from shotwell developer10:27
* mvo looks10:28
=== jussi changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs
jussimvo: others, herses how the bot works10:31
jussi@pilot in10:32
=== udevbot changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Patch Pilots: jussi
jussi@pilot out10:32
=== udevbot changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Patch Pilots:
mvonice jussi!10:32
mvo@pilot in10:32
nigelbneat :)10:32
=== udevbot changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Patch Pilots: mvo
jussimvo: thank ts210:32
* mvo hugs ts210:32
ts2:)10:32
* didrocks is kind of disappointed that the Patch Pilot isn't "Friendly" anymore ;)10:33
jussits2: can we have friendly patch pilots?10:34
didrocks:)10:34
ts2any other requests while I'm at it? ;)10:35
jibelpitti, I can't reproduce in maverick either. I only get a broken initramfs if I stop at step 2.10:35
nigelbdidrocks: you want to bot you hug you at the end of your rotation? ;)10:35
nigelb*want the10:35
jibelpitti, if I run the postinst step, the diversion is removed before running update-initramfs and the correct udevadm is copied to initramfs.10:35
didrocksnigelb: oh neat idea! :)10:36
jibelpitti, or ma I missing something ?10:36
jibels/ma/am/10:36
pittijibel: oh, you mean udev.postinst configure rebuilds the initramfs?10:36
jibelpitti, right.10:36
pittijibel: then it indeed seems that this test case isn't reproducing the problem, sorry10:37
=== ts2 changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots: mvo
ts2@pilot in10:40
udevbotAn error has occurred and has been logged. Please contact this bot's administrator for more information.10:40
ts2yippy...10:40
dholbachthanks ts210:41
geserit would be nice if the bot would tell who the bot admin is10:41
ts2@pilot in10:43
=== udevbot changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots: ts2, mvo
jibelpitti, np. I think that we can only reproduce the issue if the postinst script is not run at all or if it stops before calling enable_udevadm10:43
ts2@pilot out10:43
=== udevbot changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots: mvo
jibelpitti, let me know if you want to test something else.10:43
=== amitk-afk is now known as amitk
pittijibel: will think about it; thanks so far!10:50
mvoom26er: you have just written history, the first upload from the new patch-pilot program10:51
om26ermvo, thanks alot :)10:52
om26eri am legend :p10:52
didrocksts2: nice :)10:52
brycehmvo, om26er, congrats :-)10:52
didrocksheh, congrats mvo, om26er :)10:53
mvo\o/10:53
* pitti ^5s om26er and mvo10:54
om26erbryceh, didrocks thanks ;)10:54
om26erbug 65524110:57
ubottuLaunchpad bug 655241 in AppMenu GTK+ "some applications show duplicated menus" [High,In progress] https://launchpad.net/bugs/65524110:57
om26erhttps://code.launchpad.net/~om26er/ubuntu/maverick/appmenu-gtk/appmenu-gtk-fix-655241/+merge/4139710:57
om26erthe fix is merged upstream mvo10:59
om26erthere are quite a few duplicates too.10:59
om26erthats the last one11:01
mvoom26er: thanks! I look in detail after my lunch break, from fist sigh it looks fine, it would be nice to have a TEST CASE in the bugreport to the SRU process11:02
mvofrom the first look11:02
* mvo can't type today11:02
* om26er writes a TEST CASE11:02
=== hunger_ is now known as hunger
pittiev: FYI, I'll port usb-creator from pygtk2 to gtk3.0 and gi, if that's alright with you?11:06
evthat's fantastic, thanks pitti!11:07
cjwatsonpitti: my understanding had been that gtk3/pygi only worked properly with python3; is that true in your expereience?11:12
cjwatson*experience11:12
pitticjwatson: I have run into one bug where python3 works better11:13
pitticjwatson: but by and large it's fine11:13
cjwatsonok, that's a relief11:13
pitticjwatson: https://bugzilla.gnome.org/show_bug.cgi?id=620579 is the one FYI11:13
pittiit's being fixed11:13
pittibut it's easy enough to avoid/work around for now11:13
seb128pitti, no, it works fine with python2, though they recommends going for python3 while you are at porting if you can11:14
seb128ups11:14
seb128cjwatson, ^11:14
pittithat's right11:14
pittibut we don't currently ship py3 on the CDs11:14
pittiand I don't want to be the one who introduces it :)11:14
mvodidrocks: are you ok with me uploading #655241 ? it looks like you did most of the recent uploads?11:14
seb128right, I don't think we should go for python3 yet11:14
cjwatsonseb128: I just don't want to have to do both ports simultaneously11:14
mvodidrocks: diff looks fine and is taken from the upstream branch11:14
cjwatsonin general11:14
seb128right, my idea was to do python2 gi ports this cycle11:15
didrocksmvo: sure, cherry-pick it :)11:15
didrocksthanks :)11:15
mvothanks didrocks, uploaded11:16
seb128mvo, yeah, feel free to do appmenu-gtk uploads11:16
mvothanks om26er!11:16
seb128mvo, oh, you are the new maintainer then11:16
seb128;-)11:16
didrocksyou gave me compiz, you have appmenu :p11:16
seb128joke aside appmenu-gtk has a collection of issues if anybody feels like working on it...11:16
mvoseb128: geh, I'm pilot, not superman11:17
* seb128 hugs mvo11:18
dholbachseb128, I just noticed in Harvest that there's a number of patches and branches for it :)11:18
seb128dholbach, I will kick dx-ers to get reviews11:18
dholbachsweet11:18
geseranyone available to review and sponsor a gparted SRU? the debdiff is in comment #53 of bug #61788511:19
ubottuLaunchpad bug 617885 in gparted (Ubuntu Maverick) "gparted crash at start: glibmm-ERROR **" [High,Triaged] https://launchpad.net/bugs/61788511:19
seb128or nicely nudge rather11:19
seb128geser, I though cjwatson said he had it on his list11:19
mvogeser: I guess that is me11:19
mvoseb128, geser, cjwatson: I'm the pilot today, I can do the sponsoring (after lunch)11:20
geserseb128: wasn't that the vim merge? I asked cjwatson about gparted too but as he's busy I try to find someone else (the vim merge is still waiting on review)11:20
mvobut I'm equally happy to let someone else do it11:20
seb128geser, yeah, searching in my log I think I confused it with another patch11:20
seb128let mvo rock it ;-)11:20
cjwatsonsorry, I really haven't had time to deal with gparted - if the patch pilot could deal with it I'd be enormously grateful11:22
cjwatson(otherwise my patch pilot morning is coming up next week, but you might not want to wait for that ...)11:23
* mvo will take it11:24
cjwatsonthanks!11:25
mvoev: if you could have a look at some point at #545445, that would be nice. its  support for lxde in ubiquity, afaics its fine, I commented in the bug11:29
mvo(I had two question for the diff, but maybe they are non-issues)11:29
seb128bug #54644511:32
ubottuLaunchpad bug 546445 in ubiquity (Ubuntu) "Support for lxdm in ubiquity (autologin, only-ubiquity support)" [Wishlist,Confirmed] https://launchpad.net/bugs/54644511:32
seb128mvo, rather?11:32
mvoseb128: rather?11:33
seb128mvo, the number, see line before11:35
seb128mvo, #545445 has nothing to do with ubiquity?11:35
mvoseb128: oh, sorry11:35
mvoseb128: indeed, typo11:36
* mvo is away for lunch now11:36
seb128hum, lunch!11:36
smoseranyone else seeing bug 676790 ? I'm almost certain that its not just me (I've seen it happen every time during debootstrap on host system of maverick or lucid).11:36
ubottuLaunchpad bug 676790 in apt (Ubuntu) "huge performance regression" [Undecided,New] https://launchpad.net/bugs/67679011:36
seb128mvo, great idea ;-) I will do that as well11:36
smoserit seems fairly important to me, that it would affect cd install, and turn that into a multi hour affair11:36
=== Guest24864 is now known as jelmer
evmvo: when you get back, any idea what the details of "[ev] support "keep package selection" on install that preserves /home: TODO" in packageselection-foundations-n-update-manager-improvements are?  I thought OneConf was the approach for maintaining packages between installs.11:43
smoserhas anyone attempted a natty ISO install ?11:55
ogra_acthere were plenty probs reported in the release meeting11:56
ogra_aciirc X restarts continiously11:56
dholbachogra_ac, I think that was a kernel bug and is fixed11:56
ogra_acalready uploaded and on the isos ?11:57
dholbachuploaded yes, isos I don't know11:57
ogra_ack11:57
ogra_aci just remember it was reported on friday11:57
* ogra_ac doesnt touch x86 much anymore ;)11:58
=== xfaf is now known as zul
mvoev: well, oneconf requires a ubuntu one account. it seems like its pretty straightforward to preserve the selection via a chroot dpkg --get-selection like mechanism from the to-be-overwriten system. I don't have strong opinions here, but it sounds like the amount of work for this is about the same, the chroot dpkg --get-selections would just be one of two possilbe backend methods to retrieve the data needed12:06
mvo(as straightforrward as dealing with potentially multiple /usr, /var, /boot mount points can be I suppose ;)12:07
evunless of course they have third party repositories12:08
evin which case, as mpt pointed out, we risk removing all of their paid applications12:08
sladenwhere are the IRC logs from UDS?12:11
=== al-maisan is now known as almaisan-away
mvoev: well, oneconf only partly helps here12:15
mvoev: for payed apps we need to do "update repo to nre-release, check if software-is-still-there, if-not, fail or warn"12:16
mvoev: we could extend the "preserve-packages" to preserves sources.list12:16
mvonot sure if that still fits in the spec, it gets complicated quickly12:17
=== MacSlow is now known as MacSlow|lunch
cjwatsongeser: DMB meeting?  we have some timezone confusion apparently12:25
=== ogra_ac_ is now known as ogra_ac
evmvo: speaking of the specification, are you working on that or are you keen to keep it as just a list of work items?12:32
mvoev: the oneconf one?12:33
evfoundations-n-update-manager-improvements12:33
mvoev: so far I was happy with the workitems as more a loose selection of todos, but I'm happy to write a proper spec if you prefer that12:35
evI'm just concerned about handling the reinstallation of third-party applications.  dpkg --get-selections is easy enough if we assume a vanilla maverick to natty, but if we also copy over sources.list and try to reinstall packages, we may have broken dependencies.12:37
mvogeser: I'm happy to upload your gparted patch, could you please add a TEST CASE to the bugreport for the SRU verification - if that is not straightforward, just add something like "no known test method, regression test plus getting feedback fro mthe original reports required"12:39
mvoev: yeah, I think it will need a python-apt app, I can work on something like that if it helps12:40
mvoev: a more intelligent version of dpkg --get-selection that looks for stuff that is not downloadable, stuff that comes from third party repos and then checks if the third party stuff has upgrades etc12:43
mvoev: similar to what u-m is doing when it performs its checks. we need to figure out a UI for this12:43
mvoI think that is the hardest part :)12:44
evexactly12:45
evanswering the question of what we do when it cannot resolve something12:45
evbecause we've told the user we were going to preserve applications or whatever12:46
evbut yes, a python-apt application would be most helpful12:46
=== amitk is now known as amitk-afk
mvoev: thanks, I add a workitem for myself. I guess the exact message/Ui is on mpt, but I imagine something like a warning that the exact package selection can not be restored and a (optional) list of detailed packagename/descriptoins12:51
mvoev: I add a workitem for myself12:51
Chipzzev: wouldn't patching dpkg so the --get-selections option is aware of manually and automagically installed packages be more usefull?12:54
Chipzz(but the manually vs automatically installed information is sth on the apt and not dpkg level I suppose)12:54
=== Claudinux_ is now known as Claudinux
mvoseb128: bug #486154 looks like something that really should be done upstream, no?13:09
ubottuLaunchpad bug 486154 in metacity (Ubuntu) "System beep broken in Karmic despite heroic efforts to fix it" [Medium,Triaged] https://launchpad.net/bugs/48615413:09
mvoseb128: or do you think its appropriate as a distro patch?13:09
seb128no strong opinion about it, we use compiz by default13:10
seb128would be nicer done upstream though13:10
seb128but if you want to upload go for it13:10
mvoI don't feel confident in the diff enough for that, I would prefer upstreams comment13:13
seb128right13:13
=== MacSlow|lunch is now known as MacSlow|
=== MacSlow| is now known as MacSlow
=== almaisan-away is now known as al-maisan
=== tgall_foo is now known as tgall_out
gesermvo: I'll update the bug and add the TEST CASE when I'm back home14:11
gesercjwatson: the DMB meetings at 12 UTC don't suit me at all as I'm in the middle of a lecture at that time14:12
sorengeser: It's still going on, if you're interested.14:13
mvothanks geser14:14
=== ogra_ac_ is now known as ogra_ac
persiabdrung, So, yeah, I tend to agree with you about core-dev being about work primarily in "core", but I'm uncertain to some degree, in part because of the issues one has in bringing new stuff in, and in part because all the packagesets depend on core, which often means one has to work in core in order not to just upload workarounds for various issues.14:55
cjwatsonbdrung: if you disagree with the design of the package set split, that's one thing, but kenvandine has been working on things that *are* core packages and that are core because they're used by products other than desktop14:55
cjwatsonbdrung: in some cases we've made specific exceptions, but the volume is high14:56
cjwatsonand it's not always obvious that it *is* correct for these packages to be in desktop14:56
cjwatsonalthough netbook being merged back into desktop reduces the question marks a bit14:56
pavolzetorhi, please look at this fix https://code.launchpad.net/~pavolzetor/ubuntu/natty/liferea/liferea-test14:58
kenvandinea big chunk of what lands on the ubuntu-desktop CD isn't in the desktop package set14:59
cjwatsonbdrung: for example libindicator is in both Ubuntu and Kubuntu desktops, and therefore is in core14:59
seb128cjwatson, bdrung, persia: we also could use extra people being able to do sponsoring14:59
cjwatson(actually desktop-core I think)14:59
seb128out of set specific discussions14:59
seb128even if 99% of kenvandine's work is in desktop14:59
cjwatsonbdrung: I made an exception because it seemed most sensible at the time, but this means that people entirely focused on ubuntu-desktop can make changes to kubuntu, which is really the sort of thing core-dev is meant for15:00
persiaseb128, I'm not confortable with that rationale: I'm not convinced the sponsoring issue is a lack of people, so much as a lack of people willing to touch a small subset of critical packages.15:00
seb128if we have people who know enough about what they are doing to help on sponsoring that would be nice to have them set with upload rights15:00
cjwatsonbdrung: so simply saying "oh that should be in desktop" simplifies the situation IMO too much15:00
persiaseb128, And I'd rather grant access to folk because of my confidence in them uploading stuff, than because of my confidence in their ability to sponsor stuff.15:00
pavolzetorcoukd someone help me15:01
seb128persia, sponsoring is uploading, I fail to see the difference15:01
pavolzetorI would like help make ubuntu better15:01
seb128pavolzetor, try #ubuntu15:01
pavolzetorso :-D15:01
pavolzetorubuntu-mote => ubuntu-devel => ubuntu15:01
pavolzetortrhanks15:01
persiaseb128, Yes, I agree.  There is no difference.  Hence I don't care about a willingness to sponsor (other than that everyone should have one)15:01
seb128pavolzetor, well maybe if you state your question in a better way15:01
cjwatson14:58 <pavolzetor> hi, please look at this fix https://code.launchpad.net/~pavolzetor/ubuntu/natty/liferea/liferea-test15:02
seb128pavolzetor, but "can somebody help" usually go to #ubuntu15:02
kenvandinepavolzetor, oh... hang on15:02
cjwatsonseb128: ^- did you see that?15:02
kenvandinepavolzetor, that isn't #ubuntu :)15:02
seb128cjwatson, no, just "<pavolzetor> coukd someone help me"15:02
seb128cjwatson, thanks15:02
seb128pavolzetor, sorry, I didn't read the first line you wrote15:02
pavolzetorI have done patch15:02
BlackZpavolzetor: I will test your patch15:02
kenvandinepavolzetor, are you looking for sponsoring?15:02
pavolzetorand branch15:02
pavolzetorI am bit new in ubuntu dev15:03
seb128pavolzetor, you can try to ask mvo, he's patch pilot today15:03
seb128see the topic15:03
seb128pavolzetor, you might want to subscribe ubuntu-sponsors as well15:03
pavolzetorokey, I just must go out15:03
=== mrenouf is now known as mrenouf_home
persiapavolzetor, Don't run away, just because we're confused :)15:04
kenvandinepavolzetor, sorry we were in the middle of a debate :)15:04
pavolzetorI am student and I fixed liferea indicator in free time15:04
pavolzetorI am sorry15:04
kenvandinepavolzetor, awesome... no worries15:04
seb128persia, well, I don't say we should care about willingness to sponsor, I'm just saying that kenvandine understand enough of what he's doing to do uploads out of the desktop set if he needs to15:04
kenvandinei should look at that then15:04
persiapavolzetor, Don't be sorry.  That's wonderful.  Thank you.15:04
pavolzetorI don't wanna make big changes in source, because of diffs between ubuntu and others15:05
ivoksguys, how does one make upstart not to kill child processes?15:05
seb128pavolzetor, we will get your patch reviewed and comment, you can maybe do a merge request from it?15:05
=== mrenouf_home is now known as mrenouf|home
pavolzetorI have done it15:05
seb128pavolzetor, ok, so wait a bit and somebody is going to review the work you did and comment on it15:06
pavolzetorand I fixed gpm twcie steps, a "big bug"15:06
pavolzetorhttps://bugs.launchpad.net/ubuntu/+source/linux/+bug/25782715:06
* persia finishes playing with "comm"15:06
pavolzetorI am trying get it to gnome also15:07
pavolzetorI have tested it at two comps and it works properly15:07
pavolzetorbrb15:07
seb128persia, is your issue with kenvandine's application that you think that he doesn't have the skills to work out of the desktop set?15:09
seb128persia, I'm not sure to understand the issue15:09
persiaseb128, No, my issue is that from what I have seen, he's never had anything in "core" even as a sponsored upload.15:09
persiaI'm more than confident in his skills, but like to see some examples of work in the target packagesets before granting upload access.15:10
seb128why?15:10
persiaThat said, I might be using edit_acl or comm incorrectly.15:10
seb128it's a chicken egg issue15:10
persiaWhy?  Most folk find sponsors for stuff before applying.15:10
seb128he could be sponsoring things he doesn't work on if he had upload rights15:10
seb128he doesn't work out of desktop15:10
bdrungseb128: yes, we need more people doing sponsoring. there are 12 sponsor request open for ubuntu-desktop that kenvandine can work on.15:10
seb128it doesn't mean he couldn't make use of upload rights for sponsoring15:10
kenvandinepersia, seb128 has sponsored lots of stuff for me15:10
pittimvo: when you moved to apt-get changelog, did you keep the configuration support? (i. e. setting the mirror)15:11
seb128kenvandine, in the desktop set though15:11
persiakenvandine, I believe you.  I'm just not sure why I don't see it.15:11
seb128kenvandine, or things that ought to be in there15:11
kenvandineseb128, sort of... the fact that you needed to sponsor them though15:11
persiaMight just be race conditions then: stuff being sponsored just before it hits.15:11
seb128bdrung, we know that we need more people doing sponsoring15:12
pittimvo: ah, it's in apt-get(8) now, nevermind15:12
seb128persia, it's not even a race, if kenvandine has upload rights he could take on reviewing and sponsoring things out of the set he works usually on15:12
persiaseb128, There's a sufficiency of patches not currently embedded in sponsor requests that would benefit from wrapping that I don't find that convincing.  There's plenty of stuff that can be uploaded that isn't sponsor requests.15:14
seb128persia, I don't see the point of stopping people to get upload rights if they have the skills for15:14
seb128they might not use them now15:14
persiaWhat I'm less convinced about is whether the potential race condition between the state of the packageset at the time an upload was made and the state when I am reviewing the upload gives me an accurate picture of the historical work done.15:15
seb128but they would be set if there is any need showing up15:15
superm1ev, mvo: in terms of UI, it is sounding to me like once the person selects the option to reuse a partition you'll need to present them with another ubiquity page that gives them some details on what is about to happen with some checkboxes. Eg 1) Any  files in /home will be saved 2) The following applications that come from Ubuntu will be re-installed 3) The following applications that you paid for will be re-installed 4) The following third-15:15
superm1party sources will be re-activated.  You will need to manually reinstall applications from them; etc15:15
bdrungseb128: you should show that you can work in that area you want upload rights to.15:15
seb128persia, how can it hurt to have people skilled to be able to upload? even if they don't use that right often15:15
superm1and using the python-apt black box to build the lists for those checkboxes15:15
seb128bdrung, right, but persia said the issue was not on showing that you can15:16
seb128that's why I was asking if the issue is that he didn't show enough packaging skills15:16
bdrungseb128: i didn't saw enough outside the ubuntu-desktop set. there is no link to a launchpad bug or upload.15:18
seb128is outside of desktop harder?15:18
persiaseb128, No, I said that I'm comfortable with kenvandine's skills.  That said, I'm not core-dev, so I wonder if I'm best qualified to review his work on packages in "core".  A lack of apparent uploads (either from an LP bug or coincidence) dissuades me from approving it.  On the other hand, there are a number of other things that kenvandine does need to do which would be facilitated by being core-dev, hence my indecision.  If it was just a matte15:18
persiar of access to the "core" packageset, I'd vote to defer until I could see some work history.15:18
persia(actually, might not be an LP bug, so much as some limits in data collection)15:18
seb128let's be honest, his focus is desktop15:18
seb128that will not be different because he got upload rights out of the set15:19
seb128but that doesn't mean he couldn't make use of uploads right of the set15:19
mvopitti: yeah, I took your documentation and adapted it, the option is almost named the same (apt::changelog*s* iirc)15:19
seb128to do sponsoring15:19
seb128to be able to workaround set definition bugs15:19
mvopitti: thanks for writing the docs btw \o/15:19
cjwatsonmy focus is installer/bootloader, that doesn't mean I do nothing else15:19
seb128to be able to help on something when we need people15:19
mvo@pilot out15:20
kenvandineand not to end up blocked by or distracting others when needing sponsoring myself15:20
=== udevbot changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots:
seb128I fail to see why you have an issue with him having upload rights15:20
seb128if you trust him to upload to main out of desktop you should let him have those right, even if he doesn't use them every day15:21
cjwatsonNCommander: are you being a patch pilot today (per the schedule)?15:21
ogra_accjwatson, he should be, but doesnt seem to be up yet15:21
seb128that show we have qualified people in the team and people able to take on tasks that require those rights15:21
persiacjwatson, It's 7:21am there :)15:21
ogra_ac(we discussed it last week)15:21
SpamapSanybody have ideas what might cause the error about tar here:  http://launchpadlibrarian.net/58855548/buildlog_ubuntu-lucid-i386.mongodb_1:1.2.2-1ubuntu1.1~ppa1_FAILEDTOBUILD.txt.gz ?15:22
SpamapSdpkg-deb: building package `mongodb' in `../mongodb_1.2.2-1ubuntu1.1~ppa1_i386.deb'.15:22
SpamapStar: ./usr/bin/mongo: file changed as we read it15:22
cjwatsonpersia: hm, I forgot he'd moved15:22
SpamapSit works in sbuild and pbuilder .. but the PPA build fails15:22
bdrungseb128: i like to see a list of sponsored uploads to the core package set15:22
cjwatsonhe's still London-0500 in my head15:23
persiaThought that might be it :)15:23
persiaSpamapS, Ask in #launchpad: my random guess would be overeager cleanup of some sort.15:24
seb128bdrung, he doesn't work out of desktop15:24
seb128bdrung, but he could be helping sponsoring out of desktop or be assigned task that require upload out of desktop he had right to work there15:24
seb128bdrung, which is not the case15:24
seb128there is no point for him to pick a sponsoring bug if he can't upload15:24
bdrungseb128: there are 12 sponsor request he can work on15:26
seb128I'm not saying he lack sponsoring tasks to work on15:26
seb128I'm saying he could do things15:26
seb128like if libindicate soname change tomorrow15:26
seb128he could upload all the no change rebuilds himself15:26
seb128those being in the desktop set or not15:27
seb128rather than having to deal with pinging people or opening sponsoring requests15:27
seb128you are arguing that he might not use his upload right15:27
seb128so what if he doesn't15:27
seb128it still doesn't cost anything15:27
SpamapSpersia: ack15:27
seb128it can only be useful15:28
seb128if it doesn't it still doesn't hurt15:28
seb128it would be a no change compared to now15:28
seb128you block a potential win situation just because it might not win a lot?15:28
seb128I fail to see the point... let people be able to help, what they do then is their issue15:29
bdrungseb128: it's not about what he will do, but about what he has done.15:30
seb128why does it matter?15:30
pavolzetorI am back15:30
pavolzetorplease reply if there are some issues with my patches15:30
pavolzetorI think, gpm patch would have higher priority15:31
persiaseb128, It's how it's always been done: we expect work to be done through sponsors before we grant direct upload permissions.15:31
seb128bdrung, it's about having confidence people are ready for uploading or not15:31
bdrungseb128: is it enough to show that you have packaging skills or should you show that you can work in specific area?15:31
seb128well nobody has worked on everything15:31
persiapavolzetor, Be aware that it sometimes takes a while for folk to give feedback.  If you've subscribed the sponsors and/or made a merge proposal, someone should review it before too long.15:31
seb128even debian doesn't ask you to work on everybody before being a DD15:31
=== bjf[afk] is now known as bjf
persiaDebian and Ubuntu have very different models.15:32
seb128they just ask you to show you have the technical skills and social understanding to do what you want to do15:32
seb128persia, did you read what I wrote a few minutes back?15:32
seb128like being able to handle a soname change transition15:32
pavolzetorI think moths are too long15:32
kenvandinei think what seb128 is saying, it isn't likely that anyone will assign a bug with a patch on it to review if i can't upload it15:33
pavolzetorI am mainly interested in android, but this is my part-hobby15:33
* ogra_ac doesnt get this discussion, 128 sponsored uploads on maverick-changes dont qualify for upload rights ?15:33
persiaogra, Where are the 128 sponsored uploads?  remember that kenvandine has upload access to a number of packages.15:34
bdrungi want to see the list where kenvandine gets an upload to the core-set sponsored15:34
ogra_acpersia, on maverick-changes in my evolution if i search for his name15:35
seb128bdrung, he doesn't, he's working on desktop15:35
ogra_acwhy does he need to ?15:35
seb128bdrung, sure we don't ask to help with transitions out of the desktop set because he can't upload15:35
persiaogra, I'm having trouble finding anything sponsored.  128 seems about right based on what I see.15:35
seb128persia, I've sponsored at least 25 uploads for him last cycle15:35
seb128random number but he pings me at least once or twice a week every week15:36
ogra_acso he is obviously able to do proper packaging15:36
seb128it does cost me non null time15:36
persiaseb128, Do you happen to remember what one of them was?15:36
bdrungseb128: can you give us the list of these uploads?15:36
seb128persia, bdrung: gtk3 today15:37
=== dholbach_ is now known as dholbach
seb128nov. 02 22:19:09 <kenvandine>seb128, can you sponsor lp:~ken-vandine/ubuntu/maverick/x264/maverick-proposed15:37
seb128nov. 09 20:41:23 <kenvandine>seb128, btw i am just running ubuntu-geoip through pbuilder for natty, then it will be ready to sponsor :)15:37
seb128nov. 09 21:37:51 <kenvandine>seb128, lp:~ubuntu-desktop/ubuntu-geoip/ubuntu is ready to sponsor15:37
seb128I might have forgotten about x264 in fact15:38
seb128nice example of thing that would have been uploaded if it was not blocked on me15:38
kenvandineseb128, i think you did, and pitti saved the day :)15:38
seb128ok, great, I though I might just have zapped it ;-)15:38
=== al-maisan is now known as almaisan-away
bdrungseb128: gimme the list (pointing to the uploads like https://launchpad.net/ubuntu/+source/gtk+3.0/2.91.4-0ubuntu1 ) would convince me15:39
bdrungseb128: this list is missing on his application15:40
seb128bdrung, I changed laptop recently and don't have IRC logs over this month now15:40
seb128bdrung, I don't even see why you need it15:40
seb128ken clearly has showed he can be trusted with uploads15:40
seb128which does it matter if he use that right often or not?15:41
seb128it's wasting us hours every cycle to not have him having upload rights, that's not like we could use with extra hands15:41
geserfor me having a list of sponsored uploads (direct or indirect (e.g. requested rebuilds)) would be a proof that someone has a "need" for core-dev uploads right and not just the skill but no current use for them15:41
persiaseb128, It's not just about kenvandine: it's about every applicant.15:41
* ogra_ac really doesnt get that discussion, we have someone who *obviously* is good at packageing, he offers to help in other areas and we refuse ?!?15:42
bdrungseb128: that's the Chicken or the egg dilemma15:42
seb128persia, well same applies to everybody15:42
ogra_acwe havent been that picky in the past15:42
seb128why should somebody with technical and social skills to be part of the team not be in?15:42
geserogra_ac: we haven't had packagesets in the past15:42
seb128just because it doesn't help every day?15:42
persiaseb128, Sure.  Historically, we've expected people to demonstrate their work by getting sponsored.  I'm not sure that's broken.  I'm having difficulty finding recent examples of that happening for kenvandine.15:42
ogra_acgeser, no, but do we need to add extra bureaucracy ?15:43
kklimondafor some reason the discussion reminds of the medieval age and apprenticeship from back then...15:43
seb128persia, it doesn't work lot out of desktop, does it mean it shouldn't be welcome?15:43
persiaThere are *other* reasons kenvandine would benefit from being core-dev, but I'm not yet convinced they are sufficient (uploading new packages, working on stuff in universe).15:43
seb128persia, I would be assigning him desktop tasks I can't assign him now if he could upload15:43
Riddellmvo: I don't suppose you know what's wrong with this upgrade bug 680088 ?15:43
ubottuLaunchpad bug 680088 in update-manager (Ubuntu) "Upgrade fails "Can not mark 'kubuntu-desktop' for upgrade "" [Undecided,New] https://launchpad.net/bugs/68008815:43
ogra_acif someone has obviously the tech skills it should really be a non-issue15:43
seb128persia, he would spare me sponsoring work as well15:43
geserseb128: so everyone with the needed skills should become core-dev because he *might* need them in future?15:44
ogra_acif he/she asks for them, why not15:44
seb128everybody who has contributed to ubuntu in a significant way yes15:44
ogra_ache/she can be a patch pilot15:44
seb128that's what debian does and what we used to do15:44
ogra_acetc15:44
seb128uploads right doesn't hurt15:44
seb128they can only be useful15:44
ogra_ac++15:44
seb128he might sponsor once a month15:44
bdrungi think geser and i share the same concern. i have to leave now.15:45
seb128it's still a win15:45
bdrungogra_ac: you don't be a core-dev to be patch pilot15:45
seb128kenvandine, has uploaded hundred of packages to main15:45
persiaYou don't even have to be a developer15:45
seb128it's not like he was not contributing15:45
persiaIt's not about "main".  It's about "core"15:45
seb128same differenc15:45
persiaExcept it's not: it's also about new packages and universe, which complicates the picture.15:46
seb128being able to help where help is needed15:46
ogra_acbdrung, you need to be core-dev for being a patch pilot for all core-dev packages15:46
persiaNo, not at all.  There's *lots* of stuff in main that isn't in core.15:46
ogra_acits only a win15:46
seb128we need people who have skills to help out of their usual set15:46
persiaogra, No you don't: a huge number of patches are better addressed upstream.15:46
ogra_acthe criteria should really be tech based15:46
ogra_acrarther than artificial political issues imho15:47
ogra_acif he is technically capable and wants to help, let him do15:47
persiaogra, Which artificial political issue?15:47
cjwatsonfor the record, the patch pilot programme is intended to include things like poking people to follow up on patches, making sure people get a good first response and are directed to the right place to discuss their work, etc.  there's no need to drag that programme into this.15:47
ogra_acpersia, dunno, obviously there is no tech issue blocking here15:47
geserit's not political for me; I just prefer to give a person the rights that he really needs (just like access in security)15:47
seb128geser, well it means you are blocking all people working on specific sets to help15:48
seb128you should remove my rights to upload out of desktop going this way15:48
mvoRiddell: I commented in the bugreport, it complains about libkephal4/plasma-netbook15:48
seb128I'm doing 95% of my work in desktop as well15:48
seb128I'm just not sure what you would win by doing that15:49
* ogra_ac doesnt want to have such a discussion for every member of the arm team in the future 15:49
ogra_acand we dont have a single package set15:49
seb128I think I will just recommend desktop people to not try to work out of desktop15:49
seb128seems that's not welcome15:49
ogra_aci.e. every team member would have to be core dev15:49
ogra_acyeah15:49
seb128way to drive contributors away15:49
ogra_acits a weird setup imho15:49
persiaseb128, Rather, if there is history of work outside desktop, upload rights outside desktop are easily granted.  Without any prior history, it's uncertain.15:50
seb128"sorry but they don't like people who focus on a set"15:50
persiaThat's not at all the case.15:50
seb128it is15:50
geserseb128: I'm not strict on how many uploads one does outside his package set, but at least show that ones does them15:50
* persia is having trouble finding *one* for maverick or natty15:50
ogra_acpersia, but why has there to be any history ? tech skills should be enough15:50
seb128persia, I just gave you 315:50
seb128persia, for kenvandine15:50
kenvandinegtk3 today15:50
seb128gtk3 today got rejected15:51
seb128x264 in maverick-proposed15:51
ogra_acand the expressed will to help out with more15:51
persiaAh, rejected.  Excellent.15:51
persiaWhy was it rejected?15:51
kenvandineperms15:51
kenvandinegtk3 isn't in the desktop set yet... so now i need seb128 to sponsor it for me15:51
kenvandineeven though i applied almost the same patch to gtk2 and uploaded that15:52
chrisccoulsongtk2 isn't either is it? seb128 had to do a gtk2 upload for me last week15:52
kenvandineso until gtk3 gets added to the package set, i need to interupt someone with changes15:52
kenvandinechrisccoulson, i uploaded that fine on friday15:52
chrisccoulsonoh15:52
geserthat would be proof enough for me (as I expect that this won't be the only upload to the core set)15:52
seb128Signed-By: Didier Roche <didrocks@ubuntu.com>15:52
seb128https://launchpad.net/ubuntu/maverick/+source/compiz/1:0.8.6-0ubuntu815:52
seb128^ another one15:53
chrisccoulsoni've got no idea what i can upload really, it seems a bit hit-and-miss15:53
kenvandineedit_acl.py is your friend15:53
kenvandine:)15:53
geserdesktop might be a little bit special in this regard thanks to the intermix with core15:53
chrisccoulsonkenvandine, oh, you're right15:53
kenvandinegeser, right... ultimately the stuff i work on could be anything that ends up on the desktop CD15:53
chrisccoulsonso, i could have uploaded it myself then15:54
persiachrisccoulson, `./edit_acl.py -P ubuntu-desktop -S natty query` would show you the list of desktop packages.  Dunno if you have other packagesets offhand.15:54
cjwatsonthis bites desktop a lot because there's a lot of overlap between Ubuntu and Kubuntu desktop that one or the other side maintains15:54
seb128Signed-By: Martin Pitt <martin.pitt@ubuntu.com>15:54
seb128https://launchpad.net/ubuntu/maverick/+source/x264/2:0.98.1653+git88b90d9-3ubuntu115:54
persiageser, What intermix?  The exceptions lists?15:54
geseryes, that several packages that "belong" to the gnome desktop are in core15:55
geserI guess this is less a problem for e.g. mythbuntu15:55
seb128well still, we could assign him sponsoring bugs to review for example if he was able to upload15:55
seb128or helping on transitions for soname changes15:55
geserI'm happy with any proof that one will use that uploads right and just doesn't apply because it's "nice to have"15:56
ogra_acor any other transitions15:56
=== mrenouf_ is now known as mrenouf
seb128those cases don't show upload in uploads before because they are tasks that don't make sense when you need a sponsor15:56
geserthat's part of how I understand the whole ArchiveReorg15:56
seb128you create a chicken egg issue there15:57
=== almaisan-away is now known as al-maisan
seb128we can't get people to help on those because to be able to help they need upload rights15:57
evsuperm1: I'd like to keep things consistent with the existing interface, if possible and avoid superfluous checkboxes.  We should pick a singular outcome and structure the documentation around that.  This is automatic partitioning, after all.15:57
evoh, when you said checkboxes did you mean bullet points?15:57
superm1well i was thinking checkboxes, but bullet points might make more sense if it is automatic15:58
geserseb128: that mustn't be necessarily sponsored uploads, I would trust a word from you that you sponsored uploads for him (without any visible LP record915:58
seb128I sponsor at least an upload for him a week15:58
superm1but the important bit was relaying what the outcome would be before doing it since it could have different results depending on web access and for different people15:58
seb128that every week in the cycle15:58
geserfor me that would be sufficient15:58
geser(the word from you)15:58
seb128we add comments from pitti and me at least on kenvandine's wiki15:59
seb128add -> had15:59
kenvandineand didrocks15:59
kenvandineall wrote endorsements :)15:59
seb128not sure what else is needed there15:59
kenvandinethx guys15:59
seb128kenvandine, np ;-)15:59
persiaExamples of work would have been nice, and saved the confusion.  Turns out my issue was not querying on the "desktop-core" set.15:59
ogra_acto me this process looks really scary and broken to be honest16:00
geserI didn't read the application page yet and didn't attend the DMB meeting, will follow-up with my vote to the DMB later16:00
seb128geser, thanks16:00
seb128still seems artificially hard16:00
ogra_ac++16:00
seb128kenvandine does over an hundred upload a cycle and had endorsement from 3 active ubuntu members16:00
persiaogra, The process is basically: do something, get sponsored, ask for upload rights, get upload rights to the stuff you had sponsored work on (typically by packageset)16:01
ogra_acpersia, yeah, thats broken16:01
seb128well it doesn't solve the need to have people being able to work on things out of specific sets16:01
seb128or to help there16:01
persiaThis oughtn't be overly hard or complicated.  Sometimes the tools are confusing, or we make mistakes.16:01
ogra_acif i have the skills and am willing to help out, even if i didnt yet, i should be granted access16:01
persiaogra, Which part is broken?16:01
seb128especially that set are buggy16:01
ogra_acthe tech skills should be all that rules here16:01
seb128we have chrisccoulson and kenvandine hitting set issues every week16:02
seb128it's costing us a lot of energy16:02
cjwatsonI'm not being asked about this every week16:02
cjwatsonand I wonder why not16:02
seb128those might not been "issues" than "technlogogies used in kubuntu, xubuntu, etc as well"16:02
persiaseb128, If the set is buggy, we should fix that.  If we're having consistent set issues, we should describe the class, and fix that.  Fixing on a per-person basis doesn't scale.16:02
kenvandinecjwatson, just us needing sponsoring16:02
seb128cjwatson, because often those are cross desktop techs16:02
seb128dbus, libnotify, etc16:02
cjwatsonseb128: ok ... but you're saying the set is buggy.  which is it?16:02
cjwatsoneither it is buggy, in which case the bugs should be reported, or it is not16:02
seb128"buggy" might be wrong16:02
kenvandinebuggy might not be the word16:03
seb128it's just that we touch techs which impact on other sets16:03
ogra_acthe policy is buggy imho16:03
cjwatsonogra_ac: conflicting requirements16:03
cjwatsonmakes it hard to keep everyone happy16:03
seb128cjwatson, ideally chrisccoulson and kenvandine would have access to cross desktop techs16:03
persiaogra, How do you recommend we judge other than on prior work?16:03
pittiseb128: (which would make them eligible for core-dev IMHO)16:03
cjwatsonone suggestion would be to have both Ubuntu and Kubuntu desktop folks have upload access to the desktop-core set16:03
ogra_acpersia, do a techincal only judgement of the skills16:03
cjwatsonthat might ease the pressures without being one-sided16:04
kenvandinethe set probably is good for most people, but some of us touch a wide array of pieces16:04
persiaseb128, That's being core-dev, which is fine.16:04
persiaogra, Based on what?16:04
seb128pitti, well, the issue there is that persia, geser and bdrung think kenvandine doesn't work enough out of desktop to warrant that16:04
ogra_acpersia, exposing the proper skills of former work and expressing the will to help should be enough16:04
cjwatsonwait, geser doesn't seem to have expressed a definitive opinion yet16:04
persiaseb128, I said I couldn't *find* the work.  I've found it now (I was looking in "core", and it's in "desktop-core")16:04
seb128persia, not it's not, we are arguing over an hour now because you said kenvandine's doesn't need it16:04
cjwatsonand persia explicitly abstained16:04
chrisccoulsonhaving access to desktop-core would help me quite a bit. most of the stuff i find i can't upload is in desktop-core16:04
cjwatson(actually, bdrung also abstained)16:05
seb128ok, so maybe what we need is desktop members to have access to desktop-core16:05
seb128would that be doable?16:05
geserseb128: that's not what I been trying so say, all I want is some proof that someone is working outside a package set16:05
persiaseb128, I never said that.  I said that from what I could find, he hadn't done *any* work in the area to which he wished to be granted upload access.  This turns out to have been a problem with my tools.16:05
cjwatsonI think the DMB would have to agree on it, but it's certainly technically possible.  As I say I think that it should be both Ubuntu and Kubuntu though (the desktops in main)16:05
seb128persia, ok16:06
cjwatson(meeting)16:06
kenvandinedoesn't solve new packages and such16:06
kenvandinebut would cover more for sure16:06
persiaI'd rather keep that to core-dev, personally (and I don't mind granting core dev to people working on that set, with some demonstrated work)16:06
seb128right, it's still annoying in quite some case16:06
seb128like the GNOME3 new sources landing regularly in natty16:06
* ogra_ac totally agrees16:07
Laneythey can easily be added to the sets16:07
kenvandinenot before they are uploaded16:07
chrisccoulsonkenvandine, yeah, that probably helps me more than you, because i can already upload to universe :(16:07
persiakenvandine, Indeed: it's the new packages problem and the not-yet-in-the-packageset problem that made me almost vote for you to be core-dev *even though* I couldn't find the examples.  I'm glad I found them.16:07
seb128speaking of which what is the official way to have things added to a set?16:07
seb128out of pinging cjwatson on IRC ;-)16:07
persiaseb128, It's supposed to be ask an archive-admin, but in practice it's ask cjwatson.16:07
Laneynot archive-admin, member of owning team16:08
chrisccoulsonand it doesn't help the fact that i still can't target bugs in packages that i can upload (or maintain) to a particular release16:08
seb128I was asking to try to avoid pinging cjwatson directly16:08
persiaAnd ideally, the archive-admin request is in a bug.16:08
seb128chrisccoulson, you should apply for core rights16:08
persiachrisccoulson, That7s a different issue.  Isn't it targeted to get sorted soon?  I thought I heard things at UDS about it being on the short-term list.16:08
geserany TB member can add packages to TB owned packagesets (the are only DMB owned packagesets)16:08
LaneyYou can email the owner of the set (I think they are all TB or DMB owned)16:08
seb128Laney, where do I find the owner of a set?16:09
chrisccoulsonseb128 - i was thinking about it until i saw this discussion today ;)16:09
persiaDid the bug in LP that limits changing stuff to the TB get fixed?16:09
seb128persia, see what you did :p16:09
chrisccoulsoni think that the DMB members will have all the same objections with my application (i do 95% of my work in desktop or mozilla)16:09
seb128you made chrisccoulson not to apply! ;-)16:09
geserseb128: through the LP API (owner attribute of the package set)16:09
Laneyseb128: set.owner attribute16:09
Laneyactually I patched edit_acl to display that locally16:09
Laneycan push a branch if you want to merge it16:09
seb128hum16:09
seb128I will let other people deal with that16:10
seb128but seems the reply is "there is no easy way out of writting code to query launchpad"16:10
persiachrisccoulson, We don't care about the 95%: we care about the 5%: if you have work that you did in core or desktop-core, and you list it on your application page, there will be little confusion, and you'll just get approved or rejected for something other than the areas of your work.16:10
Laneythe LP web UI is lacking when it comes to package sets, indeed16:10
seb128chrisccoulson, joke aside you should apply16:10
LaneyI would love the +source page to display them for example16:10
seb128I think the confusion was mostly that they didn't consider desktop-core16:10
=== makl is now known as ximion
seb128chrisccoulson, should be easier for you now that this one is sorted16:11
persiaLaney, That7s been requested, for the Soyuz ArchiveReorg spec from this last UDS.16:11
Laneypersia: yeah, I know (or I'd be filing a bug)16:11
=== ximion is now known as makl
=== makl is now known as ximion
gesersome web-UI for package sets would be nice (a little overview and the like)16:11
persiaseb128, My confusion was that I didn't consider desktop-core.  I can't speak for anyone else (and some DMB members voted in favour, just not enough to confirm during the meeting)16:12
* ogra_ac still would like to rather see the policy fixed16:12
persiaogra, Would you prefer to grant upload access to people with no demonstrated experience?16:12
ogra_acpersia, no16:12
persiaThen why do you suggest that we shouldn't use demonstrated experience as a criterion?16:13
seb128"not uploading out of a set" != "not showing experience"16:13
ogra_acpersia, but if i have rights for packageset A and have proven that i'm technically capable, requesting sponsoder proof for packageset B if i express interest in that packagset is nonsense16:13
seb128you can do all your work in a set and demonstrate enough experience to be trusted with uploads16:13
geserogra_ac: I don't believe that the policy has changed much, only that the packagesets made it more hard to show the expierence in the area where one wanted to have upload rights16:13
ogra_acthe *technical* skills should be all that matters here16:14
persiaI don't think it's more hard.  It's the same.16:14
persiaogra, So everyone should be core-dev, with no distinctions?16:14
seb128geser, well we should be able to say that someone showed enough skills to be able to work out of one specific set16:14
ogra_acits not, since you refused to grant rights without sponsored proof right above16:14
seb128persia, those who showed enough experience yes16:14
seb128if they ask for it16:14
persiaseb128, How do we judge that except by them showing they did work outside the set?16:14
seb128if people are happy in their set no16:14
ogra_acpersia, everyone who has the skills sould be *easily able to* become core-dev, yes16:15
seb128persia, how is the work out of the set different from the work in the set?16:15
ogra_acyou are making the process overly complex16:15
ogra_acwhile only the tech knowledge should matter16:15
seb128persia, the packaging doesn't change because you are out of a set16:15
persiaseb128, Shows integration with wider communities of developers.  Shows consideration for other packages that depend on the packages being worked upon.16:15
ogra_acand the tech knowledge was judgeable by existing work16:15
ogra_acsponsored or not16:15
seb128you still use the same infrastructure, debian packaging, launchpad, etc16:15
persiaNo?  Compare Java packaging to GNOME packaging some day :)16:16
seb128persia, java-gnome bindings?16:16
seb128you have java in desktop as well16:16
persiaThat's packaged GNOMEy, but sure.16:16
ogra_acpersia, know what ? i'm core-dev and wouldnt touch java packaging with a ten foot pole16:16
ogra_acdo you want me to withdraw from core-dev now ?16:16
ogra_acinto some package set ...16:17
seb128I was asking the same before16:17
geserno16:17
cjwatsonthis is getting a bit emotive16:17
persiaogra, No.  You've demonstrated that you won't touch Java with a 10 foot pole, so I feel safe about you and java16:17
ogra_accjwatson, i'm cool sorry if i dont sond like16:17
ogra_acpersia, right, so why should that matter for someone becoming core-dev then16:17
ogra_acask him/her in the DMB meeting if he/she would touch java16:18
seb128persia, well you can show that you are reasonable and would not touch things you don't understand while working in a set16:18
ogra_acbut judge his/her tech skills based on existing work16:18
ogra_acwithout the requirement to having sponsored packages to the future set he/she is intrested in16:18
ogra_acthe existing work should be enough for that16:18
dholbachI think trust is important - if I trust somebody to know where to stop and where to ask that's one of the most important things (plus statements from others who worked with them, etc.)16:19
ogra_acright16:19
ogra_acmore important than sponsored uploads16:19
ogra_acand trus is something you have to build up *without* the tech skills16:19
persiaTrust is indeed important, but I have to wonder how much trust someone has if they can't find any work to do or anyone to sponsor their work in their target area.16:19
ogra_acand whichch should be based on testimonials16:19
persiaAnd while I try to know everyone, I know that I fail, and I am certain I'm not alone in that failure.16:20
ogra_acsure, and you dont have to16:20
ogra_acthats why we got wiki pages16:20
gesershould everyone from the the "kernel" set get upload rights for "core" because we trust them with kernels?16:20
seb128persia, taking kenvandine's case you got pitti didrocks and me saying he does nice work16:20
ogra_acgeser, if he knows packaging good enough, yes16:21
seb128geser, no, but when you get 3 people from the set giving you recommendation you would consider that enough16:21
pittigeser: do you trust me with kernels? (you certainly shouldn't)16:21
persiaI'd much rather focus on making it easy for folks to demonstrate their work in other areas than make granting of upload rights a popularity contest.16:21
ogra_acindeed for the kernel thats hard to judge, there wont be many different packages to review16:21
pittigeser: just as a point that there can't ever be the perfect core-dev16:21
apwheh, you want to make sure this conversation is stricken from the logs, else noone will ever bother applying for core-dev again16:21
pittiin practice, everyone knows just a subset of everything16:21
ogra_acpitti, nontheless you can upload one if you want ;)16:21
geserogra_ac: then why do we the whole package set business at all? just having motu and core-dev like in the old days would be much easier16:21
pittiwell, unless you are cjwatson, of course16:21
persiaogra, So what kernel folk do is peer-review for the first few uploads, and then bring the demonstration to the DMB, and get a new kernel uploader.  It's just collaboration.16:22
seb128geser, because some people are desktop contributors only16:22
seb128they don't want or need anything else16:22
seb128we have people who want to be able to do GNOME updates16:22
seb128it's an easier way to start16:22
ogra_acgeser, i dont say package sets are wrong16:22
seb128desktop being once example16:23
ogra_acbut the artificial threshold you put in place is16:23
seb128some people might be just interested in xubuntu16:23
=== oubiwann is now known as oubiwann-away
persiaseb128, So, if you have folk in the desktop team who need wider access, just make sure they get some stuff sponsored (you say you do this already), make sure that the examples show on their applications, and they are likely to easily be approved as core-dev.16:23
seb128then you get people over time that might feel like doing work out of the set16:23
=== oubiwann-away is now known as oubiwann
persiaWith no examples, and difficulty finding any work, it's a more complicated discussion.16:23
seb128persia, well I still fail to say why we need those if you have several trusted people who worked with them and said they would be useful contributors out of the set16:24
ogra_acgeser, if there is enough work to judge on, why do i need sponsored uploads to the target set i want to enter ? i have already proven my skills enough before16:24
seb128persia, gotcha on how to get people approved easily16:24
persiaAnd the example that raised this is not a good one, as it came from a misunderstanding on one person's part as part of an *incomplete* DMB vote.  No application was deferred.16:24
seb128I still think you are putting some requirements which are not needed16:24
ogra_acpersia, no, its a good one16:25
kenvandinethe tough thing for me is there are lots of examples of my work that didn't need sponsoring, because of the the package set16:25
seb128persia, you might have people who could help with transitions or sponsoring16:25
ogra_acsince you guys asked for sponsored uploads to the future target set16:25
ogra_acwhich is a big flaw imho16:25
seb128they will just not do as long as they can't upload16:25
seb128because they then need to find themself a sponsor16:25
persiaogra, Why?  it7s an *incomplete* vote.  Even without this discussion, kenvandne may have become core-dev.16:25
seb128which is costing as much work they spared16:25
geserogra_ac: so everyone who can package gnome ("desktop") should get access to "kubuntu" just by asking? because he knows how to package?16:25
ogra_acgeser, if he can prove he knows how to package, why shouldnt he ?16:26
seb128geser, if we trust him to not upload without asking there, etc why not?16:26
persiaseb128, I actively don't want folk working on sponsoring if they have no experience in the target area.  I7d rather they chase the random patches not ready for sponsoring and get them tested and in shape.  There's *lots* of non-sponsoring work to be done, and I don't think the primary activity of developers ought be sponsoring (although sponsoring is a welcome part of it).16:26
ogra_acif i'm able to fix a bug in gtkmm, why shouldnt i be able to fix a bug in QT ?16:26
cjwatsonpitti: (I steer *well* clear of desktop, FWIW ...)16:27
seb128persia, well then ignore sponsoring and take transitions as an example16:27
ogra_accjwatson, it was about abilities ;)16:27
pitticjwatson: nevermind, it was just a compliment :)16:27
persiatransitions and packagesets interact very badly, unfortunately.16:27
\shpersia: +116:27
geserogra_ac: I believe that only knowing how to package is not enough; I'd expect that he also knows how "kubuntu" works (how the packages interact each other)16:28
seb128persia, see you hit limitation which block people in doing work and helping16:28
ogra_acgeser, how does that matter to fix an ftbfs in a single package ?16:28
seb128would it be launchpad issues, new sources, set definitions bugs, transitons16:28
ogra_acgeser, i touched enough kubuntu packages in my ubuntu life and have no clue at all how kubuntu works16:28
mvosmoser: re #676790 - this is the natty apt?16:29
ogra_acnontheless i bet people in kubuntu were happy when their packages built16:29
smosermvo, yes.16:29
\shogra_ac: we all touched a lot of packages where we didn't know how that all works...but that's long time ago...16:30
cjwatsonKeybuk: (from #ubuntu-meeting) I don't pretend that this would be anything other than a nasty hack, but given how the kernel presently behaves it seems to me that the only thing init can do is try a few times in the hope of winning the race eventually.  Obviously you don't want to block starting other events on that16:30
geserogra_ac: it's also about trust; I don't know how to properly explain it and I'm myself not fully sure where I draw the borders16:30
cjwatsons/other events/other jobs/16:30
ogra_acgeser, so make it a requirement that someone from the desired package set expresses trust in the wikipage16:30
mvosmoser, pitti: it sounds like bug #676790 is releated to the compressed indexes16:30
ubottuLaunchpad bug 676790 in apt (Ubuntu) "huge performance regression" [High,Confirmed] https://launchpad.net/bugs/67679016:30
smoserbased on nothing but changelogs, i suspect it has to do with the compression16:30
pittiright16:31
smoseryeah, that was my suspicion too16:31
Keybukcjwatson: yeah, but in theory init is going to proxy everything anyway16:31
pittiI just don't know yet what these image builds do with apt indexes16:31
ogra_acbut dont make sponsored uploads a requirement to people who can prove their knowledge already16:31
cjwatsonKeybuk: remind me: would anything go wrong if pid 1 kept its own master fd on /dev/console (with O_NOCTTY of course) and duped it, rather than reopening for each job?16:31
\shmvo: good that I see you...I don't know if you are responsible, but /etc/cron.daily/apt does some strange apt-key net-update thingy...and it starts wget to get a keyring..without a sane timeout...can we change something there?16:31
Keybukso the fd won't be console anyway16:31
Keybukit'll be a pty16:31
cjwatsonKeybuk: this is lucid16:31
Keybukand then it'll output that to console later16:31
smoserpitti, its nothing strange, i would suspect that a CD install would hit this16:31
Keybuklucid we fixed already16:31
cjwatsonno we didn't, per bugs16:31
Keybukat least fixed well enough16:31
smoserits basically an 'apt-get install some package list'16:31
Keybukwhat are the bugs now?16:31
cjwatsonthis is why I brought up bug 64255516:32
ubottuLaunchpad bug 642555 in Ubuntu Lucid "Services not starting on boot in 10.04.1 LTS" [Medium,Confirmed] https://launchpad.net/bugs/64255516:32
Keybukcjwatson: yes, if you SysRq-K you get a kernel panic16:32
cjwatsonin the meeting where you responded16:32
Keybukif Upstart keeps its own console fd16:32
cjwatsonI assumed you'd seen it16:32
cjwatsonok16:32
Keybukno, I'm not reading bugs anymore16:32
mvo\sh: sure, is there a bugreport?16:32
geserogra_ac: for me sponsored uploads are only one way to proof that, a comment from the set "leader(s)" is for me also sufficient16:32
cjwatsonI assumed you'd seen it *in the meeting log* since you responded there16:32
cjwatsonI know you aren't reading bug mail16:32
\shmvo: not right now, I wanted to talk to you before :) because some of my servers do have no direct internet connection....and I always see this process hanging and timeouting after several hours whysoever16:33
ogra_acgeser, why leaders ?16:33
mvo\sh: heh :)16:33
Keybukoh, no, I just have certain highlights ;)16:33
KeybukI was coding16:33
mvo\sh: sure, let me quickly check the code16:33
ogra_acand why cant you judge by existing work ?16:33
\shmvo: especially when the firewall drops those packages16:33
\shmvo: I think there needs to be just a --timeout option to wget (60 secs should be enough imho)16:34
seb128is grep-dctrl working for other people in natty16:34
seb128?16:34
cjwatsonthere is a lot of plymouth-hatery in that bug, which makes it difficult to pick out the real issue16:35
Keybukwell, yeah16:35
Keybukor if there is actually an issue16:35
geserogra_ac: perhaps leaders is the wrong word, more "key persons"; persons who I'd trust from that team and not somebody who just entered into the team in the last meeting or so16:35
cjwatsonbut making sure that there's a console fd would dismiss that part of the question for good16:35
ogra_acgeser, agreed, still though, technical knowledge should be judgeable by former work already, i agree on the trust thing16:36
chrisccoulsonseb128 - it doesn't work here, but i guess that's because the package lists in /var/lib/apt/lists are stored compressed now16:37
seb128chrisccoulson, I was wondering about that yes16:37
seb128iz pitti bog? ;-)16:37
chrisccoulsonheh:)16:38
pittichrisccoulson: "it" ==?16:38
pittioh, grep-dctrl16:38
chrisccoulsonpitti - did you break grep-dctrl? l;)16:38
geserogra_ac: for extending permissions, the technical knowledge often isn't the hard part, for those applications I watch more on comments from "key persons" from the team where one applies16:38
chrisccoulson:)16:38
pittichrisccoulson: presumably; can you please file a bug about it and assign to me?16:38
pitti(sorry, I'm in meeting, and then off)16:38
chrisccoulsonpitti - yeah, sure16:38
ogra_acgeser, right, what seemed wrong to me above was just the requirement that ken had sponsored uploads to the core-dev set16:39
seb128chrisccoulson, I can do it if you want16:39
seb128chrisccoulson, I was the one who asked about it ;-)16:39
chrisccoulsonseb128 - yeah, feel free to do that :)16:39
seb128chrisccoulson, ok16:39
=== dendrobates is now known as dendro-afk
=== jelmer is now known as Guest34932
=== Guest34932 is now known as jelmer
cjwatsonKeybuk: I posted to that bug to try to narrow things down a bit16:53
pittiev: FTR, the UI works, but I have spent an hour trying to track down why the communication with the backend locks up; presumably a threading deadlock, so I pushed the stuff to a branch for now; will get back to that tomorrow16:57
evthanks16:58
evindeed, I should've never gone down the path of threads with that application.16:58
pittiit's a pain to debug indeed, but let's see16:58
pitticould be that gi/pygobject just isn't thread safe yet, I'll investigate16:59
pittigood night everyone17:00
evg'night pitti17:00
cjwatsonpitti: OK if I just accept this d-i upload?17:04
bdrunggeser: you said that you found sponsored work from kenvandine - how?17:06
=== dendro-afk is now known as dendrobates
ari-tczewbdrung: tumbleweed has a script to looking sponsored bugs17:14
mvo\sh: commited the timeout fix to bzr now, thanks for the report!17:15
\shmvo: thx to you...(and I didn't file any report ;))17:16
mvo\sh: no worries17:17
=== deryck is now known as deryck[lunch]
geserbdrung: did I? persia found it in "desktop-core"17:18
Sarvattis there a way to mark a bug fix in a sync request without doing it in debian's changelog?17:21
Laneyyou can make that bug into the sync request bug17:23
persiabdrung, I screen-scraped LP to get kenvandine's uploads, and then compared with the output of edit_acl querying the desktop-core set, and found matches.17:24
seb128zul, hi17:26
zulseb128: hey17:26
seb128zul, do you think you could sru the fix from https://bugzilla.samba.org/show_bug.cgi?id=7791?17:26
seb128it's bug #39301217:26
persiaRAOF, pitti Welcome to the sponsors team :)17:26
ubottuLaunchpad bug 393012 in autofs "smb: Error while copying file, "Invalid argument"" [Undecided,Confirmed] https://launchpad.net/bugs/39301217:26
seb128zul, https://bugzilla.samba.org/attachment.cgi?id=6062&action=view17:27
zulseb128: sure17:27
seb128zul, thanks17:27
\shoh wow...attachmate wants to take over Novell, and Microsoft wants to get hold on some IPs of Novell...oh wow...this will be fun and I can already read all the FUD news17:27
seb128zul, we got quite some users unhappy because smb copies from nautilus are broken in maverick, that should fix it ;-)17:27
cjwatsonpitti: ... accepting, figuring it's OK17:28
* smb does not like his copies to be broken17:28
zulseb128: gotcha...but they shouldnt be running windows in the first place ;)17:28
seb128;-)17:28
seb128zul, thanks!17:28
ogra_acgah, damned, no way back, i'm in the ubuntu-sponsors team17:30
\shogra_ac: lol...you were in the past, too..well not sponsors, but MOTU17:31
ogra_ac(thats like the patch pilot flight license, right ?)17:31
ogra_ac\sh, indeed17:32
persiaogra, The main power of the sponsors team is the ability to unsubscribe the sponsors team from things no longer of interest.17:33
ogra_acpersia, patch pilot license, as i said ;)17:34
=== dendrobates is now known as dendro-afk
=== dendro-afk is now known as dendrobates
\shogra_ac: I hope the patch pilots don't have problems like the A380 pilots ;)17:40
ogra_acwe dont use rolls royce ;)17:41
ogra_acwe are launchpad powered instead ;)17:42
=== tkamppeter__ is now known as tkamppeter
slangasekSpamapS: bug #582963> so what does the apache2 ask-for-passphrase do if plymouth shuts down in the middle of prompting?17:47
ubottuLaunchpad bug 582963 in apache2 (Ubuntu Maverick) "SSL pass phrase dialog can't read input" [Medium,In progress] https://launchpad.net/bugs/58296317:47
geserogra_ac: and no eject button :)17:51
ogra_acheh17:51
Kmosdate17:54
=== dendrobates is now known as dendro-afk
=== dendro-afk is now known as dendrobates
gesermvo: added TEST CASE to bug 61788518:01
ubottuLaunchpad bug 617885 in gparted (Ubuntu Maverick) "gparted crash at start: glibmm-ERROR **" [High,In progress] https://launchpad.net/bugs/61788518:01
gesermterry: can you please do a vala-dep-scanner upload soon? so we are one rdepends nearer to kill libvala-0.10{,-dev}18:04
mterrygeser, ah ok!  a new version is coming out shortly, so yeah18:16
geserthanks18:16
=== deryck[lunch] is now known as deryck
=== didrocks1 is now known as didrocks
ogra_acstgraber, so i managed to get lightspark to build up to the point where it starts linking against the intel SSE assembler code (where it indeed fails). from a packaging POV its easy to add armel support, but the assembler is a huge lot18:29
stgraberogra_ac: ok, cool. And I guess there's no C equivalent for that assembler code (that'd likely make it "unoptimized" but at least work) ?18:30
ogra_aci dont think so18:30
ogra_acthe assembler really hooks deeply into SSE, rewriting that is beyond me18:32
stgraberis that something anyone at Linaro might be interested to work on ?18:33
stgraberI guess it depends on how much you really want a flash implementation and of ongoing discussions with Adobe by the manufacturers (I'm guess there's some as unfortunately flash is usually kind-of a must have :()18:34
ogra_acwell, i guess the manufacturers get access to flash internally18:37
=== yofel_ is now known as yofel
ScottKcjwatson: re  bug 635273, I don't have a strong opinion.  Depending on what the fix it, it might be ~OK.  I'd make barry figure it out.18:43
ubottuLaunchpad bug 635273 in python-support (Ubuntu Lucid) "Building debs with SWIG libraries do not work" [High,Confirmed] https://launchpad.net/bugs/63527318:43
=== bilalakhtar is now known as papercutter
SpamapSslangasek: you mean the race condition between ping and ask? yeah.. I thought about that when I copied it from some other package. :-/18:50
ScottKskaet: Sorry I missed the meeting, an appointment ran over.  One item of note for 10.04.2 is that the Tech Board approved putting KDE point releases in -updates so you can expect Kubuntu to update to KDE 4.4.5 (and 4.4.7 for kdepim) for 10.04.2.18:51
SpamapSslangasek: I believe apache2 will fail to start at that point.18:52
=== papercutter is now known as bilalakhtar
SpamapSslangasek: its kind of an interesting problem..  we kind of have to have this happen before getty's/X ..18:54
=== al-maisan is now known as almaisan-away
bdrunggeser: sorry. i meant persia.18:59
bdrungpersia: can you give me your findings?19:00
=== tgall_out is now known as tgall_foo
bdrungSarvatt: i think the no-source change syncs are useless19:25
Sarvattbdrung: ack, wasn't sure on those because pitti is doing no change reuploads of X packages in main to shrink the livecd size19:26
bdrungSarvatt: when will be the next cd spin?19:27
SarvattI just requested the ones in xserver-xorg-video-all19:27
Sarvattdunno, I agree they are kind of useless but was trying to help out since I saw him doing that19:28
jmelisHello, is this channel the right place to ask about packaging?19:28
bdrungSarvatt: if pitti wants the rebuilds, then give him the rebuilds. ;)19:29
bdrungjmelis: yes19:29
bdrungjmelis: maybe #ubuntu-motu is appropriate too19:29
jmelismy question is: I have a package, and I have created a system-V init script in the debian folder named <package>.init. I'm working with 10.04. The problem is that after installing the package the init script is not installed in /etc/init.d... How can I debug this? System-V init scripts are still working, or do I have to switch to upstart?19:31
persiajmelis, Are you calling dh_installinit?19:31
skaetScottK,  Riddell,   will add that Kubuntu will update to KDE 4.4.5 (and 4.4.7 for kdepim) into the minutes.   Thanks for letting me know.19:31
jmelispersia: no. where should I do that?19:31
jmelispersia: in the rules file, correct?19:32
Sarvattok I'll invalidate all the sync requests that aren't new upstream versions then. sorry for all of the sync requests but its a heck of a lot easier to do everything in debian and request a sync since I'm not a core-dev :)19:32
persiajmelis, Yes, in the rules file.  If you're using dh(1), it's probably automated.  If not, it's usually done in the install: rule, somewhree around dh_installmenu19:33
bdrungSarvatt: no, you misunderstood me19:33
jmelispersia: can you point me to a guide that explains this? I haven't had much luck finding one...19:33
Sarvatthow so? from what I understood you think it'd be better to do a no change rebuild instead of syncing something like this? https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-tdfx/+bug/68011019:34
persiajmelis, Erm, I don't tend to do it manually, but ask in #ubuntu-packaging, and I'm sure someone can take you through it step-by-step19:34
jmelispersia: thanks a lot!19:35
persiaSarvatt, If that upload was needed in Debian, doesn't make any difference which way you do it.  Note that things in sync will autosync, so sometimes no request is required.19:35
jmelisi'll give it a shot myself, if it doesn't work I'll go to #ubuntu-packaging19:35
persiaHeh, OK :)19:35
SpamapSslangasek: In looking at this, I don't know if we'll have other services like this, but in this case we have to make sure that apache starts before plymouth is stopped. Can we somehow wait in plymouth-stop for apache?19:36
* persia isn't as sure of being active in #ubuntu-packaging later19:36
SpamapSit would appear to me that plymouth quit could be called from several locations.. :-P19:36
* highvoltage somehow didn't know about #ubuntu-packaging before now19:37
SpamapShighvoltage: me too :p19:37
bdrungSarvatt: i thought that syncing a no-change version from debian has no benefit, but if a rebuild is really desired, it's ok19:38
bdrungSarvatt: if pitti wants the rebuilds, then give him the rebuilds by syncing from debian.19:40
=== Claudinux_ is now known as Claudinux
Sarvattbdrung: either way works, it's just easier for me to request syncs than find sponsors for all of the no change rebuilds so I did it that way for packages in xserver-xorg-video-all and I completely understand if a no change rebuild would be preferred, no problems doing it that way instead here outside of it being a lot harder to find sponsors :)19:41
bdrungSarvatt: but i prefer a sync over a no change rebuild :)19:42
bdrungSarvatt: we have ack-sync for sync requests and sponsor-patch for everything else. so there is no big difference from the sponsoring side of view19:45
bdrungSarvatt: processed all sync requests - all were fine19:51
Sarvattbdrung: thanks a ton for that!19:52
bdrungSarvatt: np19:52
Sarvattsorry to dump so many sync requests in there, unstable is being reserved for squeeze on the driver side and everything's being done in experimental19:54
bdrungSarvatt: no problem. we can live with that. in the lucid cycle, we had to requests syncs from unstable. ;)19:55
bdrungall sync requests processed19:58
SpamapShmm... is there a way to lock plymouth so it won't exit after 'plymouth quit' until you've unlocked it?20:01
SpamapSthat would work to make sure plymouth is around to enter the SSL pass phrase...20:02
SpamapShmm.. I actually need it to also not deactivate20:03
SpamapSI wonder if pause-pogress would work20:10
slangasekSpamapS: I don't see any sane way to wait in plymouth-stop for apache to start.  For now, perhaps open a new bug report against the package to document this know issue?20:53
slangasek+n20:53
bdrungBlackZ: libgphoto2-2.4.10.1 fails to build: Unmet build dependencies: libgpmg1-dev - but libgpmg1-dev is there21:11
BlackZbdrung: it's related to build-dependencies with [linux-any], if you want, patch your pbuilder ( see http://bugs.debian.org/363193 )21:12
BlackZbdrung: it's fixed in the launchpad buildd AFAIK21:12
bdrungBlackZ: the merge will took more time (this bug and big diff). ask the patch pilot instead or ask me tomorrow again.21:16
bdrungBlackZ: i sponsored enough today.21:17
BlackZbdrung: heh, will do :)21:17
bdrungBlackZ: look at that spice in the first graph: http://reports.qa.ubuntu.com/reports/sponsoring-stats/21:17
BlackZbdrung: uh, cool!21:19
micahgvery cool :)21:19
=== dendrobates is now known as dendro-afk
bdrungcrimsun: can you reassign bug #678183 to the correct package?21:28
ubottuLaunchpad bug 678183 in vlc (Ubuntu) "audio-only playback always stops after 1.5 minutes on any file types" [Undecided,Invalid] https://launchpad.net/bugs/67818321:28
=== dendro-afk is now known as dendrobates
Awsoonnhi all, what package is the network proxy config tool in?21:33
hallynupstart question.  Is it actually the case that if package X installs an /etc/init/X, that its removal should not remove /etc/init/X?  (I wouldn't have thought so)21:37
cjwatsonnot particularly upstart-specific21:38
sorenhallyn: It's got nothing to do with upstart. It's all dpkg.21:38
cjwatsonin general, files in /etc are conffiles and are only removed on purge21:38
hallynoh, ok.  so --purge should remove the conffile at least?21:38
cjwatsonyes21:38
sorenhallyn: this is also why most init scripts (sysv style ones, that is) almost always have a check for the presence of the daemon they're supposed to start.21:39
hallynok, thanks guys.21:39
sorenhallyn: Sure.21:39
hallynoh, so does dpkg do it automatically for debian/*.upstart files?21:40
sorenhallyn: Not that I know of.21:41
ebroderhallyn: debhelper does it automatically for files installed into /etc21:42
soren?21:42
sorenSorry, then I don't understand the questoin.21:42
cjwatsonindeed, at debian/compat level 3 and above21:42
sorenhallyn: What is "it" in your question?21:42
ebrodersoren: dpkg determines what should be removed at purge vs. uninstall based on a "conffiles" file in the control information (i.e. DEBIAN/conffiles or /var/lib/dpkg/info/foo.conffiles)21:43
sorenebroder: I know.21:43
hallynsoren: sorry, i meant removing the init scripts21:43
hallynbut dh_installinit manpage i *think* answers that21:43
sorenhallyn: That's what I thought. Then no. It doesn't.21:43
hallynhm, dh_installinit says it sets up postrm commands...21:44
ebroderhallyn, soren: Err, sorry - I think I inverted the question when I answered it21:44
sorenhallyn: It removes the rc?.d/ links.21:44
hallynsoren: yeah that was my first interpretation, but that doesn't help :)21:44
hallynebroder: meaning it does not?21:45
hallynall right - thanks guys, at any rate it's not quite what i expected, sadly, and i know what i need to do :)  thanks21:45
ebroderhallyn: debhelper will mark /etc/init/* as a conffile. dpkg will only remove conffiles when the package is purged21:45
sorenhallyn: debian/*.upstart are installed as upstart jobs. They land in /etc and are handled like any other conffile.21:45
hallynok, so still --purge removes them21:45
sorenhallyn: "conffile" is a very specific term here.21:45
sorenhallyn: Different from "config file".21:45
sorenhallyn: That may not be completely obvious.21:46
hallynit's not21:46
sorenhallyn: Right. So when people talk about this stuff, "conffiles" is a term used to describe a set of files that dpkg handles in a very specific way.21:46
sorenhallyn: For almost all packages that use debhelper (which is almost all of them), this set includes all files shipped in /etc.21:48
sorenhallyn: Others can be added to the set. It's reasonably rare, though.21:48
sorenhallyn: "shipped in /etc" is also a pretty specific phrasing. Some packages end up having files in /etc that aren't conffiles.21:49
sorenhallyn: They do this by shipping the file in /usr/share somehwere and then copying it into /etc at postinst (or preinst, I suppose) time.21:49
hallynsoren: ok, thanks.  and i presume the idea is that those should be preserved as much as possible?21:52
cody-somervilleugh21:56
cody-somervillewhy am I getting ubuntu-desktop team bugs now? :(21:56
cody-somervilleoh wait, thats the mailing list21:56
sorenhallyn: Sorry, i don't understand the question.21:59
=== ogra_ac__ is now known as ogra_ac
hallynsoren: well, between package upgrades, and temporary removals, etc, the conffiles are treated specially so that local customizations are preserved22:00
hallyn(i was just blabbing, ignore me)22:04
ScottKhallyn: It's described in debian-policy.22:14
GulfstreamWhat language is Ubuntu written in? I am thinking about simply changing the name... WHat language skills are needed?22:16
cjwatsonUbuntu is made up of lots of different pieces of software written in quite a few different languages22:17
cjwatsonthe primary ones, depending on how you count it, are probably C, Python, POSIX shell, C++, Perl, a few others22:17
* ogra_ac thinks shell comes before python :)22:18
hallynScottK: thanks, bookmarked.  will be reading that22:18
ogra_acif the order is based on most freqently used langs22:18
=== dendrobates is now known as dendro-afk
Gulfstreamwhich language is best for fixing Ubuntu?22:22
ogra_acyou need to really look at the piece of ubuntu you want to fix22:22
ogra_ac(i.e. the package that has the issue)22:22
ogra_acand find out which language this uses22:22
Gulfstreamokay22:23
GulfstreamWhich language would you recommend someone learn?22:25
ogra_acshell and C would likely be a good start ....22:27
ogra_acbut thats personal preference22:27
sladenGulfstream: pick a goal, and then work backwards to see how to achieve it22:27
ogra_acsome pepole would probably recomment to start with python22:28
ogra_acyeah, what sladen says makes sense22:28
sladenGulfstream: perhaps investigate some of the "papercuts" bugs.. some of those are as simple as changing a string (eg. menu item) to be more accurate22:31
GulfstreamOkay22:32
Gulfstreamis C# used much?22:33
cjwatsonthere's some, not lots in the core22:33
cjwatsonwe ship a couple of applications by default that are written in C#22:33
GulfstreamThanks for the info... I will see what I can learn22:35
hungerGulfstream: "Our" C# is using different libs than would normally be used on windows... so do not spend too much time on the ui stuff when wieding a windows book:-)22:36
SpamapSslangasek: The bigger issue is that when we write an apache upstart job, we are going to have to revisit the non-plymouth passphrase dialog... :-P23:05
slangasekoh, this isn't even an upstart job yet?23:06
ebroderSpamapS, slangasek: Isn't the bug that plymouth will willingly exit when a client is in the middle of an ask-{for-password,question} command?23:06
SpamapSactually I don't think it will23:07
SpamapSits that gdm may start before apache has asked for a password23:07
SpamapSebroder: I'm trying to grok plymouth's code/docs to verify that.. or to figure out how to play with it without having to constantly restart a vm23:08
ebroderSpamapS: At least with plymouth-x11, "plymouth quit" will cause plymouthd to exit, and the plymouth clients get an error23:08
ebroderSpamapS: Install plymouth-x11, then do sudo plymouthd and sudo plymouth show-plash23:08
ebroder*show-splash23:09
NCommander@pilot in23:09
=== udevbot changed the topic of #ubuntu-devel to: Archive: Open | Development of Ubuntu (not support, not app development) | #ubuntu for support and general discussion for dapper -> maverick | #ubuntu-app-devel for application development on Ubuntu | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | Current Friendly Patch Pilots: NCommander
gesercjwatson: (/me reading the meeting logs) I've no problems to access packagesets information over LP API with anon login. Or is a specific method failing?23:09
jonoNCommander, did you just sign in for your patch pilot work?23:09
=== apachelogger is now known as FanOfNightrose
SpamapSebroder: ahh, you're my hero. ;)23:09
NCommanderjono: yeah.23:09
jonoNCommander, cool, thanks!23:09
NCommanderjono: I can't upload though23:09
NCommanderjono: at least not today23:10
jonoNCommander, no worries, if you can review patches that is fine23:10
* NCommander does not have a working card reader so no GPG keys, I can put stuff in my queue however to upload when I can23:10
jonoand help people get in touch with the right folks23:10
jonocool23:10
jonothanks NCommander :-)23:10
ajmitchNCommander: you're the lucky person first up for it? :)23:10
ebroderajmitch: mvo went earlier23:10
jonoajmitch, mvo was23:10
jono:-)23:10
ajmitchaha23:10
=== Amaranth_ is now known as Amaranth
* NCommander generates a temporary GPG key for thsi laptop so I can upload23:13
NCommanderBAH :-P23:13
=== FanOfNightrose is now known as apachelogger
=== apachelogger is now known as phononlogger
cjwatsongeser: I tried making edit_acl.py use anonymous login for query and it fell over in some way I forget23:21
cjwatsongeser: if you can make it work, send me a patch :)23:21
cjwatson--- ChangeLog   2010-08-04 03:43:05 +000023:22
cjwatson+++ ChangeLog   2010-11-22 13:33:13 +000023:22
cjwatson@@ -1,3 +1,4041 @@23:22
udevbotError: "@" is not a valid command.23:22
cjwatsonyow, grub is not a particularly quiet project23:23
gesercjwatson: will give it a try, the FTBFS page is also using anon LP API login and can access packagesets data. I'll look what breaks.23:23
cjwatsonI may just have been being stupid23:23
SpamapSebroder: does plymouth-x11 allow typing in answers to ask-question btw?23:24
SpamapSebroder: mine seems not to23:24
ebroderSpamapS: I can get ask-for-password to work, but not ask-question23:44
SpamapSebroder: hmm.. so.. plymouth quit doesn't seem to kill it ... just blocks actually23:56

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