[00:06] barry: have you been introduced to wrap-and-sort? [00:07] barry: wdiff is your friend. [00:08] barry: you could file a wishlist bug for the maintainer to use wrap-and-sort from devscripts if it's bad [00:18] if user during installation selects encrypted home, then we don't use gnome-keyring for passwords? [00:19] thomx: No, we still use gnome-keyring for passwords. What makes you think otherwise? [00:20] RAOF: just installed cloudsn and it tells me that i use plain-text passwords by default rather than gnome-keyring. that made me think about it [00:21] is there a way to check that i'm using gnome-keyring "by default"? [00:21] or, set it somewhere? sorry for that i'm asking about it in -devel, but i guess no one in #ubuntu will know about these things [00:22] There isn't really a way to *not* use gnome-keyring. Anything which uses gnome-keyring will use it, and will fail if it's not running. [00:22] gnome-keyring-daemon is up and fine [00:42] barry: Feel free to package the python3 bits while you're up to your elbows in python-qt4. [00:55] Ack. Are there any archive admins around? [00:55] RAOF: Hm? [00:56] StevenK: Can you please do the magical dance for update-manager updates? [00:56] RAOF: The ... what? [00:56] The https://wiki.ubuntu.com/ArchiveAdministration#Special_case:_update-manager_updates [01:03] StevenK: ^^^ [01:03] RAOF: Yes, I saw, I'm reading, and distracted [01:04] Thanks. Just checking :) === asac_ is now known as asac [01:26] RAOF: For natty? [01:27] StevenK: Yes, please. [01:27] Hm, which doesn't exist. That's easy [01:27] Quite. [01:28] Natty doesn't exist? That'll make the SRU burden go down a lot! [01:29] RAOF: Bah. :-) [01:29] :P [01:29] RAOF: main/dist-upgrader-all doesn't exist in natty-updates [01:29] cp -a natty-proposed/dist-upgrader* natty-updates/ [01:29] Or I missed a main. [01:29] But yeah. [01:29] It *should* have existed, I suspect, we've been dropping that ball. [01:30] And mvo's been picking up the slack with static URLs to -proposed. [01:30] So is this a case of out of date documentation? [01:30] RAOF: No, the docs are correct, the admins are out-of-date. :P [01:31] RAOF: I noticed this slack a week or so ago, corrected a bunch of d-i oopses, and some dist-upgrader oopses, but didn't fix all of the latter because mvo's workaround let me slack. [01:41] RAOF: Done. [01:41] StevenK: Thanks muchly. [01:41] RAOF: It won't turn up on archive.ubuntu.com for ~ 80 minutes [01:42] This, like asteroids, do not concern me. [01:42] Or possibly does not. [01:42] Heh [01:42] * RAOF thinks the rest of the day should be devoted to routine work with low consequence for failure. [01:48] * mwhudson has a strange problem [01:49] mwhudson: Do expound! [01:49] perl -w /usr/share/debconf/frontend $POSTRM remove is hangin [01:50] it has the packages postrm script as a zombie subprocess [01:50] as far as i can easily tell, the postrm script exited normally [01:50] can i make logging super verbose about what's going on somehow? [01:56] mwhudson: There's a DEBCONF_DEBUG env var [01:57] StevenK: DEBCONF_DEBUG=1 apt-get remove $package ? [01:57] mwhudson: (I think that's it) -- if you set it to developer, you'll drown in debug logs [01:57] ah ok [01:58] hooray for virtual machine snapshots is all i can say [02:02] well that was uninformative === asac_ is now known as asac [02:05] ah, http://www.fifi.org/doc/debconf-doc/tutorial.html#AEN198 seems relevant [02:07] mwhudson: Indeed. debconf is a little anal about FD usage [02:10] could "sudo service apache2 restart >&2" do something with fds that debconf doesn't like? [02:14] the answer seems to be yes [02:14] in fact, i bet it's the >&2 that's the problem [02:17] huh, no apparently not [03:10] c [04:12] * SpamapS does the happy dance as 11.04 boots on his new Macbook air [04:22] SpamapS: Oooh. Correctly? Last I heard was that i915 made everything blow up. [04:22] http://www.backtrack-linux.org/forums/backtrack-5-beginners-section/41096-tips-dual-boot-macbook-air-3-2-a.html [04:23] These guys show the way [04:23] Specifically, have to make a new ISO with an updated isolinux [04:23] http://www.uni-koblenz.de/~ceinig/howto/artikeldruck.php?id=92 [04:23] RAOF: I'm still just in the installer.. it may die a horrible death after this. ;) [04:24] Ah, well. If the livecd's booted then it likely works better than Sarvatt's machine :) [04:24] Now I just have to resist the urge to reinstall my main machine too.. Natty really surprised me.. Unity has grown on me quite a bit. [04:25] ..and is dead for me on oneiric because of lack of multi-monitor.. :( [04:27] Multimonitor works pretty nicely in unity for me. As long as you don't try and change the setup, of course. [04:27] w/ nvidia twinview, it blacks out the entire second monitor [04:27] for no apparent reason. :-P [04:27] Ah. If you restart unity it'll probably work? [04:28] nope [04:28] It seems to be a deliberate "don't put stuff here" blackened area [04:28] no clue why [04:28] Unity is (still!) terrible at noticing when it should draw onto a differently-sized framebuffer. [04:37] hrm.. black screen after grub.. :( [04:39] Yeah.. something's dying after 'Loading initial ramdisk' :-P [04:48] i915.diescreaming=1 should ‘fix’ it, FSVO fix. [04:48] is that.. is that a joke? [04:50] why no.. it isn't [04:50] haha.. thank you :) [04:51] Passing any module parameter to a module that it doesn't understand will cause it to fail to load :) [04:51] So I like to be inventive :) [04:52] You've probably noticed the lack of 3d by now. [04:53] there's a reason my laptop with an i915 is still running hardy :) [04:53] and probably no multi-monitor capability either.. :-P [04:54] which is.. sort of the point of having this machine.. to hook up to projectors [04:55] I'm *pretty* sure efifb isn't going to do hotplug for you, no. [04:55] so the question is, where are we at w/ i915? Can I backport a kernel or something? [04:56] The intel kernel module is currently undergoing a period of ‘why doesn't this damned display work properly’ fixing. You could try drm-intel-next, but there are likely to be significant fixes soon. [04:56] Whether or not those fixes actually fix your Apple system is another matter, of course. Mac hardware is special. [04:57] If only somebody else made an equally sexy intel machine. :) [04:58] At this point.. I'm thinking VMware fusion is my better option. :-/ [05:00] well I suppose a bug report is in order [05:02] Yeah. [05:03] It'll be difficult to get any useful info out of there, though; from what Sarvatt described, I think you'll need to netconsole to get anything out. [05:03] On the plus side for you, a member of Ubuntu-X has one of the shiny new Airs… [05:04] I am a server guy. I detest messing with any of this stuff.. :-/ [05:04] Just give me terminals.. lots of terminals. [05:05] then clearly you should have bought a thinkpad :) [05:05] Hah.. I refuse to give in to ugly hardware! :) [05:05] i mean noone is going to take you seriously without a seriously ugly laptop [05:05] I have to appeal to the ruby kids [05:05] i don't think you can run textmate in a terminal, can you? [05:06] going for the rock star look? [05:06] ok bug 817324 filed [05:06] Launchpad bug 817324 in linux (Ubuntu) "New MacBook Air blank screen after grub" [Undecided,New] https://launchpad.net/bugs/817324 [05:12] I find it odd that the livecd works, though. [05:13] works, but not w/ 3D [05:13] or at least, it was in classic mode [05:15] RAOF: rebooting into the live cd now [05:15] good morning [05:15] Ra raw! Morning, didrocks. [05:16] RAOF: well i915 is actually loaded.. [05:17] But you don't have 3D? Hm. 11.04 should have sandybridge 3D. Well, ish. [05:17] xdpyinfo says GLX [05:17] does the live env just always choose classic? [05:17] Install mesa-utils and run glxinfo to really check that it's working. [05:21] Ok thats a meaningless burp of drivel to me.. but.. I do see "Software Rasterizer" [05:21] glxgears looked nice, and got 708fps [05:23] nah looks like fb still [05:23] no multi monitor either [05:24] hey RAOF ;) [05:27] Hey didrocks. [05:27] SpamapS: Want to pastbin your /var/log/Xorg.0.log and dmesg to satisfy my idle curiosity? ;) [05:27] RAOF: you are 0.5ms too late asking for that.. I just shut it down [05:27] and need to sleep.. as I have to be on a plane in 8 hours [05:28] Ok. Idle curiosity can be satisfied later. [05:28] Boo, hiss. Air travel! [05:28] We hates it! [05:28] RAOF: yes, I'll remind you when I am fighting the urge to return this to the store. :-/ [06:56] who should I ping to get a review for my usb-creator branch? [07:00] evfool: It _should_ be on the sponsorship queue, which should get it seen in not-too-long. [07:01] RAOF: I've proposed a merge more than a month ago, and it did not appear on the sponsors queue since then [07:01] Hm. It should do. What's your merge? [07:01] https://code.launchpad.net/~evfool/usb-creator/unityprogress/+merge/65844 [07:02] RAOF^ [07:02] Ta [07:02] RAOF: I've subscribed the Ubuntu Sponsors Team a few minutes ago to see if it does help to appear on the sponsors queue [07:03] Ah. Yeah. That merge is not targetted to an Ubuntu branch, so hasn't appeared on the sponsorship queue. [07:05] good morning [07:05] RAOF: I also have other merges not targeted to ubuntu branches, but they do show up on the sponsors queue [07:06] evfool: It's possible that those branches have ubuntu-developers notified about merge requests. That doesn't seem to be the case for usb-creator. [07:07] RAOF: that might be [07:08] ping ev [07:08] evfool: ev maintains usb-creator. He may have seen it, but is likely busy with other things. [07:10] thanks TheMuso, I'll ping him again, to see if I can get it reviewed and merged before alpha3 === smb` is now known as smb [07:29] evfool: I'll review it today [07:29] apols for missing that [07:30] ev: no probs, but I don't want the branch to be forgotten :) [07:32] sure :) [07:55] Hi, can anyone explain how mysql gets started automatically when I do apt-get install mysql? [07:55] I can't seem to find where this happens in the postinst script [08:00] oh, found it... it's handled through upstart via /etc/init/mysql [08:00] .conf === Quintasan_ is now known as Quintasan [09:58] Somebody around that can try a test-build on armel ? === tkamppeter_ is now known as tkamppeter === hunger_ is now known as hunger [10:22] there's no Restart menu option? [10:23] huh. [10:24] jml, no, log out and reboot from the login screen if needed ;-) [10:25] jml, it's only temporary until they add the option to the dialog [10:25] jml, i.e see https://wiki.ubuntu.com/DeviceMenuAndUserMenu [10:36] seb128: thanks. [10:38] jml, yw [10:40] jml, ok, I forgot to reply to menu edit question you asked the other day, you can if you don't use indicator-appmenu, check org.gnome.desktop.interface can-change-accels in dconf-editor and open the menu and hit the keybinding you want over the menu item to replace [10:41] seb128: did I *really* ask that question? [10:41] jml, you asked if you could make nautilus go back to "delete" [10:41] the move to bin action [10:41] seb128: ahh, right :) [10:43] seb128: thanks. that's probably too much messing about for me. [10:43] yw [11:16] hello all, does anyone know how to set up a default wallpaper using a preseed file? === MacSlow is now known as MacSlow|lunch [11:27] sponsorng page not updated anymore ? Last updated at: Wed, 27 Jul 2011 11:00:47 +0000 :) [11:45] dupondje, bdrung: there's a bug in the sponsoring page [11:45] I'll file a bug [11:46] hum [11:47] hang on [11:47] dholbach: it runs and doesn't crash (did you read my mail?) [11:47] bdrung, http://paste.ubuntu.com/653733/ [11:47] it does crash [11:47] at least on qa.u.c [11:48] let's see if it still runs locally [11:50] let me see one which MP it chokes [11:51] ok, it breaks on the first one [11:52] bdrung, http://paste.ubuntu.com/653737/ [11:52] I added some debug prints [11:52] and there's the output [11:52] dholbach: you can output stuff if -v is specified [11:53] yeah, first merge proposal [11:56] bdrung, I'm asking in #launchpad [11:56] barry: ping [11:57] dholbach: it runs locally without crashing [11:57] dholbach: current run: http://people.ubuntu.com/~bdrung/sponsoring/ [11:58] bdrung, that seems buggy [11:58] like bug #814099 is closed for 2 days [11:58] Launchpad bug 814099 in sqlite3 (Ubuntu) "Please merge sqlite3 3.7.7-2 (main) from Debian unstable (main)" [Wishlist,Fix released] https://launchpad.net/bugs/814099 [11:59] seb128: that's strange [11:59] bdrung, wgrant noted it might be a corrupt cache - let me check [11:59] the most recent "time in queue" is from the 21 on your page [11:59] bdrung, ^ [11:59] one week old [11:59] let me remove the cache and rerun it [12:02] bdrung, that's what I'm doing right now - it might well be that it was just the cache on qa.u.c that was corrupt (for whatever reason) [12:05] anyone seen doko today [12:05] zyga: I think he is at debconf [12:05] dupondje, bdrung: it works again [12:06] mvo: ah, too bad [12:06] * zyga wants someone with pip/virtualenv/setuptools knowledge [12:07] Hello everyone, just curious as to whether or not there are any plans to allow a 64-bit kernel on a 32-bit install for Oneiric Ocelot [12:07] dholbach: here too. http://people.ubuntu.com/~bdrung/sponsoring/ seems to be up-to-date now [12:07] given apt now appears to have multiarch support [12:08] bdrung, thanks [12:09] dholbach: sweet :D [12:15] * zyga dives into python-defaults [12:38] hi, there is a package for libgl1-mesa-glx in 32 bit, on oneiric ? [12:56] I have renamed a binary package in foomatic-db. will the renamed package get into the archives automatically or does it need manual action by an admin? [13:04] tkamppeter: new binary packages end up in the NEW queue and needs to be manually accepted by an archive admin. [13:04] bdrung, who are the current archive admins so that I can contact them? [13:05] tkamppeter: why do you need to contact them? [13:05] tkamppeter: https://launchpad.net/~ubuntu-archive [13:07] bdrung, how often do they go through the queue? I need foomatic-db-xml be accepted to stop gutenprint from FTBFSing and some weeks ago I have uploaded a completely new package, cloudprint and I have still no reaction on it. [13:08] tkamppeter: binary packages tends to get accepted quickly (a few days normally) [13:11] bdrung, and completely new packages? [13:13] tkamppeter: that varies. that can take one or two weeks. if other packages depends on it, poke someone from https://launchpad.net/~ubuntu-archive/+members#active [13:14] cjwatson, hi [13:15] tkamppeter: he is at debconf and has probably no time for AA duties [13:19] tkamppeter, why did you need to rename that binary? [13:24] seb128, it is to make sure that build processes of printer drivers get the Foomatic database as original XML (so that the driver's PPDs can get built) and not as compressed PPD archive. [13:25] seb128, one can request the XMLs by installing foomatic-db-xml via Build-Depends:. [13:25] tkamppeter, what was the issue before this change? [13:27] seb128, the printer driver package build-depends on foomatic-db. foomatic-db is once the XML package, but also a virtual package provided by foomatic-db-compressed-ppds. [13:28] On the last upload of gutenprint the buildds seems to have installed foomatic-db-compressed-ppds. [13:28] you can't be a real and a virtual package at the same time [13:29] seb128, then the renaming would fix this by making foomatic-db virtual only (provided by foomatic-db-compressed-ppds and foomatic-db-xml). [13:30] why do you need a foomatic-db? [13:30] can't you just have the 2 other ones provides it? [13:32] seb128, in addition, I will change all driver packages which need the Foomatic XMLs to build-depend on foomatic-db-xml. [13:32] tkamppeter, I still don't understand what was the issue that you are solving [13:32] seb128, the two other ones provide foomatic-db now. There is no real foomatic-db package any more. [13:34] seb128, perhaps you know that all the years Foomatic had XML files (in /usr/share/foomatic) which provided the data to generate PPD files for printers. This PPD generation process was very slow and the XMLs very disk space consuming. [13:35] seb128, so I came to the idea some years ago to pre-build and highly compress these PPD files and let CUPS take the PPDs from the compressed archives. [13:36] right [13:36] what I fail to see is how changing the name of the binary solve anything [13:36] seb128, this I turned reality in summer 2010 as a GSoC student coded the appropriate tool for me (the pyppd package). [13:37] seb128: afaiu there is a foomatic-db (with the xml files) and foomatic-db-compressed-ppds (provides foomatic-db) and tkamppeter has a package build-depending on foomatic-db but the buildd picked the wrong one [13:37] seb128, so I added a new binary package named foomatic-db-compressed-ppds to the set of binary packages of the foomatic-db-source package, holding the compressed PPD archive. [13:38] geser, well, if foomatic-db-compressed-ppds doesn't really provide foomatic-db it should not use a provide? [13:39] seb128, many driver packages (and also printing tools in Debian which Ubuntu does not use) simply need Foomatic's PPDs and depended on foomatic-db therefore. [13:40] dholbach: [13:40] tkamppeter, ok, I newed it, still seems an unoptimal way to deal with it [13:40] seb128, as the compressed archive worked as well for them, I simply made the archive providing foomatic-db and this satisfied these driver and tool packages. [13:40] dholbach: hi ! ;D [13:40] hi alex____ [13:40] seb128, thanks. [13:41] tkamppeter, you're welcome, thanks for the explanations [13:41] dholbach: I can see the new UPG at ppl.c.com/~dholbach ;) [13:41] seb128, they told me that time that a build-depend will always pull the physical package and it did so all the time, until breaking on the last gutenprint build. [13:41] alex____, nice - I hope you like it :) [13:42] seb128, perhaps the buildds had already installed the compressed archive before starting to look into the build-depends. [13:42] tkamppeter, it's likely yes [13:42] seb128, this new layout now assures that the right thing will be done. [13:42] ok [13:44] dholbach: yeah, btw I got a request from the owner of the Swe Loco. He wanted me to introduce new to ubuntu development. but I'm quite new to all this as well [13:45] dholbach: new members* [13:45] alex____, how can I help? [13:45] dholbach: I just wanted you to know [13:45] dholbach: it was because of the omgubuntu article [13:46] alex____, give me a second, I'll dig out a link that you could use for presentation material [13:47] try http://spreadubuntu.org/en/material/presentation/introduction-ubuntu-development and http://spreadubuntu.org/en/material/presentation/fixing-bugs-ubuntu [13:47] alex____, I'll go on holidays tomorrow, but if you drop me a quick email, I could put you in touch with other Swedish Ubuntu developers who might help answer further questions you might have [13:48] alex____, the presentations not only have slides, but also notes which should help with the preparation [13:49] dholbach: ah, norway :) have fun. No worries, I can manager, but I guess the presentation would take place in an IRC channel. [13:49] manage* [13:50] I guess you could upload screenshots of the slides (wherever it makes sense) [13:50] and probably try to get the other Swedes on IRC as well :) [13:50] drop me a mail and I'll introduce you [13:50] and thanks :) [13:50] I'm looking forward to it [13:51] to what? I don't know how much I can help actually, but okay [13:51] to some Swedish Ubuntu developers [13:51] ah [13:52] I personally think giving a presentation is a great idea, especially if it's an introduction in your mother tongue [13:52] as I said, it's not an IRL pres :P [13:53] still :) [13:54] hmm, I guess it's still a good idea, even on IRC... [13:54] yeah === dholbach_ is now known as dholbach [14:14] cjwatson, is there a way to run the lucid server installer CDROM such that it won't eject the CD at the end of the install? I'm on a remote box where I'd really like theCD to stay i the bay. [14:18] woot.. in flight wifi is amazing. :) [14:20] SpamapS: expensive? [14:20] $9.95 for the flight duration (2hrs) [14:20] which airline? [14:21] Alaska [14:21] latency is impressively good [14:21] it's 3G so it's usually somwhere between 40ms and 100ms [14:21] at least last I tried it on US airways [14:50] RAOF: howdy, care to help me with an intel video/unity issue? [14:52] RAOF: thinkpad x201, lspci at http://paste.ubuntu.com/653832/ [14:53] RAOF: I'm unable to get 1920x1080 to my external monitor [14:53] which issue? [14:53] RAOF: directhex: oneiric [14:53] hm. haven't upgraded yet. kinda dreading it [14:53] i want to report a bug in ubuntu [14:53] directhex: well, it was broken in natty too [14:54] kirkland, in different ways. a gushing edge kernel fixed the worst bug [15:00] seb128, thanks, gutenprint is building correctly now. [15:01] tkamppeter, you're welcome [15:01] kirkland, it's middle of the night for RAOF, maybe ask on #ubuntu-x [15:01] seb128: ah, okay, thanks [15:04] hello [15:04] are here developers [15:04] ? [15:04] i have a bug [15:04] report [15:24] is deb http://archive.canonical.com/ubuntu natty partner by any chance mirrored anywhere? [15:25] I'd managed to get mirror.anl.gov whitelisted in our stupid corporate firewall (the .gov helped) but unfortunately that does nothing for java [15:25] so I'm manually copying over the java security update .deb files [15:25] (which are blocked otherwise by the stupid stupid websense) [15:26] well. https or ftp would work too - but I already tried. canonical doesn't support those apparently [15:28] cjwatson, or anyone... how/when do new seeds populate to tasks ? [15:28] my request for 'uec' task renamed to 'cloud-image' was merged, but i dont know how/when it would show up [15:29] h.... never mind. seems like its good now. [15:33] tgardner: cdrom-detect/eject=false as a boot parameter [15:33] * Riddell nudges dholbach or someone to review https://code.launchpad.net/~jr/ubuntu-packaging-guide/libraries/+merge/69293 [15:33] smoser: I took care of that straight after merging your change - it's manual [15:33] partly, anyway [15:33] Riddell, I generally liked it, just wanted somebody else to double check the symbols bits [15:33] cjwatson, thats what apw suggested, which worked. [15:33] I'm happy to add that as a comment [15:33] dholbach: yeah, me too :) [15:34] added :) [15:34] * dholbach hugs Riddell [15:34] Riddell, I mentioned your good work in the weekly development update post I just did :) [15:53] hey, how does natty perform on tablet machines? [15:57] hm, a tool (lxc-ps) wants to run 'mount -t cgroup', but needs to see /proc/self/mounts contents, not /etc/fstab. Is there a precedent for handling this? [15:57] I suppose I could createa fresh mounts namespace and bind-mount /proc/self/mounts :) but that seems heavyweight [15:57] scripting the parsing of /proc/self/mounts myself gets just hairy enough that i don't want to do it in a one-liner (in a larger perl script) === deryck is now known as deryck[lunch] [15:58] and 'mount' doesn't seem to have a command line option to specify what file to use as mtab [15:59] I'm excited about the Qt/qml, and I see Ubuntu becoming the member of the family [16:00] experience from touch devices, bottlenecks? [16:02] guess i'll go with grep -E '^[^ \t]+[ \t]+[^ \t]+[ \t]+cgroup' /proc/self/mounts [16:02] feels hacky, but should work [16:12] dholbach: can we get those posts on planet somehow? [16:13] Laney, they're on the fridge, so they should turn up on planet too [16:13] ah, groovy [16:14] not sure why they're not on their yet [16:18] why don't we see these new people in -motu? :-) [16:28] If I want to add a package (lxc) to my PPU set, should I create a whole new application wiki page, or just update my original one? [16:30] hallyn: I think it makes more sense for you to apply for ubuntu-server-dev :) [16:30] (lxc is part of that fwiw) [16:32] Daviey: curse you [16:32] Daviey: lxc bugs don't show up in the server set bugs, though. [16:32] so i'm not convinced it's int eh set [16:32] it should be, of course [16:33] hallyn: it is part of the ubuntu-server set, we are just not subscibred to the package bugs (fixing now) [16:33] https://bugs.launchpad.net/ubuntu/+source/apt-file/+bug/817622 => could somebody take a look ?:) [16:33] Ubuntu bug 817622 in apt-file (Ubuntu) "apt-file got broken in oneiric" [Undecided,New] [16:34] hallyn: fixed. [16:35] Daviey: cool, thanks. I'd been counting on it being in there, so there were some bugs sitting around untouched for too long :( [16:35] Daviey: hey, DEP5 q [16:36] do I mark a bug as 'Forwarded: yes' if i have intent on sending it upstream in the next week or two? :) [16:36] dupondje: that code change doesn't make sense [16:37] Daviey: I'm waiting to commit my source until you answer that :) [16:37] micahg: but it fixed my apt-file ? :) [16:37] hallyn: ubuntu-server package set in oneiric gives upload access to: http://pb.daviey.com/zTwD/raw/ [16:37] dupondje: let's move to -motu [16:37] hallyn: Forwarded: yes is implied if omited [16:38] Daviey: yes, but I feel like I"m lying if I say 'yes' but haven't yet done it [16:38] Daviey: if i plan to do it in next few weeks, is 'yes' ok? [16:38] hallyn: are you likely to upload again before release? [16:38] haha, at this point probably [16:38] almost definately [16:38] hallyn: then I would be inclined to mark it no :) [16:39] ok [16:39] Daviey: oh noes, i don't see open-vm-tools in that list :) [16:39] "shame" [16:50] hmm...I've got a python app that I've added to the unity launcher...but when it's running, unity doesn't see it as running [16:50] anyone know how I can debug this? [17:17] lucas: How could I query all_bugs in the udd and not only show 1 report if it has been merged with others? [17:17] er s/not// === _LibertyZero is now known as LibertyZero [17:24] with the autosync snafu of the last day, we pulled in some stuff requiring libjpeg-dev, should we update libjpeg62 to provide that (unless we're going to do the libjpeg8 transition)? [17:36] I'm trying to build a Natty version of apport for an SRU and I'm running into a failure I could use some help with [17:36] http://pastebin.ubuntu.com/653912/ [17:37] In the top 21 lines that is the failure building with sbuild on Oneiric [17:37] in the bottom half if I build in a natty chroot it works out fine [17:38] and of course I've palyed with icons in my changes [17:38] wow I haven't === cking is now known as cking-afk === deryck[lunch] is now known as deryck [17:53] bdmurray: on which filesystem (ext3 or ext4 or tmpfs) did you try to build? I've seen this error in the past already for apport but only in my tmpfs pbuilder (or some archive rebuilds) but the build on the buildd succeeded [17:54] geser: I tried it in my PPA too and it failed .. ext3 [17:54] hmm [17:55] micahg: do you know how big the transition is? [17:56] bdmurray: there was some discussion about this error in the past but I don't know if it was found why it happens (and how to fix it). [17:57] * micahg checks [18:00] hmm, my script seems to be broke... [18:00] laney@ries> zcat source/Sources.gz | grep-dctrl --eregex -FBuild-Depends -sPackage "(libjpeg-dev|libjpeg62-dev)" | wc -l [18:00] 309 [18:01] maybe I should do that on an Ubuntu box [18:01] * micahg gets 37 [18:01] 373 [18:02] yeah, probably not then :P [18:02] mdz: tb meeting? [18:03] sabdfl: tb meeting? [18:05] kees: mdz indicates (via #debian-devel@OFTC) that he's unable to connect to freenode at present [18:06] slangasek: ah! thanks [18:07] and no idea if cjwatson is available, it's 8pm @ DebConf [18:08] slangasek: e-mail to technical-board@ said not [18:09] slangasek, micahg: right, I've marked cjwatson and pitti as not-available [18:09] I can't find keybuk, and mdz's network is hosed [18:10] while you have some spare time, please respond to the DMB voting query ;-) [18:10] keybuk is presumably not paying attention to his schedule because he's at the upstart sprint [18:10] kees, I'm sort of here now [18:10] I am riding the train, though, so conditions are not ideal for a meeting [18:12] in perhaps 10 minutes I'll be in a more stable situation and can join the meeting [18:12] kees, ack, and thanks :-) [18:12] mdz: okay [18:22] slangasek, it seems that the natty powerpc kernels may be in universe [18:23] slangasek, http://ports.ubuntu.com/pool/universe/l/linux/linux-image-2.6.38-11-powerpc64-smp_2.6.38-11.47_powerpc.deb [18:23] i wonder if you have time to check for me === tlp_ is now known as tlp [18:34] apw: yes, looks like that needs adjusted for natty-proposed; that's the only place this kernel is, right? (no -security or anything that I should be hunting down?) [18:36] mdeslaur: I'll try to verify 313812 today. (gonna take awhile to d/l an iso to start :) [18:36] barry, hi [18:37] seb128: hi [18:37] barry, could you have a look to https://code.launchpad.net/~jbicha/ubuntu/oneiric/gedit-plugins/gedit-plugins-3.1.2/+merge/69386 [18:37] ? [18:37] * barry looks [18:37] barry, not sure why he needs to clean the .pyc manually, shouldn't dh_python2 do that? [18:39] yeah i think so. i never do that explicitly [18:39] hallyn: cool, hggdh said he's look also, might want to coordinate with him [18:39] s/he's/he'd/ [18:39] mdeslaur: ok, thanks for the headsup [18:39] barry: if you get some time, could you test that because when I only called it once, it didn't work right [18:39] Can anyone tell me what's the policy for rebuilding a package in natty? [18:40] jbicha: i'll try, but i'm a bit backed up right now [18:40] dh_python2 may not be designed for use outside the normal install location [18:40] jbicha: you can give it a directory though [18:42] jbicha: heh, i see you do though [18:42] that might be a bug in dh_python2 [18:43] and I have no idea why calling it a second time helps [18:43] slangasek: yep just in -proposed right now .. thanks [18:43] Quintasan: SRU process, like anything else [18:43] micahg: Thanks [18:46] jbicha: afaict, that's just what you (currently) have to do if you give it a directory. we've been discussing improving that, but i found the same thing [18:48] jbicha, the second call is ok, but the find call to clean the .pyc is weird [18:50] seb128: that was there before I got to it, do you have a better one-liner for that? [18:50] jbicha, well I though that dh_python2 ought to clean that for you [18:50] barry, ^ should it be the case? [18:51] it's a .la removal line [18:51] jbicha, your merge request has "+ find . -name '*.pyc' -delete" [18:51] ok, let me check into that [18:52] thanks [18:53] ScottK: btw, i've just now verified that jose's patch to python-qt4 looks good. i built the package with his patch, and then was able to install python-qt4-dbus and python-dbus where before i couldn't. is there some other specific test i can perform to make sure it's working, or is build+installability enough? [18:56] seb128: ok, I removed that extra line and repushed [18:57] jbicha, thanks [18:59] barry: hi [18:59] barry: do you have a moment? [19:00] zyga: sure [19:00] barry: I ran across an interesting issue with: natty's virtualenv and dist-packages [19:01] barry: should one be able to use virtualenv, install a package inside with --install-layout=deb and then expect to use it normally? [19:01] barry: what I observed is this: 1) you can import a package normally [19:01] barry: 2) setuptools-based scripts will not see it (despite the .egg-info meta-data) and will proceed to download it again, for dependency resolution [19:02] barry: what do you think? [19:06] * zyga just reproduced the issue to be sure [19:07] * zyga looks at virtualenv directory to see how it looks like [19:07] barry: interestingly, I could install but I cannot develop [19:07] I thought this is identical but apparently not! [19:07] zyga: i just tested it here, and i can see that by default /lib/python2.7/dist-packages isn't on any sys.path [19:08] hmm [19:08] it is for me [19:08] o_O [19:08] * zyga tries to do it again [19:08] oh [19:08] I always use --no-site-packages [19:08] perhaps that affects this [19:08] site-packages is, and /usr/lib/python2.7/dist-packages is and /usr/local/lib/python2.7/dist-packages is [19:08] yep [19:09] w8 [19:09] nope, i still don't get dist-packages [19:10] barry: can you mkdir the relevant directory and try again [19:10] barry: meanwhile I'm checking my local setup [19:11] barry, can you do a python upload? [19:11] smoser: i saw doko's reply. it's on my list :) [19:11] k. thanks. [19:12] zyga: yes, if you create the dist-packages dir, it ends up on sys.path [19:12] barry: ok, as soon as you create dist-packages [19:12] barry: right [19:12] barry: which is interesting and odd [19:12] barry: I'm mainly interested in your expert opinion on how it _should_ work [19:12] zyga: it's odd that our favor does not create dist-packages [19:13] barry: I'm inclined to use dist-packages and --install-layout=deb because that does not create the .pth file I really hate, I'm trying to create something like pbuilder for python and the fact that all the packages mutate .pth file is problematic [19:13] barry: I'm also using python2.6 and python2.7, perhaps this is unique to 2.6 (never go to test 2.7) [19:14] jjohansen: around? [19:14] hallyn: yep [19:14] zyga: my thinking is that it should work analogous to upstream behavior with site-packages. iow, we add dist-packages and if that's doesn't work the way it would for someone use the source tarball, then it's a bug [19:14] jjohansen: on https://blueprints.launchpad.net/ubuntu/+spec/server-o-lxc-sandboxing, mind if I mark your item as postponed? [19:14] zyga: i think it works the same in 2.7 which is what i'm using [19:14] barry: what is _really_ strange for me is that setuptools does not see this correctly [19:15] barry: I have a few packages, when I try to install them with deb layout they all install correctly [19:15] barry: but when I try to develop (say the last one) instead of installing it tries to fetch all the existing installed packages from pypi [19:15] hallyn: please do [19:16] zyga: wait, when you say "setuptools" what do you mean, exactly? :) remember in debian/ubuntu, setuptools is really distribute, and i think our version is hacked to understand dist-packages. so if it's upstream's actual setuptools, then it doesn't surprise me too much [19:16] barry: ah, distribute [19:16] jjohansen: done [19:16] barry: though shortcut [19:16] hallyn: thanks [19:16] barry: stuff that says: from setuptools import setup, unlike the stuff that says from distutils import setup [19:16] yes, it's all very brain hurty [19:16] jjohansen: mind you, the patch they'r elooking at now could still be useful for very simple generic sandbox, but it's quite limited :) [19:16] np, ttyl [19:16] barry: I tried digging into setuptools code but I'm not finished yet [19:17] hallyn: yeah, but we can look at it for next cycle [19:17] zyga: you'll want to tie a rope off at the opening to the pit so you can crawl out afterward :) [19:17] barry: what I find confusing is that with the .egg-info directories and `pip freeze` show what was installed to dist-packages just fine [19:17] zyga: so, try using distribute from ubuntu and see if that helps [19:17] barry: haha [19:17] barry: what do you mean? [19:18] barry: you mean unlike the copy in virtualenv? [19:18] s/setuptools/distribute/ in your import statement [19:18] barry: but setuptools _is_ distribute [19:18] barry: in the virtualenv package in natty we patch it to use distribute all the time [19:18] barry: so it's just the name, it's distribute underneath [19:19] right of course. i'm thinking of something else, nm [19:20] barry: so your suggestion was to change my existing virtualenv to use the host's distribute, correct? [19:20] zyga: wait, sorry, let's step back a sec [19:20] ok [19:21] zyga: so, you create a virtualenv with --no-site-packages, then you install some dependencies into the virtualenv, then you try to install the package you're developing and it ignores the dependencies, right? [19:22] barry: mostly, I install them with --install-layout=deb, then I can install all the packages I want just fine, *then* any attempt to devlop a package involves re-fetching dependencies [19:23] zyga: and does it do the same thing if you don't use --install-layout=deb? [19:24] barry: no [19:24] barry: then it all works correctly [19:24] barry: installing to site-packages [19:24] barry: and creating/modifying the .pth file I hate [19:25] barry: peraphs I'm pushing the wrong door, I really wanted to avoid mutable file inside the virtualenv, I discovered dist-packages as a possible solution by mere accident [19:25] barry: if there is a better way of doing that then I'm all ears [19:25] barry: essentially dpkg-like guaranees (I shall not modify files shipped by another package) for random stuff on pypi [19:27] barry: +bonus point if the same scheme works on windows/mac [19:27] barry: so that the pbuilder-like tool works on those systems without magic ubuntu patches [19:29] zyga: so, a couple of things. first, i searched around on bts and launchpad and didn't find anything relevant. second, we are a little behind upstream, but offhand don't see anything relevant, unless we're talking about fixes in newer versions of the bundled pip or distribute [19:30] zyga: so, i think you're encountering a bug in our version of virtualenv. it would be good to file a bug on that [19:30] barry: I reviewed them too but I did not find anything that seemed relevant [19:30] barry: I'll be happy to [19:30] zyga: cool, thanks [19:30] zyga: next... [19:31] zyga: i'm not sure how much a bug fix would help you anyway, since windows and mac definitely will not have the dist-packages hacks [19:32] hh [19:32] hmm [19:32] zyga: for a cross-platform solution, i think you will probably want to use upstream's virtualenv [19:32] barry: how can I get rid of .pth? [19:32] barry: btw: is there any spec, apart from the source, for what this file is designed to do and how? [19:33] zyga: heh [19:33] zyga: http://docs.python.org/library/site.html [19:33] actualyl two setuptools and easy_install [19:33] * zyga reads [19:34] barry: that does not describe the non-declarative syntax of easy_install.pth [19:35] that's probably defined or described in one of the links from http://pypi.python.org/pypi/setuptools [19:35] http://peak.telecommunity.com/DevCenter/EasyInstall [19:36] briefly mentions it but doesn't go into detail [19:36] barry: ah, I need to re-read them [19:36] barry: last question [19:36] barry: do you know if simply yanking that file post-install will affect stuff in unexpected ways? From what I understand it's not really doing anything if the package is in site-packages already [19:36] zyga: i wish i could be more helpful. this is all a horrible mess [19:36] barry: I agree [19:37] barry: I wish I could help to untangle the mess but I dare to say the damage's been done and now it's hard to stay compatible and fix this [19:38] zyga: i'm not sure exactly what's in easy_install.pth, but be careful because sometimes .pth files use the hack that lines starting with import are exec'd to basically run arbitrary python. buildout does this for example [19:38] zyga: right, and really this needs to be sorted out in distutils-sig [19:39] barry: yeah, I assume it's also needed when you use multiple versions of the same library [19:39] * zyga looks at that ML [19:39] yep [19:39] barry: as in that case AFAIR they don't show up in stadard path by default [19:40] zyga: right, that's pkg_resources.require() functionality [19:40] right [19:40] ok [19:40] thanks for your time [19:40] * zyga goes to file the bug on virtualenv [19:41] zyga: you might want to file that on bugs.debian.org [19:41] zyga: i wish i had better news ;) [19:41] barry: even though I never tested how this works on debian? [19:42] zyga: i'm betting it will work the same way there since we try not to delta too much. but if you feel more comfortable, file it on lp against ubuntu and we can link it to debian bts later [19:42] barry: k [19:44] barry: may I quote you on the bug? [19:44] zyga: sure, np [19:51] bug 817702 [19:51] Launchpad bug 817702 in python-virtualenv (Ubuntu) "virtualenv does not create dist-packages directory" [Undecided,New] https://launchpad.net/bugs/817702 [19:54] zyga: thanks, subscribed === zyga is now known as zyga-afk [20:07] Hey folks! I used to be able to preseed auto-login with d-i passwd/auto-login boolean true, but this seems to have changed in Oneiric, anyone know where I could find the new setting to use? === Quintasan_ is now known as Quintasan === yofel_ is now known as yofel === zyga-afk is now known as zyga [21:09] barry: .pth files are insane, the rationale for their existance is totally bogus! [21:10] ;)