=== emgent`out is now known as emgent [00:04] hello === mthaddon changed the topic of #ubuntu-devel to: Archive: Feature Freeze, DocumentationStringFreeze | Ubuntu 8.04 LTS Beta released | Development of Ubuntu (not support, not application development on Ubuntu) | #ubuntu for support and general discussion for dapper/edgy/feisty/gutsy, #ubuntu+1 for hardy | #ubuntu-motu for getting involved in development | http://wiki.ubuntu.com/UbuntuDevelopme [01:44] mthaddon: You looked to have chopped off the topic [01:45] StevenK, oops - had removed the launchpad outage message I'd added [01:46] mthaddon: Yup, I think the outage message caused the topic to be cut off in the first place. === mthaddon changed the topic of #ubuntu-devel to: Archive: Feature Freeze, DocumentationStringFreeze | Ubuntu 8.04 LTS Beta released | Development of Ubuntu (not support, not application development on Ubuntu) | #ubuntu for support and general discussion for dapper/edgy/feisty/gutsy, #ubuntu+1 for hardy | #ubuntu-motu for getting involved in development | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs [01:46] Yay [01:46] :) [01:46] heya gang :) [01:46] xchatlogs are a wonderful thing... [01:47] irssilogs too :P [02:14] I'm assuming that when one tries to upload the first package to a recently activated PPA and it FTBFS due to index files in the PPA not being found is a known bug? [02:15] Yup. [02:17] TheMuso: Yeah, and somebody decided it wasn't important enough, so they untargetted it... [02:17] TheMuso: not deemed important though. just retry the build. [02:17] TheMuso: Retry the build in 5 minutes. [02:18] Fujitsu: How? Can I do that from the interface now? [02:18] Yeah. On the FTBFS page there's a "retry build" button. [02:18] Thanks./ [02:56] Hey all, I'm looking to do some open office development, has anyone setup an OO dev environment on Ubuntu here? [02:58] <_MMA_> Hydrant: calc is who you wanna talk to. [02:58] calc: do you have experience with OO dev environment on Ubuntu ? [03:11] why KDE is so slow and consumes too much ram? [03:11] gnome runs fine [03:12] calugor: what's your definition of slow ? [03:12] slangasek: Did you get the word from Riddell that it's OK to nuke kdepim-kde4? [03:12] apps take a few seconds to start [03:13] calugor: turn off some eye-candy [03:13] calugor: what are specs on the machine ? [03:13] :/ [03:13] <_MMA_> Why GNOME is so slow and consumes too much ram? Fluxbox runs fine. :P [03:13] calugor: there might be some services running that aren't needed [03:14] p4 512mb ram [03:14] _MMA_, fluxbox with gnome apps :D [03:15] calugor: so you're running a pure KDE environment, are you running the same apps in both GNOME / KDE ? [03:17] used kde apps while running KDE but perhaps there was someother stuff running in the backgroud [03:18] calugor: KDE can be very fast if configured well in my experience [03:18] calugor: it may be other things are going on that are hurting performance too [03:19] http://www.psychocats.net/ubuntu/purekde [03:19] I will try this [03:22] calugor: I don't know that will do anything for you [03:22] calugor: is the performance really so bad? [03:23] after it started using swap , yes [03:23] calugor: ah using swap eh [03:23] calugor: with 512MB RAM it shouldn't [03:24] calugor: are you running something fancy or just plain old KDE ? [03:24] calugor: sorry to ask, what is your user level experience? I don't want to water down answers or go too advanced either, if you give me an idea I can help more [03:25] average [03:25] calugor: did you check what else might be using memory? [03:25] calugor: I guess you can use ksysguard to check that [03:25] or KDE crash guard, whatever it's called [03:27] xorg was 300mb [03:27] calugor: something's up [03:27] calugor: what video card? Have you done something fancy since install? [03:27] no [03:27] calugor: oh, btw, AFAIK video memory is in that stat, I forgot [03:28] calugor: you must have a 256MB video card or something ? [03:28] yes, I will try again [03:28] calugor: no, if I'm right it's not a problem [03:28] calugor: ok... so it's going to swap... how big is your swap drive? [03:29] 1.5gb [03:29] calugor: how do you know it's going into swap, and how much swap is being used? [03:30] I need to test again and pay attention to the ksysguard or whatever it is called [03:30] thanks [03:52] Hydrant: back [03:52] Hydrant: what was your question? [03:56] calc: hey [03:57] calc: I'm trying to do some OO development on Ubuntu, and was curious if anyone had that setup [03:57] calc: I mean, if I wanted to build OO with debugging support it looks like I have to compile myself === ember_ is now known as ember [04:03] Hydrant: debug builds for all packages are already available [04:04] * calc looks up url [04:04] calc: thx [04:04] * calc thinks the url should be in the default sources.list [04:04] https://wiki.ubuntu.com/DebuggingProgramCrash [04:04] what's the package name ? [04:05] depends on which package [04:05] i think they append -dbg to end of each iirc [04:05] oh maybe -dbgsym [04:05] read the page :) [04:05] k, I'm used to debian [04:06] * calc thinks debian should switch to doing it this way :) [04:06] much easier you don't have to care about it in your packaging afaict [04:06] tbh i don't even know how it does it [04:08] nice, installing now [04:08] do you know by any chance why unoidl isn't in Ubuntu [04:09] * lamont decides to reinstall the gutsy version of kobodeluxe... unless I can figure out how to get the gutsy graphics in the hardy version [04:10] Hydrant: no, hmm [04:10] Hydrant: i wouldn't be surprised if most uno stuff doesn't work in general though, since we have no stable free java yet [04:10] calc: so Ubuntu might not be a good dev platform ? [04:11] * calc keeps his opinion of java to himself [04:11] haha... my opinion of Java is likely much worse than yours [04:11] Hydrant: feel free to file bug about missing unoidl, not sure why its not there [04:12] calc: hrrm... if only compiling OO didn't take all night [04:12] i don't mind the language itself that much just the surrounding issues [04:12] Hydrant: hehe [04:12] calc: JVM is the devil [04:12] i'm booting up my other ubuntu vm to look at upstream ooo [04:12] calc: k, thx... [04:13] calc: I've gotta get a project done tonight, kinda in trouble if I can't get OO working right [04:13] i thought i was going to get 2.4 uploaded today, heh yea right [04:13] Hydrant: what do you need to get working on OO? [04:13] calc: if it's in heron I might be able to put it in [04:13] calc: I've been working on adding some spread sheet functions to OO Calc... spent a week or so trying in PyUno, found a major bug with it today that blocks that [04:13] i got to my first bug which i thought was an easy one, fix up mime types, then found out shared-mime-info doesn't even support most Office mime-types :\ [04:14] Hydrant: yuck [04:14] calc: today I've decided to just go back to C++'s loving embrace, but now I'm missing unoidl :-( [04:14] Hydrant: if you want to do it with Ubuntu and it looks like you are having weird bugs your best bet would probably be to use the official sun version, but you will lose gnome integration [04:15] calc: don't care about that [04:15] i don't see unoidl even on upstream version [04:15] where is it normally located? [04:15] calc: this is my dev environment, actual target server to deploy to is running Arch [04:15] calc: no idea, I'll check on another server [04:15] its not in /opt/openoffice.org2.4/program/uno* [04:16] try slocate [04:16] i don't see it anywhere in /opt/openoffice.org2.4 [04:16] slocate unoidl returns nothing [04:16] i think your mythical unoidl is just non-existant (or at least in 2.4) [04:17] http://udk.openoffice.org/common/man/tools.html#unoidl [04:17] methinks the OO docs are in shambles [04:17] calc: so are you the OO maintainer? [04:17] maybe its an internal program that they don't ship [04:17] Hydrant: yea [04:17] because in the official sun debs its not there either [04:17] calc: You have my deepest sympathies :-) [04:17] they have lots of internal tools that i think they don't generally package up [04:17] Hydrant: heh yea :) [04:18] calc: but it ought to be in the SDK [04:18] calc: arrgh [04:19] hi calc :) [04:19] i don't see a sdk deb so it must be part of what they call core or testtool(?) [04:19] emgent: hi [04:19] the debs i am talking about are the sun debs by the way [04:19] calc: where can I get official sun debs? [04:19] eg OOo_2.4.0rc5_20080312_LinuxIntel_install_en-US_deb.tar.gz [04:20] calc: how broken ? [04:20] they changed their website yesterday so i am not sure how to get them now, looking [04:20] http://download.openoffice.org/other.html#en-US [04:20] there it is [04:21] http://openoffice.bouncer.osuosl.org/?product=OpenOffice.org&os=linuxinteldeb&lang=en-US&version=2.4.0 [04:21] its only available for 32bit and doesn't upgrade cleanly so if you go to upgrade you might have to do juggling later [04:21] eg placing packages on hold, etc [04:22] well, I'd remove existing OO altogether [04:22] I've likely broken it anyways [04:22] and you won't have office 2007 support, etc [04:22] since upstream doesn't support it at all yet [04:22] meh, who cares... [04:22] just noting things that upstream lacks ;-) [04:22] thx [04:22] official OOo gets Office 2007 support in 3.0 [04:23] speaking of lacking, I can't get the d/l to work [04:23] lol [04:23] i didn't try downloading it myself [04:23] it was just released today so that might be why [04:23] the gods hate me [04:25] heh that was what i thought when i found all the missing support in shared-mime-info ;-) [04:27] I'm getting the tar, hoping it's not source [04:27] OO hurts my brain [04:27] I spent all day tracking down a bug from 2004, that was "going to be fixed" in 2.0 [04:27] not only that... but the buggy source that never worked is on the wiki as an example of how to get this working [04:27] ARRGH! [04:29] k, I got a deb from a mirror [04:29] my fiancee has been doing a lot of labels, cards, etc. etc. in OOo, and every time I have to redo the templates because they're off by part of an inch. Extremely irritating [04:30] plus OOo seems to have this fixation with A4, and pays no mind that my printer is set to print letter [04:30] my personal favorite was the fact that fonts aren't stored in the OO file... so when we did a server upgrade the fonts changed in some inperceptible manner and broke all docs [04:30] but at least it does the job... [04:31] they were all aligned forms and had to be re-done [04:31] makes me want to implement label printing support in inkscape though. :-) [04:31] but OO is working alright [04:31] yikes [04:31] overall it's alright, it's free, it's open source [04:31] ... an obfuscator would be less confusing [04:31] ... but open source [04:31] there's an extraordinarily annoying bug where it doesn't have a template [04:32] so if you ever pick a business card template, you'll never be able to make blank business cards. You must always have a template [04:32] I've debugged more than my share of OO bugs [04:32] Oh, I have a better favorite [04:32] bryce: file a bug http://www.openoffice.org/issues/query.cgi :-) [04:33] I found if I used the "Attention:" template, it was easy enough to just backspace out the "Attention:" text from each card, and load from that [04:33] bryce: they are generally pretty responsive as long as you describe your problem enough for them to understand you [04:33] calc, really? Hmm [04:33] ... debugging an OO BASIC script... you know how OO BASIC doesn't support threading, says so in the docs... I spent a week debugging a problem that wound up only being possible if it was a threading issue.... I wound up having to write an OS-like event dispatcher with interrupts and everything to make sure the stupid thing wasn't threading... ah well [04:34] * calc is going to check on wordperfect then file mime bugs on it and be done with mime crap for now [04:35] mime issues are a royal pain [04:35] Hydrant: yea have to fix them in both OOo desktop/mime files and in shared-mime-info [04:36] Hydrant: i just filed four high priority bugs to fix them for Office 2007 file types [04:36] i installed xp/office 2007 and saved out every format it supported and tested them and lots weren't supported [04:36] er weren't supported by shared-mime-info [04:37] calc: hrrm [04:37] calc: I'm sure that kinda testing is appreciated [04:37] Hydrant: yea i should have done that testing a few months ago [04:37] I've been impressed with how little thinking I have to do with Ubuntu [04:37] now its going to be hard to get it supported before release :-\ [04:38] of course it probably doesn't work on any other linux either, since its an upstream bug [04:38] I've run Gentoo and Debian for years... for my personal desktop now I prefer Kubuntu only for not having to do so much work just to manage things [04:38] calc: not to say it works in others, more so that you have to build it yourself as you go [04:38] calc: I'm more used to a blank slate and building a lot of things from scratch [04:39] * calc used debian from 1998-2004, slackware/redhat/etc before that [04:40] calc: ooh, hardcore [04:40] calc: I've only been using Linux since 2000 [04:40] thats a pretty long time pre gnome 2.0 anyway ;-) [04:40] calc: mostly as a user [04:41] calc: yup... thats tldp days [04:41] i started using linux in high school, but that was a long time ago (1995) [04:41] calc: I've learned that most Linux "gurus" who have used it for 3 weeks don't know of tldp, so it's a good litmus test nowadays to weed out people [04:43] calc: not that that's important, I just get irritated when I get lectures on Linux from people who use it for 3 days... but now I'm just ranting on [04:43] calc, huh, there's already a ton of bug reports about templates [04:43] tldp? if i know what you are talking about i thought it was just LDP [04:43] calc: true [04:44] calc: I wonder whatever happened to it, back in the day it was the most critical thing ever [04:44] bryce: yea templates were left out but it was other things besides that [04:44] aha, here's the one where you can't do a template - 42873 [04:44] bryce: works file format, xml document format needs subclassing, etc [04:44] at least it probably could help to subclass it [04:45] if i understand how mime works at least [04:45] I got 2.4 on now [04:45] Hydrant: linux got easy to use? ;-) [04:45] ah, and here's one about the offsetting - 62349 [04:45] calc: yeah... I always Linux to succeed, now that it has I kinda miss the old days :-) [04:45] bryce: lp bugs or fdo bugs? [04:46] calc, oo.org bugs (searching that link you sent) [04:46] bryce: oh ok [04:47] so looks like pretty much all the issues I ran into already have bug reports [04:47] oh no, OOo Later means... you are screwed :-\ [04:48] i think it depends on who looks at the bug if it gets marked to be fixed or marked to go to bug hell [04:49] ah well [04:50] probably I can talk one of my minions into implementing it in inkscape for me ;-) [04:50] once i get access to the ubuntu master bug i can start adding things people want to get fixed and we can push that list [04:50] convince upstream to consider those bugs higher priority [04:50] calc: how do I get the SDK on ? [04:50] It only seems to have RPMs [04:51] Hydrant: not sure, i thought everything was in each download [04:52] nope :-( [04:52] how is ubuntu oo-sdk made? [04:52] is it in the regular rpm download? [04:52] Hydrant: just from regular compile [04:52] k, I'll alien and hope for the best [04:56] ScottK2: yes, I did; already nuked, thanks :) [05:56] good morning [05:57] Hey dholbach. [05:57] hiya TheMuso [06:00] who can help me debugging bug 208113? [06:00] Bug 208113 on http://launchpad.net/bugs/208113 is private [06:01] hum [06:01] somehow debuild crashes on me on amd64 (libperl S_regmatch) [06:02] when I try to valgrind it, memcheck crashes in a similar way as bug 150477 [06:02] Bug 150477 on http://launchpad.net/bugs/150477 is private [06:02] I just did a memory test and all seems good :/ [06:05] calc: the OOo rc5 doesn't yet support the 3D effects for impress? [06:06] have you guys ever seen the error msg "Tainted filename: ''." in ppa uploads? [06:09] oh wait I see [06:45] Good morning [06:54] sorry for not being on topic, but: i just tried to install 7.10 on a disk with dm-crypt + lvm encryption without formating my dm-crypt partition. alternate installer. i choose to setup encrypted lvm devices and was asked for a passphrase twice(!) [06:55] hi [06:56] now i cant access the two logical volumes in this dm-crypt partition. i think the installer called luksformat on this dm-crypt partition and destroyed the lvm information [07:04] calc downloading, build will take a a time ... [07:08] doko_: ok [07:10] Good morning [07:17] i've got a little package management question : I have KDE4.0 installed on hardy from the repositories, but I need QT 4.4, and so I'm compiling it all from svn. Now, how do I remove the KDE4 that i installed from the reopsitories? === \sh_away is now known as \sh === Seveaz is now known as Seveas [07:30] I just installed kubuntu 8.04 alt amd64, and it has the same bug that 7.04, and 7.10 had: [07:30] https://bugs.launchpad.net/ubuntu/+bug/86666 [07:31] also why doesn't it install firefox by default? honestly, I doubt that many people waste their time w/ konq [07:32] tomahasamoot: ....because it's a gtk app... [07:33] Hobbsee: that's true, but the answer is to do the proper intagration, as it's the only real web bowser [07:33] tomahasamoot: feel free to help out. [07:33] and plenty of people use konq, only. [07:36] Hobbsee: I'd like to see the break-down... as konq can't even do gmail, it's not a real contender for serious use... so I'm not sure what kind of useage pattern these konq users have? [07:36] gmail got fixed? [07:36] afaik, anyway [07:37] that's what I was told, but it wasn't working when they said it was... I'll try again [07:38] Some people have sane email providers, I'm afraid. For them, Konqueror is an option. === doko_ is now known as doko [07:39] the point is that if it doesn't work for something so basic as gmail... [07:40] * Hobbsee wasn't aware that "basic" meant "littered with ajax and other bits of that nature" [07:42] no... still not working [07:43] basic in the sence that it's one of the most important websites out there [07:44] does gmail render correctly in opera/safari/{other non-mozilla/ie browsers}? [07:44] i can't quite remember [07:44] Hobbsee: I don't know [07:45] * Hobbsee recalls something about user agents, etc. === tkamppeter_ is now known as tkamppeter [07:59] asac: do you need any more info on bug 208141? [07:59] Launchpad bug 208141 in network-manager "acx chipset unhappy with network-manager" [Undecided,New] https://launchpad.net/bugs/208141 === cprov-out is now known as cprov [08:07] doko: any chance you could reupload the sun-java5 dapper update with a bug number in the changelog? [08:09] Is said upload going to fix the plethora of CVEs? [08:13] apport removed need-i386-retrace from a bug of mine 20 minutes, but hasn't done anything since. [08:13] *20 minutes ago [08:13] Fujitsu: http://java.sun.com/j2se/1.5.0/ReleaseNotes.html doesn't mention CVEs in general :( [08:14] morning all [08:14] Fujitsu: I had to kill it, because the chroot was broken; maybe yuo can retag it yourself? [08:14] pitti: Will do. [08:14] pitti: The latest upload to Hardy fixed a lot. [08:14] doko: nevermind, accepting now; but pleeeease, pretty pretty please add bug # to changelogs in the future [08:16] pitti: `This release contains fixes for one or more security vulnerabilities. For more information, please see Sun Alerts 233321 233322 233323 233324 233325 233326 and 233327.' [08:16] They hide it remarkably well. [08:17] pitti: I'll add the upstream report numbers next time [08:17] doko: most important is the SRU LP bug# [08:22] asac_: do you need any more info on bug 208141? :-) [08:22] Launchpad bug 208141 in network-manager "acx chipset unhappy with network-manager" [Undecided,New] https://launchpad.net/bugs/208141 [08:24] huaynac_: hi [08:24] oops [08:24] dholbach: hi :) [08:24] dholbach: never seen this kind of access denied thing [08:25] dholbach: is it a hidden SSID? [08:25] yes [08:25] dholbach: does unhiding help? [08:25] asac_: let me try === dholbach_ is now known as dholbach [08:31] slangasek: i got some negative feedback on the bug i was waiting for answers for [08:32] Hobbsee, ScottK: WDYT about syncing http://packages.qa.debian.org/s/slony1/news/20080223T003124Z.html ? 8.3 is our default version in Hardy, so it would be good to have a matching slony [08:34] asac: I don't get the crazy warnings any more, but it does not connect either - updated bug 208141 [08:34] Launchpad bug 208141 in network-manager "[acx] cannot connect to hidden SSID" [Undecided,New] https://launchpad.net/bugs/208141 [08:34] Hobbsee, ScottK: (that's bug 191782) [08:35] Launchpad bug 191782 in slony1 "Slony1 1.2.13 for hardy? ( postgresql 8.3 capable )" [Undecided,Confirmed] https://launchpad.net/bugs/191782 [08:35] dholbach: when did you first test that chip? [08:35] dholbach: e.g. do you know if it worked in gutsy? === dholbach_ is now known as dholbach [08:37] dholbach: could you also please remove the remembered connection from nm-editor before trying the connect [08:38] asac_: I did that [08:38] (mentioned it in the bug too :)) === asac_ is now known as asac [08:39] dholbach: what about gutsy? [08:39] dholbach: another info helpful would be to know if WPA works [08:39] asac: ugh... don't have a gutsy CD here any more [08:39] I can try WPA though [08:39] dholbach: ok try WPA first then [08:43] asac: just normal "wpa personal"? [08:43] dholbach: yes [08:49] dholbach: i assume it doesn't work? [08:58] asac: no, it's seriously unhappy [08:58] attaching the log in a bit [08:58] asac: I remember: the problem in gutsy was a kernel issue (firmware was not loaded) [09:01] asac: log attached === dholbach_ is now known as dholbach [09:03] dholbach: it crashes? [09:03] asac: it only showed "WEP key something" when I selected the network, then I tried to select "shared key" and it crashed [09:04] how do I turn on apport on the livecd? :) [09:04] dholbach: thats on CD? [09:04] yes, on the livecd [09:04] thought it was an instlal ;) [09:04] dholbach: your driver doesn't support wpa [09:04] no, I wanted to see if everything works before I install ubuntu on my girlfriend's laptop :) [09:04] otherwise you won't see AP_SCAN [09:04] 2 [09:05] * dholbach can try WEP [09:05] you can, but chances are low [09:05] wpa is better than open in regards of getting feedback in user-space [09:05] ok... what do you think the problem is in this case? [09:05] dholbach: if you run iwconfig while connecting to a non-hidden network [09:05] the hidden ssid seems to irk NM, but that doesn't seem to be the real problem [09:05] do you see that th essid is set? [09:06] ok... let me get back to public non-hidden [09:09] dholbach: http://acx100.sourceforge.net/wiki/WPA i think an option would be to go for ndiswrapper [09:09] not sure how easy you can test that on livecd though [09:10] dholbach: the problem for open network is that the driver either never associates (which we are trying to figure out now) or it doesn't send a notification to user-space when association has happened [09:12] this sucks :) [09:12] dholbach: so ... do you see a essid set (using iwconfig) while connecting? [09:13] I'm just restarting the machine - network-manager was even more confused after it crashed (and I restarted it) - will let you know in a sec [09:14] ok [09:16] * pitti pokes soren about bug 196149 [09:16] Launchpad bug 196149 in dnsmasq "Please sync dnsmasq 2.41-1 (universe) from Debian unstable (main)." [Wishlist,Incomplete] https://launchpad.net/bugs/196149 [09:17] pitti: Oh, right. That one fell completely off my radar. [09:18] asac: yes, I can see it [09:22] yeah.. thats a driver issue then [09:22] dholbach: you could ask ben if its really the latest [09:22] why? [09:22] or try ndiswrapper [09:22] dholbach: because it associates, but doesn't tell userspace that association has happened [09:22] hum [09:22] thats required for networkmanager, but not for dhclient [09:22] dholbach: there is one more thing you could try [09:22] you could try to use AP_SCAN 1 [09:22] instead of two [09:22] so run AP_SCAN 1 sudo NetworkManager ? [09:22] so run AP_SCAN=1 sudo NetworkManager ? [09:22] no [09:22] that won't work :) [09:22] so? :) [09:22] get that branch: http://bazaar.launchpad.net/~ubuntu-core-dev/network-manager/ubuntu.0.6.x.ap_scan [09:22] oh ... hmm you can't do that on livecd i guess [09:22] dholbach: i think its really worth tto check with kernel folks that the driver is really the latest [09:23] it says 0.3.36 which is the latest release [09:23] 20080210 is a snapshot, I guess [09:23] I'll ask [09:23] http://sourceforge.net/project/showfiles.php?group_id=75380 [09:24] yes, i am not sure if 0.3.36 is the same as 20080210 [09:24] modinfo doesn't show version info [09:25] it shows in iwconfig [09:26] the driver version? [09:26] strange [09:27] Nickname:"acx v0.3.36" [09:33] if we have luck its really outdated and things are fixed in latest [09:41] hmm.... would be really nice, if update-manager would display which repository a package comes from [09:57] asac: it is outdated and the changes are pretty huge [10:00] dholbach: great. good to figure out now [10:00] dholbach: please point th ekernel team to the bug ... i added a linux-ubuntu-modules target by now [10:09] soren: did you test dnsmasq 2.41 in hardy with kvm? [10:09] soren: FF-wise it looks acceptable, so I'll make this dependant on your testing results [10:11] * soren blushes [10:12] Actually, I need to merge it. [10:12] Rather than have it synced. [10:12] I just noticed I changed stuff in dnsmasq after filing the sync request. [10:13] oh, ok [10:13] I'll mark it as invalid then [10:13] soren: if you are happy with the new version, FF exception granted [10:14] soren: ^ copied that to the bug, closed [10:16] * soren questions the integrity of his todo list [10:17] pitti: The changes I did in 2.40-1ubuntu4 were to address the issues you raised in bug 190905 [10:17] Launchpad bug 190905 in dnsmasq "Main inclusion report." [High,Incomplete] https://launchpad.net/bugs/190905 [10:18] soren: sweet [10:18] How do you debug memory leaks/problems? Valgrind? [10:19] http://wiki.ubuntu.com/Valgrind [10:19] pitti: Ok, so I'll do the merge test, report, and we can close the MIR as well? [10:19] soren: no, please don't close the MIR yet [10:19] I'll do it after promotion [10:20] but that needs the package be uploaded, built, and published first [10:20] otherwise promotion will fail [10:21] Ok, cool. [10:21] openoffice 2.4 is released, is it planed to provide it in Hardy ? [10:21] kagou: yes [10:22] oh nice seb128 :) [10:22] kagou: we try to not stick to candidate versions when the stable version is available [10:22] kagou: what would be the interest? [10:23] indeed [10:25] asac: I'd appreciate your input on bug 206875 [10:25] Launchpad bug 206875 in seamonkey "Please remove iceape from Hardy" [Medium,Incomplete] https://launchpad.net/bugs/206875 [10:26] would we need the new dpkg-dev in hardy if we wanted to use the new source package formats in intrepid, or does that only matter for dpkg and new .deb formats? [10:29] pitti: 11:14 < asac> Fujitsu: iceape is just in gutsy, right? [10:29] 11:15 < Fujitsu> asac: It will be once the removal is processed. [10:29] 11:15 < Fujitsu> asac: Somebody apparently forgot to remove iceape from Hardy, even though the binaries are now all provided by seamonkey. [10:29] 11:16 < asac> Fujitsu: yes, so the security issues are actually phantom ones :) [10:30] pitti, keescook: http://dvlabs.tippingpoint.com/blog/2008/03/19/cansecwest-pwn-to-own-2008 [10:30] pitti: Oh, I see I missed your comment. [10:30] pitti: so far we are holding up pretty good :) [10:30] Fujitsu: thats just a transitional package then [10:30] pitti: iceape source is still in Hardy... [10:31] iceape | 1.1.4-1ubuntu2 | hardy/universe | source [10:31] hm, seems that madison on drescher lies then [10:31] ok, so I'll kick that? [10:31] pitti: if the source is still there, then please remove [10:31] i invalidated the bug for now [10:32] based on your comment [10:32] The version in Hardy is lower than that in gutsy-security, which could do strange things. [10:32] pitti: did the sync of attal/attal-themes succeed? (the output you added to bug #205810 looks like it didn't) [10:32] Launchpad bug 205810 in attal "please sync attal (1.0~rc1+cvs20080318-1) and attal-themes (1.0~rc1+cvs20080318-1) from unstable/main to universe" [Undecided,Fix released] https://launchpad.net/bugs/205810 [10:33] sistpoty|work: in progress [10:33] pitti: thanks! [10:34] pitti: Thanks for removing that. [10:35] <\sh> pitti: when you are on sync crusade, please sync phpgroupware from debian unstable (bug #206948) [10:35] Launchpad bug 206948 in phpgroupware "phpGroupWare should be included in Hardy" [Undecided,Confirmed] https://launchpad.net/bugs/206948 [10:35] <\sh> pitti: it was removed from debian and hardy because ajmitch was MIA ... now debian has a new maintainer [10:37] <\sh> pitti: request is acked (+2) by motu-release aka scott and hobbsee :) [10:45] \sh: doing [10:45] pitti: i'm not against the sync. [10:47] \sh: argh, sync-source.py barfs on that [10:47] doko: I'm going through the milestoned bugs and I see https://bugs.launchpad.net/ubuntu/+source/gcc-4.2/+bug/184141 , would you like some help in creating that list? [10:47] Launchpad bug 184141 in gcc-4.2 "libstdc++6 has a binary incompatibility" [High,Confirmed] [10:49] james_w: yes, asked cprov and kiko (I do have the list for a) [10:50] doko: ok, sounds like you are on top of it, great [11:02] pitti: bug 191782 approved. [11:02] Launchpad bug 191782 in launchpad "Slony1 1.2.13 for hardy? ( postgresql 8.3 capable )" [Medium,Confirmed] https://launchpad.net/bugs/191782 [11:03] ScottK: great, thanks [11:18] asac: firefox-themes-ubuntu (main) b-deps on firefox-2 (universe); this needs to be fixed [11:19] pitti: can be demoted ... though steve already did that [11:19] thought [11:19] asac: demoted, thanks [11:20] asac: so I'll unseed it, ok [11:20] pitti: thanks to you [11:20] ? [11:20] obviously yes. [11:20] pitti: thanks [11:34] dholbach: wishlist: would be great if "update-signature yesterday" worked :) (date = time.localtime(time.time()-24*3600)) [11:35] bug 208204 [11:35] Launchpad bug 208204 in five-a-day "pitti dholbach: wishlist: would be great if "update-signature yesterday" worked :) (date = time.localtime(time.time()-24*3600))" [Wishlist,Confirmed] https://launchpad.net/bugs/208204 [11:35] * dholbach goes back to lunch :) [11:35] dholbach: yay your irc2lpbug script [11:35] * pitti hugs dholbach [11:36] hehe [11:55] pitti: when will you write a script that *fixes* bugs while being on irc? [11:55] hardy says the nvidia driver is installed but not in use. how do i use them? [11:57] Hobbsee: I have one, but it depends on libjustdoit to work :P [11:57] +++++++++++++++++++++++++++++ [11:57] sistpoty|work: ahhh [11:58] so, I guess we'll need to wait until debian bug 456789 is fixed [11:58] Debian bug 456789 in gclcvs "gclcvs: FTBFS: Uses way too much RAM." [Serious,Open] http://bugs.debian.org/456789 [11:59] erm... not that one /me shuts up again [11:59] sistpoty|work, only with versions that ship /usr/share/libjustdoit/caffeine.in [11:59] heh === heno_ is now known as heno [12:14] does anything call pm-powersave? [12:41] <\sh> pitti: uga....worked? [12:41] <\sh> pitti: ah ok....fakesync...doing it (phpgroupware) [12:42] heya [12:43] I install hardy using wubi, did 100MB of udpates, restarted and then got "initramfs" prompt. whats gone wrong? [12:46] tuntun: This inst a support channel, please ask in #ubuntu, thanks. [13:18] * cody-somerville waves at davmor2. [13:18] cody-somerville: 0/ [13:23] i just installed the Hardy Beta on my laptop and it looks pretty sweet - well done [13:32] evand: hi [13:32] Tm_T: howdy [13:33] evand: final is going, judges are voting [13:33] great! [13:33] nervous? [13:33] evand: I had several hardware failures so my presentation was a tad bit raw but it went mostly ok [13:33] ok [13:34] not anymore, I got my presentation ready while others were presenting theirs :-P [13:35] heh [14:09] pitti, is there a way to override the usage of /sys for batteries and force hal to /proc ? [14:09] seems the cmpc doesnt report all info in /sys but /proc worked before [14:11] * ogra_cmpc has 1033975 h battery time remaining with 36% atm [14:11] impressive [14:11] ogra_cmpc: fix whatever report the infos to have those in sys? [14:11] * Haegin is jealous [14:11] ogra_cmpc: That sounds fixable in the kernel [14:12] mjg59, seb128, right i just wanted to know if we still support a manual override for falling back [14:16] ogra_cmpc: No [14:16] theks [14:16] theanks [14:17] gah === iceman__ is now known as iceman [14:24] evand: first try of wubi on an up-to-date vista and it seems to be working fine :) [14:24] so far :) [14:24] good news [14:27] I'll let you know when it finishes :) [15:06] ogra_cmpc: not that I know of [15:40] evand: no luck, but I'll do it === ogra_cmpc is now known as ogra [15:44] ok [15:46] evand: worked fine Just need a way to backup vista now :( [15:47] davmor2: dd :) [15:49] evand: might be partimage doesn't like it (invalid compression). But it's faster than dd so I'm going to be hitting it with a mallet to see if I can get it to work. [15:49] heh [15:52] evand: partimage is complaining it can't read the image. So I'm wandering if it got corrupted somehow. So I'm going to reinstall Vista do all the updates again and try again. If it fails again I'll look at other alternatives :) [15:52] ok [15:54] fabbione: yeah, I've been keeping tabs on it by remote [16:14] network manager is breaking my balls, here === \sh is now known as \sh_away [16:30] sabdfl: 0.6.6-0ubuntu4 ? [16:31] 3 [16:32] sabdfl: ubuntu4 should be on the mirrors already. it ships some iwl tweaks [16:32] asac: here's what blows my mind [16:32] with a 4-line wpa.conf (that basically just names the essid and password) and dhclient it ALWAYS works [16:32] and fast [16:33] wtf is up with NM? [16:33] is it just a POS? [16:33] network manager depends on some driver features that are not needed for dhclient [16:34] e.g. send associate event to user space once your interface successfully associates with your AP [16:36] asac, is that what dhcdbd provides ? [16:36] ogra: no, that provides a dbus server to run dhcp [16:36] ah, k [16:49] asac: Is network manager not restarting the network after it updates itself a bug, feature, known limitation? [16:50] tkamppeter: I think I see the problem in bug 195782, just posted a followup [16:50] Launchpad bug 195782 in hplip "Users not automatically added to "scanner" group: No scanning functions of HP multi-function in Hardy" [Medium,Confirmed] https://launchpad.net/bugs/195782 [16:50] ScottK2: what do you mean by restarting the network? [16:50] /etc/init.d/networking restart? [16:51] asac: I just updated using apt and was not connected to the network when the update finished. [16:52] pitti: Would you please look at Bug 208097 and tell me who I ought to chase after to see if the package can be removed? [16:52] Launchpad bug 208097 in python-aptsources "FTBFS in Hardy due to python-distutils-extra changes" [Undecided,New] https://launchpad.net/bugs/208097 [16:53] ScottK: hm, the change of p-distutils-extra is my credit/fault [16:53] but I actually didn't introduce any compatibility breaking things, hmm [16:53] ScottK2: can you still connect? [16:53] asac: Yes. [16:53] asac: Manually reconnecting worked fine. [16:53] ScottK2: please upload your syslog somewhere (that contains the incident) [16:54] asac: Sure. [16:54] asac: I'll take me a few minutes to sort through it. [16:54] ScottK2: i usually ask for the complete one [16:54] ScottK: it has no rdepends at all, so I don't mind removing it if you think it's useless [16:55] and it's not even in debian [16:55] pitti: It may be useful to someone, but it's been untouched since Feisty. [16:55] My thought is if it was being used, it'd have an rdpend somewhere. [16:55] asac: OK. I'll get it. [16:58] asac: This is Kubuntu KDE3 Hardy. http://paste.ubuntu-nl.org/61387/ [17:00] have a great weekend everybody [17:02] dholbach: cheers [17:02] bye asac [17:03] ScottK2: didn't it reconnect at line 40++ ? [17:04] oh blind me :) [17:05] OK. I didn't watch the upgrade, just came back to it and found it not connected. [17:05] ScottK2: can you reproduce by running apt-get install --reinstall network-manager? [17:05] I'll give it a shot. [17:05] ScottK2: 12:48:08 <- it starts [17:06] ScottK2: 12:48:16 <- IP retrieved [17:06] asac: I just did the reinstall and it's off the network [17:06] ScottK2: 12:48:18 KTS-D430 NetworkManager: Activation (wlan0) successful, device activated. [17:06] ScottK2: yeah ... wait a bit [17:06] OK. [17:06] it should reconnect after a while [17:06] (yes, it tears down the network ... which is a different beast) [17:07] asac: Is this while seconds or minutes? [17:07] depends on when the wpasupplicants scan kicks in [17:07] ScottK: can you see APs in the list already? [17:08] asac: Yes [17:08] ScottK: try nm-tool to see if NM already sees networks [17:08] zul: has debian bug 365097 been addressed in our bacula? [17:08] Debian bug 365097 in bacula-common "bacula-common: Uses the same passwords on every Debian installation" [Important,Open] http://bugs.debian.org/365097 [17:08] (maybe kde applet doesn't behave like gnome thing) [17:08] pitti: yes it has [17:08] asac: It does [17:08] zul: thanks (shoudl be pointed out in the MIR) [17:09] pitti: yep [17:10] asac: Still no network. What now? [17:11] ScottK: please stpo the kde applet and see if it works for you with gnome-applet [17:11] (i guess it should work) [17:11] on kde [17:11] just to see if its something applet related [17:12] OK. I'll have to reconnect to install that and then I'll trip the reinstall again. [17:13] asac: What's the package name for the gnome applet? [17:14] keescook: awake already? [17:14] ScottK3: network-manager-gnome [17:14] ScottK3: you start with nm-applet --disable-sm [17:14] pitti: I am, yup [17:15] asac: Thanks [17:15] keescook: WDYT about https://wiki.ubuntu.com/MainInclussionReportLibArchive? [17:15] keescook: it's a nice library, but currently dup's code from cpio and tar, thus is security relevant [17:15] keescook: so I'd like to get your "ack, we'll be aware of it" at least (or a "no, no, not that one!!!") [17:16] pitti: I'm not quite "no no not that one" about it, but I'm not really a fan of it. [17:16] 3 CVEs last year [17:16] same for me [17:16] (though they are fixed in hardy) [17:17] note that the cpio and tar code are the bsd ones [17:17] so they are not new and unused copies [17:17] ironically, the older the archive code, the buggier it is, seemingly. [17:19] does it have a test suite? [17:19] (I think that should be added to the MIR QA questions...) [17:19] keescook: yes, it has [17:19] let me try it [17:20] jdstrand: what is your opinion on libarchive? [17:20] AFAICS it is supportable, but actually needs a little effort over 5 years [17:20] keescook: all I know about it is that it embeds bsdtar and cpio [17:20] I'm not super keen on that [17:20] yeah, that makes me rather unhappy too [17:21] switchign over our tar to bsdtar by default is late for hardy, but considerable for intrepid [17:21] honestly, if cpio and tar were switched to use libarchive, that'd work for me. ;) [17:22] I wish we would solve the "main package can't have universe build depends" [17:22] IMHO that shouldn't be 'solved' in any way [17:22] that's hurting us right now because we have the choice to drop interesting features because we don't want to promote things or to not provide those to users [17:22] considering this is LTS, and libarchive is still seeing vuln reports, I'd prefer it not go into main. is there is really good reason for this to be in main? [17:22] keescook: test suite isn't run by default on package build, but a simple "make check" does [17:22] a quick search shows 4 CVEs in the last 2 years-- 3 all in archive_read_support_format_tar.c [17:23] jdstrand: I guess tar itself had the very same? [17:23] pitti: well, I consider not shipping cool stuff because we don't want to new cracks an issue === kitterma is now known as ScottK2 [17:23] pitti: well, I consider not shipping cool stuff because we don't want to support new cracks an issue [17:23] seb128: yeah, hrm [17:23] and there is no way to solve that [17:23] right now we don't build epiphany using webkit [17:23] pitti: that isn't the bsdtar code [17:23] we don't have the gvfs archive code [17:23] etc [17:24] jdstrand: ah, so the default tar has completely different code? i. e. gnu tar and bsd tar are totally separate? [17:24] it's just reading a tar archive bits (AFAICS from looking at this for 10 seconds) [17:24] asac: Works with the gnome applet, so I guess we get to blame the KDE applet. [17:24] * ScottK2 goes offline to remove gnome bits from his machine ... [17:25] seb128: if you feel it is an important feature, I would say "okay" as long as the test suite is enabled by default (and will fail the build on a failure). [17:25] keescook: well, that gives the ability to browse iso and archives from nautilus in a transparent way [17:26] pitti: 'the 'bsdtar' program is a full-featured 'tar' replacement built on libarchive [17:26] (from the README) [17:26] the code I was talking about is in libarchive [17:26] "libarchive: a library for reading and writing streaming archives" [17:27] libarchive of course provides bsdtar, but it is different code [17:27] (it didn't always though) [17:27] * keescook thinks we should add fuzzing to the Security section of the MIR for things that process stream or file data :P [17:27] keescook: which was a thing rated quite high on the new ideas website and that fedora already provides now [17:27] * keescook nods [17:28] but I agree that having to support extra crack to get features is not optimal [17:28] but neither is not having those feature, especially when other distro do have those and user are wanting to try them [17:28] hm, I damaged the code on two places, and the test suite still reports 'all passed', hmm [17:28] libarchive wasn't written yesterday, and they've been okay about fixes. so, like I said, since it has a testsuite, if that can be enabled in the build, I would object less. :) [17:29] pitti: ewww [17:29] keescook: I updated embedded-code-copies this week for libarchive === kitterma is now known as ScottK2 [17:30] * LaserJock hugs pitti [17:30] pitti: uhhh, icky [17:30] asac: Anything else I need to gather for a bug report? [17:31] ah, my 9th damage was finally caught [17:32] keescook, jdstrand, seb128: so, if it's so uber-cool, and you security guys are ok with it, do you agree with main? [17:32] the weakness of the testsuite bugs me. [17:33] doesn't look like cpio has tests [17:33] fyi [17:33] ScottK2: you could trace the dump the dbus traffic for the relevant endpoints [17:33] s/trace the// [17:33] tar and libarchive do [17:34] keescook: admittedly those were pretty random changes [17:34] asac: Thanks. [17:34] ScottK2: i have the feeling that kde applet doesn't deal with some dbus message [17:35] Probably right. Looking at the bug list for it, it looks like it needs more than a little love. [17:35] ScottK: what i don't understand is that you connect in the syslog [17:36] ScottK: do you still see that in syslog (e.g. stage 5 finished) when --reinstall ? [17:36] asac: One thought that just occured to me is that I may be connected and the applet doesn't know it. [17:36] * ScottK has been believing the U/I. Maybe it's just wrong. [17:36] ScottK: if the syslog snippet captures the relevant bits, yes. [17:37] ScottK: it ends with: Activation (wlan0) Stage 5 of 5 (IP Configure Commit) complete. [17:41] pitti, seb128: okay, here's what I propose: 1) enable failure-sensitive testsuite in the build, 2) promote libarchive to main, but leave the bsdtar binary in universe. [17:42] keescook: WFM [17:42] WFM too [17:42] keescook: although the bsdtar package is innocent; it's the library which has the code [17:42] but we don't need it for gvfs [17:42] (if we did, it could use /usr/bin/tar in the first place) [17:42] right, but it doesn't actually link to libarchive, which bugs me. :P [17:42] pitti: but I really thing we have an infrastructure problem there which is worth discussing and trying to solve [17:42] keescook: oh, urgh; static linking apparently [17:43] s/thing/think [17:43] seb128: can you please promote the source and the libs? -ESSH again, sorry (and I need to leave soon) [17:43] * pitti kicks his ISP [17:43] pitti: sure, thanks [17:43] the one single reason why I want to move to a different place [17:44] someone needs to enable the testsuite in the build too, seb128, can you do that? [17:44] I'm at it [17:44] keescook: yes, do you mind if there is a small delay between promotion and the change? (ie, can I promote it now and then fix libarchive?) [17:44] ah, cool [17:44] pitti: oh, okay [17:44] * seb128 hugs pitti [17:45] seb128: we can also promote it later [17:45] whichever you want [17:45] but I need to leave in 5 [17:45] seb128: no problem with a delay -- I just care about the final hardy state. :) [17:45] I promote it now [17:45] I want to rebuild gvfs using it today ;-) [17:45] why does mksquashfs hang, I wonder? [17:45] pitti, I saw a note from Fujitsu that vlc can't be demoted to multiverse. Do you know why by chance? [17:45] seb128: I'll upload tomorrow then [17:46] pitti: ok === kitterma is now known as ScottK2 [17:48] asac: I just tried the reinstall several times and it never got past stage 2. I think when it got to 5 in the syslog you were looking at it was when I manually activated it. [17:49] seb128, keescook: new package prepared and tested; will upload tomorrow morning when the promotion is published [17:49] asac: Looking at the syslog, it looks to me like a key management issue. Is this really what it reads like: Mar 28 13:43:46 KTS-D430 NetworkManager: nm_dbus_get_wireless_user_key_done(): nm_dbus_get_user_key_for_network_cb(): dbus returned an error. (org.freedesktop.NetworkManagerInfo.GetKeyError) org.freedesktop.NetworkManagerInfo.GetKeyError [17:49] I'll send the diff to Debian, too [17:49] pitti: ok, thanks [17:50] * pitti waves [17:52] * ion_ integrates pitti's wave [17:53] ScottK2: yes, that looks scary, but why could it connect a few seconds later? [17:53] pitti: could you shove libghemical-* through binary-new, to make LaserJock happy? [17:53] asac: Good question. [17:54] asac: I'll see if I can pursue it in #kubuntu-devel because I'm going to guess this is wallet related somehow. [17:55] pitti, keescook, seb128: disregard my comments on cpio-- cpio/cpio.c isn't compiled in our build [18:00] pitti, thank you. I have answered. My idea now for a better script is to generate an fdi file with entries for all possibilities which match the wild cards of /etc/udev/rules.d/55-hpmud.rules. This would be 5 * 256 entries, or would this slow down HAL a lot? [18:03] <__mikem> If I wanted to do opengl development, what package should I install? [18:05] How can I increase the maximum number of semaphores? I need to do this as part of some testing I'm doing at my job. [18:07] pitti, keescook: /bin/bash ./libtool --tag=CC --mode=link x86_64-linux-gnu-gcc -Wall -g -O2 -static -I/libarchive -Wl,-z,defs -o [18:07] bsdtar tar/bsdtar-bsdtar.o tar/bsdtar-getdate.o tar/bsdtar-matching.o tar/bsdtar-read.o tar/bsdtar-tree.o [18:07] tar/bsdtar-util.o tar/bsdtar-write.o -larchive -lacl -lz -lbz2 -lattr -lacl [18:08] I thought trackerd got shot in the head for hardy? [18:08] so it looks like it statically liks to libarchive and other libs [18:08] lamont: Nope, it just doesn't index ~ by default. [18:08] it still chews up an entire core forever [18:09] 1887 lamont 39 19 56124 35m 2908 S 99 1.7 129:31.59 trackerd [18:09] lamont: Hogtied, not shot. [18:09] There's a bug then. [18:09] well, there's also 433GB of /home [18:10] ]The following packages will be REMOVED: [18:10] libdeskbar-tracker* tracker* tracker-search-tool* [18:10] there. bug fixed [18:10] Your older settings may still have indexing on. [18:10] Oh, probably... this machine started life as feisty, iirc [18:10] and my total involvement with trackerd has been to terminate it with predjudice whenever I notice that it's chewing up the machine... [18:11] I understand that it does wonderful things, if it ever finishes indexing [18:11] OTOH, I also haven't run evo since about hoary time [18:12] in other news, I need to figure out WTH gnome did to my sshagent stuff, and how to fix my sledgehammer to do what I want, or else figure out how to use seahorse or whatever [18:13] ion_: and I have no clue where I'd even look to find out what my trackerd settings are [18:14] You could open the tracker settings window. :-) [18:15] slangasek: okay, how do i make d-i install an udeb in the alternate disks? the mythbuntu disks aren't quite there yet.. [18:16] laga: set the udeb priority to standard? [18:16] laga: what's the udeb? [18:16] slangasek: mythbuntu-diskless-client-builder [18:17] where would i report bugs in translated strings in package gdm? [18:17] upstream? [18:18] caci: upstream or language-pack-gnome-locale-you-are-using [18:18] laga: ok, I'm having a look inside the image [18:19] thanks [18:19] sladen, Hi [18:19] (rather, I will be when jigdo finishes :) [18:20] seb128: any reasonable chance something happens? [18:20] can we imagine that samba password will be integrated by PAM for Hardy or is it too late for this ? [18:21] ideally we would get help from the server team to drive that forward [18:21] slangasek: note: i havent made any serious attempt to get it installed. just a warning ;) [18:21] laga: I'm not trying to install from it, just look inside it [18:23] pitti, keescook: re bsdtar and staic libs> just libarchive is statically linked, the others end up dynamic (libtool-fu?) [18:26] laga, seeding it should be enough to have it in the installer [18:27] then you just need a preseed file that calls it ... i.e. ltsp-build-client/run=true is the call for the ltsp udeb [18:27] ogra: i am seeding it and it actually is included on the disk.. or are you talking about preseeds? [18:27] both [18:27] yeah, the preseed is in place as well [18:27] seeding is to get it in, preseed to execute [18:27] laga: if mythbuntu-diskless were reuploaded with mythbuntu-diskless-client-builder set as Priority: standard, that should be sufficient [18:28] ogra: so, seeding makes sure it is installed, and preseeding triggers code inside the postinst. [18:28] laga: it currently isn't Prio: standard in the archive, so it doesn't get set Prio: standard in the CD [18:28] well, preseeding sets vars [18:28] then you could forego any preseeding [18:29] (and in the general case we would use preseeding instead of twiddling priorities, because once you've set it Prio: standard, it's Prio: standard for everything; but in this case, mythbuntu-diskless-client-builder isn't going to be on any images except the ones you want..) [18:29] in case of ltsp i check the debconf value for the run variable ... if thats true i run, else i skip [18:29] ogra: yes, that's what i meant [18:29] just curious: if i left out the preseeding, would the udeb still get installed? [18:30] installeed, yes, thats handled by the installer seed usually [18:30] slangasek: just to make sure: we're talking about this priority: http://www.pastebin.ca/961148 [18:30] ogra: ah, nifty. i thought preseeding just set debconf values. [18:30] yes, it is [18:30] laga: not the priority for the source package; you want to change the priority *only* for the udeb [18:31] so you would add a Priority: standard field to the per-binary stanza of debian/control [18:32] okay [18:32] thanks everyone for their help :) === Positronic_ is now known as Positronic [18:36] blueyed: ping? [18:37] slangasek: pong [18:37] blueyed: hi. anyone else stepping forward on bug #203736 yet for sponsorship, or should I have a look? [18:37] Launchpad bug 203736 in hdparm "FFe: Please merge hdparm 8.6-1 (main) from Debian unstable" [Wishlist,New] https://launchpad.net/bugs/203736 [18:38] slangasek: dunno?! please sponsor it, if you want/can, of course. [18:39] blueyed: ok, looking [18:40] slangasek, do you want that i open a bug (for tracking informations) about integrating samba/pam password ? [18:40] kagou: if there isn't one already, then I think it should be opened as a bug, yes (on the pam package) [18:48] slangasek, ok, i check this [18:55] apport makes firefox take too long to crash [18:58] Disable it! [19:01] slangasek, i'v done Bug #208419 can you check that i'v well explain the problem (my english is not good enough) [19:01] Launchpad bug 208419 in pam "Integrate samba password in PAM" [Undecided,New] https://launchpad.net/bugs/208419 [19:04] Should privacy bugs get the security tag? I'm currently feeling uptight about Bug #208399. [19:04] Launchpad bug 208399 in lshw "[Hardy] lshw 02.12.01-2 phones home (with tcpdump example)" [Undecided,New] https://launchpad.net/bugs/208399 [19:07] dajhorn: I'd prefer a patch that added a "--check-for-new-version" flag or something rather than just ripping the whole thing out. [19:07] lots of software do these kinds of new-version checks (firefox extensions, for example). [19:08] kagou: looks good, subscribing myself/ubuntu-server [19:08] keescook: Would it be similarly appropriate for bash or gzip to phone home? Network access from Firefox is expected. [19:09] keescook: I consider it very bad form for any software I installed as a distro package to do its own new version checks [19:09] dajhorn: it's a bit odd, I'll give you that. [19:10] slangasek: yeah, true. that's the other point in the bug. I think we should take that patch. [19:12] keescook: I noticed the change doing a security audit, but I didn't notice a changelog entry or other commentary. [19:12] * keescook nods [19:12] dajhorn: code review or saw the traffic? [19:13] keescook: First I noticed different linkages. [19:14] keescook: The upstream code is clean, it is just a simple query. Nevertheless, I trust the Ubuntu distro, not each and every upstream developer. [19:15] yeah, it's certainly not a least-surprise issue [19:15] or, rather IS a least-surprise issue [19:15] (rebuilding it now, should have it uploaded shortly) [19:16] I'm sure the bug was introduced in good faith. I'm much more interested in how this fits policy. === hunger_t is now known as hunger [19:16] Somebody subscribed the security list on the original bug before it got nailed shut. [19:22] keescook: I see the commit now. (Thanks.) [19:22] * dajhorn returns to bug hunting. [19:22] dajhorn: you're welcome -- thanks for catching it [19:25] blueyed: your hdparm diff introduces a debian/hdparm.default file not present in the current package or the Debian package, and not documented in the changelog [19:26] slangasek: are you sure? I cannot find something relevant in the diff. [19:26] slangasek: which line? [19:27] I cannot find even "hdparm.default" as string.. [19:30] blueyed: I was sure, now I'm less so :) Let me see what's going on.. [19:31] blueyed: ok, no, it's present in the Debian version, it just shows up as a diff relative to what MoM produces [19:31] slangasek: if you have a second, could you take a look at bug 208436. Should we go ahead and patch the version in hardy? [19:31] Launchpad bug 208436 in cairo "Latest cairo has an api change which causes problems" [Undecided,New] https://launchpad.net/bugs/208436 [19:31] so MoM is getting something wrong [19:32] slangasek: could be, I've used DaD [19:33] protonchris: wouldn't it be just as straightforward, and more future-proof, to fix this in cairomm [19:33] ? [19:34] slangasek: well I am concerned that the api change will cause problems in other packages as well. [19:35] protonchris: well, should be testable by rebuilding all of the cairo reverse-depends and seeing if any FTBFS. I'm ok with patching it for hardy, though I don't think it should be treated as a blocker [19:37] I'm inclined to say that breaking stable APIs is a Bad Thing. [19:37] (Especially when upstream has reverted the breakage) [19:37] but haven't considered the breakage severe enough to do a new release yet [19:38] and APIs << ABIs :-) [19:38] Well, I am up for either patching cairo or cairomm. Both have upstream patches that should be easy to include in the packages. [19:39] I'd fix both, if it were me. [19:40] They're both small and "obviously correct" patches. [19:44] Thanks. I'll fix cairo first and cairomm if I have a chance. [19:45] The cairo fix should render the cairomm fix less urgent, but I think it's more obviously correct to apply both patches. [19:59] Do I need a FFe for https://launchpad.net/bugs/208436 [19:59] Launchpad bug 208436 in cairo "Latest cairo has an api change which causes problems" [Undecided,New] === nakkytar is now known as nacitar [20:00] protonchris, Does the upload introduce new features? [20:01] blueyed: hmm, there's a bug in the Ubuntu patch for hdparm - there's code in the preinst to remove the hdparm init script on upgrade, but the init script is still being shipped in the package. Any chance you'd be willing to take a look at that before upload? [20:01] protonchris: no, you don't [20:02] slangasek: introduced with the merge or already with the current package? [20:04] slangasek: looking [20:04] pitti: my opinion as a user on the libarchive issue is: if you do, please do not switch over gnu tar to bsd tar [20:04] blueyed: already in the current package [20:05] pitti: gnu tar, in typical gnu style, most likely has a lot of command-line options that don't exist in bsd tar (though I would have to check that) [20:05] I use tar from the command-line quite often, and I do think quite some people would be upset with having to re-learn tar [20:06] slangasek: so the init script should be dropped, too? (and the version in preinst adjusted) correct? === luisbg_ is now known as luisbg [20:08] slangasek: it probably gets added with every merge again, hmm? maybe only the version in the preinst needs to be bumped?! [20:08] * Chipzz wonders how feasable it would be to dlopen tar and invoke the needed functions that way instead of including a seperate copy [20:09] blueyed: no, it's not the preinst that needs to be bumped - the debian/rules needs to be fixed to not install the init script at all [20:10] well - the init script should *also* be bumped, but as a one-time change accompanying the debian/rules fix [20:10] s/init script/preinst script/ [20:10] slangasek: ok. [20:23] slangasek: well, dh_installinit gets called by debhelper.. and it finds the .init/.default files and installs them. We use "--noscripts", but that's something different. [20:24] slangasek: I'll try adding --onlyscripts, too. [20:24] blueyed: patching the init script out of the package completely may be the easier solution [20:24] slangasek: yes, but it gets merged back every time, doesn't it? [20:25] I don't know [20:25] if so, that's an error on the part of the person doing the merge :) [20:27] yes.. :) but we could make it easier for the merger next time.. === nakkytar is now known as nacitar [20:49] slangasek: updated diff for bug 203736 [20:49] Launchpad bug 203736 in hdparm "FFe: Please merge hdparm 8.6-1 (main) from Debian unstable" [Wishlist,New] https://launchpad.net/bugs/203736 [20:49] works with "--noscript". [20:49] ..well, and "--onlyscripts" [20:58] hm... if I have an or'd builddep left | right, which one will get drawn in by default by soyuz? (and will it draw in the other one, if the first cannot get satisfied?) [20:59] (/me only remembers, that it used to be the oppisite of pbuilder long ago) [20:59] sistpoty: It will pull in the left build-dep first, and go through the list until it finds one available. [21:00] infinity: thanks... (I'm trying to get ghc6 up on lpia again, and the problem is that haskell-utils is missing, which however should build with hugs as well) [21:00] sistpoty: Ahh, I was planning to bootstrap that sometime when I was bored, but if you can do it with a few bootstrap uploads instead, that works just as well, and frees up some time for me. [21:01] infinity: *if* haskell-utils still works with hugs, it should be just one upload and a few give-backs... hppa is a different story though :( [21:02] infinity: if haskell-utils doesn't work, I could do a local bootstrap, and do some bootstraps upload for lpia in my ppa, if that helps? [21:02] sistpoty: I wouldn't waste too much sleep on hppa. If we want it fully built, lamont and I can take a weekend and do something about it. [21:02] heh [21:03] sistpoty: And, no, PPA bootstraps won't help me much, because it still introduces trust issues (nothing personal, of course), so if we *need* a bootstrap, I'll have to do it myself. [21:04] infinity: well, you could take the ghc6 binary packages for lpia to very much ease a personal bootstrap ;) [21:04] sistpoty or infinity: if either of you have time to take a look. I have attached a debdiff to Bug 208436 . [21:04] Launchpad bug 208436 in cairo "Latest cairo has an api change which causes problems" [Undecided,Confirmed] https://launchpad.net/bugs/208436 [21:04] protonchris: sorry, I'm just a motu and cannot upload to main then [21:05] protonchris: Looks correct to me. [21:05] sistpoty: Whoops. Thanks anyway. [21:05] infinity: thanks. [21:07] evand: ping [21:10] evand: don't worry sorted :) [21:12] ok [21:16] what's "worse" in german? Like in "worse than ..."? [21:16] no germans around? [21:16] evand: just a quick query about M-A which you had emailed to me but I couldn't find. But I found it :) we weren't sure if vista was supported and I felt sure you said you were working on it for intrepid but I thought it was through a gsoc project or something :) [21:16] "schlechter als" according to /usr/bin/translate-bin [21:16] "worse than", that is. [21:17] danke [21:18] ion_, simira: translate-bin is right with "schlechter als" [21:20] any weirdnesses with NM/iwl3945. My laptop doesn't seem to "see" connections anymore [21:21] slangasek: For the record, that cairo bug has caused some FTBFSes (which I was just about to report), so rather than report the bugs, I think I'll just upload the fix. :P [21:21] slangasek: https://lists.ubuntu.com/archives/ubuntu-autotest/2008-March/018788.html for example. [21:23] Meh, firefox needs an option to "prefer browsers on current active desktop" when opening links in new tabs. Searching all my desktops for the firefox that was the "top window" and got the link is annoying. [21:23] infinity: ok then :) [21:25] dpkg-buildpackage: error: fakeroot not found, either install the fakeroot [21:25] HATE NEW COMPUTERS, ARGH. [21:26] Also hate packages that need dpatch in the clean target. *sigh* [21:26] infinity: That's why you should have your own metapackage handy that depends on all the essentials. :-) [21:27] protonchris: Thanks for doing the heavy lifting, uploading your change now. [21:28] infinity: no problem. I am new to packaging. So if you have any advice, I would love to hear it. [21:28] protonchris: The patch was pretty much perfect, as-is. Changelog was decent, patch itself was correct, bonus points for correctly using the patch system that the package already had in place, etc. [21:29] protonchris: I'm just sponsoring the upload with your name in the changelog, as I didn't feel the need to "fix" anything. [21:29] infinity: great. thanks. [21:31] protonchris: If you're not already, can you subscribe to the bug, to follow up and make sure the upload doesn't explode? [21:32] protonchris: Ahh, and you already have. Excellent. [21:32] sistpoty: re hppa: what infinity said; [21:32] lamont: Doesn't nick hilighting on arch tags make you grumpy some days? :) [21:32] lamont: heh, it's priority whishlist on my radar... [21:35] infinity: thanks. I'll keep an eye out for any explosions. [21:37] can anybody translate "Vorbereiten zum Ersetzen von" [21:38] LaserJock: "prepare to replace " [21:38] "Prepare for replacing" says /usr/bin/translate-bin, and i think the "von" means "of" [21:38] Today ubuntu is build to support all legacy CPU's is there any chance for maybe building a ubuntu port with for modern CPU's with all the nice CPU features and instructions set enabled? [21:38] or is ubuntu 64bit the best option? [21:38] bardyr: yes, that's called "amd64" ;) [21:39] hmm [21:39] laga, Okay, thanks :) [21:39] i'm by no means an expert on that topic, but amd64 might be a bit better with regards to opimization.. recompiling ubuntu just to take care of the newest instruction sets on your CPU won't do much good [21:40] maybe for some very specific apps [21:40] could I get somebody with Hardy and a fast internet connection to install gcompris? [21:40] LaserJock, 20mbit hardy :) [21:40] LaserJock: yes [21:41] I just need to make sure version 8.4.4-1.1ubuntu1 installs [21:42] LaserJock, installing atm [21:43] LaserJock, installed without a problem [21:43] it installs just fine [21:43] hum [21:43] ah, now it's started [21:43] that took a while [21:43] omg, what have i installed :D [21:43] bardyr: ok, thanks [21:44] bardyr: one of the best educational applications the free software world has ever created ;-) [21:44] meh... seems like soyuz isn't that smart as I thought... haskell-utils build-depends on ghc6, which depends on haskell-utils, which is not installable, but it doesn't consider the alternative :( [21:44] gcompris has an interesting UI [21:45] LaserJock, kickball is great :D [21:45] LaserJock, and i love the voice [21:45] goal: fine motor coordination. i wonder if kids who need to be taught to use a mouse will understand that ;) [21:45] bardyr: you can also do fun things like play chess and play with electrical circuits if you have gnucap installed [21:47] LaserJock: i get some warning about missing gnomevfs libaries (i'm running kde). i guess that's ok [21:48] hmm, not sure [21:48] I'd have to look at it [21:49] LaserJock: http://www.pastebin.ca/961415 [21:51] laga: interesting, I've not seen that before, although I normally run it in Gnome [21:52] it seems to run fine [21:52] at least tux landed on the boat [22:01] bug 89269! [22:01] Launchpad bug 89269 in acpi-support "power.sh: wrong laptop_mode activation" [High,Triaged] https://launchpad.net/bugs/89269 [22:04] I made my irb calculate 89269!, but it's unsurprisingly taking a long time. :-P [22:05] heh [22:07] blueyed: hdparm uploaded, thanks! [22:11] slangasek: thanks. Do you mind looking at major acpi-support issues? Like the one above? [22:11] I have a bzr branch for it.. [22:13] infinity: you used my name [22:13] blueyed: ENOTIME right now, sorry [22:13] infinity: I don't highlight on arch tags [22:13] lamont: LIES. [22:14] blueyed: ah, uploaded and rejected because I built with the wrong options, oops fixing :) [22:14] infinity: statistics?? [22:19] It's still computing 89269!. Perhaps i should have done that on a faster box, or not in an interpreted language. ;-) [22:21] ion_: maybe Java would've worked ;-) [22:22] I don't have a stick long enough to dare touch javur. :-( [22:25] infinity: please give back happy and alex on lpia :) [22:25] I object to both, just because they have silly source names... [22:26] heh [22:26] sistpoty: So, you've made haskell-utils installable at some point, I take it? [22:26] maybe I should have written: "please give back happy alex on lpia" :P [22:26] infinity: correct... though this upload should vanish soonish, as it will draw in hugs when installing ghc6 [22:28] ghc6: Depends: haskell-utils but it is not installable [22:28] sistpoty: Still waiting on a publisher run, perhaps? [22:28] infinity: grmpf... I thought "ACCEPTED" means that this had already happeneD? [22:28] sistpoty: Nah. [22:29] infinity: ok, then it still needs its time [22:33] infinity: would you mind giving "revelation" a give back on hppa.? [22:34] sebner: Done. [22:34] infinity: thank you :D [22:36] infinity: I am thinking about leaving cairomm alone since it is a sync from debian. If I go ahead and patch cairomm to be completely correct, I'll make more work for us when we want to get back in sync with debian. What do you think? [22:37] protonchris: If it's not currently ubuntuish, don't bother. As I read the bug and the patches, the cairomm patch, while correct, is entirely unnecessary with cairo already fixed. [22:38] infinity: I agree. So does that make the cairomm bug fixed or invalid? Bug 205701 [22:38] Launchpad bug 205701 in cairomm "Latest hardy cairomm is broken" [Undecided,Confirmed] https://launchpad.net/bugs/205701 [22:39] infinity: by the way, thanks for all your help today [22:39] protonchris: It makes the cairomm bug unfixed, still open, but low priority. [22:39] protonchris: It's certainly not "broken" anymore. [22:40] i want to document my package selection and make installations with a defined package-set reproducible ... [22:40] aquo: "dpkg --get-selections" [22:40] infinity: i know that, but i am thinking about creating meta-packages [22:40] aquo: And you can pipe that same output to "dpkg --set-selections" on another system, before running "apt-get dselect-upgrade" [22:41] infinity: --get selections is not modular enough [22:41] aquo: If you want to learn how to create metapackages, this isn't the channel for you. #ubuntu-motu might be a better starting spot. [22:42] ok, thank you [22:44] cjwatson: hey [22:46] cjwatson: was told that you might know why bash_completion isn't installed in hardy as default? [22:47] awen_: cjwatson probably isn't around, there are other people that know though, so you may get an answer if you don't direct your question to one person [22:47] wow.. have not expected pressing the power button to shutdown my system.. (hibernating since last kernel update) :p [22:48] Isn't that expected to present a shutdown dialog or do nothing? (kde4) [22:48] james_w: okay, was just told that he was the best bet... but if anybody else knows :) [22:50] awen_: I knew that this topic was dealt with in the last few days (sorry, don't know exactly when, or what the outcome was)... maybe grepping irclogs might give you some clue? [22:50] awen_: (right here in ubuntu-devel) [22:50] yeah, I think cody asked about it yesterday or the day before [22:51] wouldn't mapping the power button to "hibernate" make far more sense? (if it's possible through pm-utils) [22:52] sistpoty: okay... i'll try looking through them, thanks [22:52] blueyed: in gnome it launches the log out/supsend etc. dialog [22:52] blueyed: at least in Gnome you can set what it should do in the power management preferences [22:53] james_w: Through gnome-power-manager probably then. [22:53] imho in kde as well... *being bold and just trying* [22:53] Great. KDE3 did so, too IIRC. Needs a fix in KDE4 then, like the shutdown in general :p [22:53] wohoo, I'm still there :) [22:53] sistpoty: kde3? [22:54] blueyed: yes [22:54] sistpoty: lucky you :D kde4 even does nothing, from the shutdown menu currently. I've always manually called "sudo pm-hibernate", which works great. [22:55] sistpoty: :-) [22:55] heh [23:00] * awen_ found it, thanks... but thinks might + may is included one too many times in the conclusion [23:08] Any chance to get http://incoming.debian.org/ndisgtk_0.8.3-1.dsc synced? It fixes bug #108656 and the changes are very very small [23:08] Launchpad bug 108656 in ndisgtk "ndisgtk can't remove installed driver" [High,Fix released] https://launchpad.net/bugs/108656 [23:13] Hi. What is the preferred way of including autoreconf in the build process of a package? [23:15] the preferred way is to not have to ;) [23:16] I need to fix a Makefile.am [23:16] but if you're going to: build-depend on the autotools you need; call autoreconf somewhere before calling ./configure; and remove all of the autogenerated files in the clean target, to not cause gratuitous interdiffs. [23:16] TomJaeger: Are you the Tom Jaeger of bug #194214 fame? If so... thanks for finding the solution! [23:16] Launchpad bug 194214 in xorg-server "Keys get "stuck" down" [Unknown,Confirmed] https://launchpad.net/bugs/194214 [23:17] indeed, that'll get included before too long [23:17] yup, that'd be me [23:19] slangasek, interdiffs that remove files are okay? [23:19] because the original tarball already has all those autogenerated files in it. [23:30] TomJaeger: removal of files present in the orig.tar.gz are ignored when creating the .diff.gz [23:30] TomJaeger: so they wouldn't show up in the interdiff at all [23:30] okay, thanks