[00:14] Sorry to ask, but where and to whom do I talk about backports from hardy to gutsy? [00:30] !backports | Fjodor [00:30] Fjodor: If new updated Ubuntu packages are built for an application, then they go into Ubuntu Backports. See https://help.ubuntu.com/community/UbuntuBackports - See also !packaging [00:35] Fujitsu: Thanks. More specifically, though, I would just like to ask the person in charge of that specific package why the compcomm-settings-manager thingy isn't backported when alle the other compiz(fusion) things are... [00:36] Basically, that's why I ask here, as I would deem this a dev (or packager) question, and not just a simple usage one that would be better asked in #ubuntu [00:37] It may just be that nobody thought to backport it. [00:39] Fjodor: anyone here would just have to look up who the backporter in question is, presumably you can find that information even easier than we can since you have the other backports in front of you [00:40] slangasek: Now we are moving! Where do I look up backporters? [00:40] Fujitsu: Indeed, so I just need to get in contact with the guys who are handling these packages (see below) [00:41] Fjodor: in /usr/share/doc/$package/changelog.Debian.gz [00:41] slangasek: Excelent. Thank you! [00:46] i'm trying to update my own apt archive [00:47] how do i make a Release.gpg file? i thought it was just by signing the Release file [00:47] but that gives 'unknown gpgv error' from apt [00:48] poolie: Do you have the key on apt's keyring, and is the error from a chroot (where gnupg might not be installed)? [00:48] yes, and no respectively [00:48] Nothing more useful than `unknown gpgv error'? [00:49] no, it sucks [00:49] well [00:49] W: GPG error: http://bazaar-vcs.org ./ Release: Unknown error executing gpgv [00:50] -o debug::acquire::gpgv=1 [00:50] poolie: ^-- try with that [00:50] thanks [00:51] ok, that shows a command [00:51] * Fujitsu worships all-knowing elmo. [00:51] which when i run it gives, "no valid openpgp data found" [00:51] it is a valid ascii-armoured file though... [00:51] What's the URL to it? [00:52] ah, that's because it gave the exec args with argv0 repeated [00:52] http://bazaar-vcs.org/releases/debs/gutsy [00:52] slangasek: Just a final thank you and update. I just mailed the last person to update the package. Thanks for the pointer [00:52] ok, "not a detached signature" [00:52] poolie: that dosn't look like ... [00:53] right.. not one of those :) [00:53] Fjodor: no problem [00:53] elmo, --detach-sign wants to produce Release.sig - i just rename it to .gpg? [00:54] poolie: yeah [00:54] yay [00:54] thanks [00:55] I think you want a -a in there. [00:55] is it meant to be ascii or not? [00:55] the previous one was not... [00:55] poolie: I suspect either works [00:55] poolie: ascii is more friendly and is what ubuntu/debian uses [00:55] The one I saw first was ASCII, but the latest one is not. [00:55] Ah, that's better. [00:56] i am getting more convinced about using a PPA :) [00:56] PPAs don't sign yet. [00:56] i know [00:56] i mean, just avoiding this in general [00:56] or, avoiding building equivalent automation [02:00] * lamont wonders how well a breezy->gutsy upgrade will work [02:00] it'll be very brutsy [02:00] lol [02:00] * persia suggests hitting Dapper along the way for the new meta-packages bonus [02:00] well, it's a rather slow rsync to get the iso where I can burn it, so I have an hour to kill. what the hell. [02:00] persia: was planning on dapper [02:00] and then maybe I'll leave it at dapper. :) [02:01] And a chance to reinstall if it all goes horribly horribly wrong? [02:01] install testing is all about love [02:01] * lamont just upgraded his laptop to 4GB, and now it runs slower. [02:01] most annoying [02:02] hello lamont [02:02] howdy [02:02] lamont: how much do you like to suffer :) [02:03] zul: netboot hates me right now, and the machine has breezy on it through a strange array of coincidence [02:03] OTOH, 1u ia64 boxen are kinda noisy [02:03] * Fujitsu would advise going to at least Edgy, as there were a lot of migrations. [02:03] esp when you're sitting 2 feet in front of it [02:04] keeps the winter chill away though :) [02:04] heh [02:04] Fujitsu: Aren't most the migration hints still there for Gutsy? [02:04] dapper is fine for doing chrootish things... and besides, it's not for me. it's for christmas [02:04] the evms-must-die one isn't [02:04] I remember the Itanium at LCA '02 (I think it was) -- my bag was 3 feet away from it, and it was hot [02:04] persia: They're meant to be, but I've tried Dapper->Gutsy, and it... didn't work well. [02:05] I suspect that Edgy->Gutsy would work much better. [02:05] when I get home, I get to recover a box that I just dist-upgraded from feisty to gutsy, which doesn't boot anymore because it has root on /dev/md0 [02:05] * ScottK tried Dapper -> Gutsy as well. [02:06] * ScottK got there, eventually, but with no TTYs. [02:06] ScottK: yeah.. that's a known bug. [02:07] Yeah. I know, but it doesn't make it more pleasant. [02:07] 'struth [02:07] * ScottK looks around for someone who will actually fix it. [02:07] I thought it had been [02:07] dunno. [02:07] pester keybuk and/or mvo [02:08] * ScottK doesn't have any boxen needing Dapper -> Hardy for real, so can't be bothered. [02:08] amazing how our beloved child becomes a step child to hide under the table [02:08] * Hobbsee should upgrade her edgy machine. [02:08] Well I've got Christmas card lables to print. [02:09] yeah! netboot loves me, it's the console that didnt [02:09] My server here will be jumping from Dapper to Hardy [02:09] * lamont learned that nssldap doesn't care if the server cert is expired. [02:09] whereas ldapsearch does [02:09] Two servers, even [02:10] kewl. lan A was actually eth0. :-) === bigon is now known as bigon` [02:11] lamont: Is that surprising? [02:11] I've found that it depends... [02:11] sigh. [02:11] yes [02:11] I did find the right place to unplug the cable so that it would ask for a proxy. go me. [02:11] Hah. [02:12] stupid so-called "transparent" prxoies [02:12] and the box gets to have gutsy on it. [02:12] * Hobbsee notes that one eventually gets used to the compiz-slow-refresh [02:12] * Hobbsee wonders why compiz has only crashed once today [02:13] Hobbsee: it hasn't crashed for me all week. :-) [02:13] lamont: is it purged? [02:13] yeop [02:14] odd, that [02:14] or (in the other case), no 3d acceleartion [02:14] * Fujitsu finds that purging fixes a lot of things, including trackerd eating the CPU. [02:14] Fujitsu: don't forget network-mangler [02:14] * Fujitsu likes network-mangler. [02:14] Is it related to queue-mangler? [02:15] if trackerd would eat network-mangler, that'd be just fine [02:15] no [02:15] Fujitsu: purging network-manager should not down the interfaces on my box. [02:15] that's rude [02:15] I thought the alternatives to network-mangler were being removed, making it harder to purge... [02:15] lamont: Restarting it shouldn't either, but that's what it does. [02:15] persia: "ifupdown"? [02:15] slangasek: Yes (or at least my configuration is becoming unhappy over time) [02:15] * Fujitsu finds that quite adequate unless versatile wireless is wanted. [02:15] * lamont uses ifup and dhclient [02:20] persia: hmm, works for me so far [02:20] * persia hopes it continues to do so [02:21] * lamont has been using ifupdown forever [02:21] well, a few years anywway [02:21] * Fujitsu thinks we might have some slight complaints if it stops. [02:22] heh [02:22] and how. [02:23] Fujitsu: It depends. If there was a perfect alternate solution that didn't break things when it was enabled, it would likely be acceptable to many. [02:25] persia: that rules out network-manager [02:25] at least until it gets more use-models in it than "there is one network" [02:25] I believe there is some sanity on the 0.7 roadmap. [02:26] s/one network/one network with guaranteed uptime of the primary router and any DHCP server/ [02:26] persia: yes, one that doesn't require me to log in before it brings the network up [02:26] Multiple active interfaces and system-wide config are on said roadmap, I believe. [02:27] slangasek: That's not so bad for the notebook use-case, but it's frustrating to have the network stop working because of dropped packets to the router. [02:27] persia: it's bad for the notebook with kerberos use case [02:28] slangasek: Right. s/notebook use-case/exceedingly simplified single-user single-person self-managed environment in perfect surroundings/ [02:29] right. :) [02:30] persia++ [02:30] slangasek: Could I convince you to release texlive-bin from binary NEW? [02:31] Oh, btw., does this channel still concern itself with matters of gutsy (specifically the logout dialogue in GNOME)? [02:34] * Fujitsu doesn't see how that is specific to Gutsy. [02:35] Fjodor: Yes, but not in the sense that things should be complained about. There's a spec for a new solution planned for hardy... [02:35] StevenK: what does it break this time? [02:36] slangasek: gnome? :-) [02:37] $world [02:37] slangasek: It shouldn't break anything [02:37] Fjodor: https://wiki.ubuntu.com/DesktopTeam/Specs/ExitStrategy [02:37] slangasek: btw, congrats on becoming a real person today [02:37] slangasek: It's a direct sync, so it's Debians fault [02:37] StevenK: it's tex, of course it breaks something [02:37] Bwahaha [02:37] * lamont hates 9600 baud serial consoles [02:38] persia: Hmmm, so if pressing the logout button results in an inaccessible system for several minutes before the logout dialogue appearing (sometimes without options for sleep and hibernation), that is hardy(+) material? I will be reading the link in 2 minutes... [02:39] * lamont wonders if a 2GB EFI partition is a bit excessive, for a kernel develoer [02:39] Fjodor: Not necessarily, but for a blocking problem or regression, filing a bug is a better solution. [02:40] persia: Ok, will do then. I just hoped to get someone involved on the line, as my gf is bugging me about it all the time ;-) [02:40] persia: Thanks [02:40] Fjodor: those sorts of support issues plague all of us. [02:40] and after you upgrade from gf to wife, it just gets more so... [02:41] Fjodor: It's more that the solution is likely complex, and so a quick fix is definitely wrong. You might be able to get hints on a workaround in #ubuntu [02:41] lamont: One should use GF LTS to avoid upgrade issues? [02:42] Fujitsu: once you install wife, downgrading is inadvisable [02:42] lamont: Indeed, but however much empathy you get, the hard fist of wife-alpha1 still perseveres ;-) [02:42] persia: Ok, thanks, but I have tried there for some time now... [02:42] Fjodor: In that case, there probably isn't a workaround :( [02:43] sometimes, signal-to-noise suffers on #ubuntu. not always though. [02:43] and yeah, that points to a lack of a workaround, or that the people who have the workaround are in the wrong timezone relative to you [02:44] dear partitioner. when I create a md device with n partitions, and there are n drives with one partition each available, please preseed the answer for me. [02:44] kthxbye [02:45] persia: Well, I have done some preliminary diagnostics. That's why I went here afterall. I have had complaints about it on my gf's laptop, I see it on my own laptop, I see it on my desktop, and I see it on my office desktop. That's my reason for deeming it reproducible... [02:45] * slangasek looks at lamont blankly [02:45] * Fujitsu hasn't seen that issue since early in Gutsy's dev cycle. [02:45] lamont: But yes, you are right [02:45] Fjodor: Right. That would be a bug then, so the people who work on that package will be notified, and can discuss alternatives for a solution. [02:45] n partitions, or n components? [02:46] * Fujitsu hopes the latter. [02:46] * lamont sees slangasek's blank look, and raises him an eyebrow. [02:46] create an md device gives me a list of partitions to put in the array. there are certain cases where the answer is obvious and forced. [02:47] right, so "an md device with n components" [02:47] since one rarely partitions an md device [02:47] * lamont currently sees long delays at login because of restricted-manager thinking about life. [02:48] slangasek: partitions on other devices... the list is of partitions to use to make the MD device [02:48] Just for speeding things up, where do I report the bug? If noone has the direct URL, I'll find it myself, but you devs seem to be able to find such things at a wink of the eye [02:48] given that partition 1 on the drive needs to be (in this case) efi, the raid device lives on partition 2 [02:49] hrm... I guess this machine does get a desktop install [02:49] * lamont forgot the magic incantation to tell the alternate CD to install just server-bits. [02:49] and doesn't want to go back in any case [02:49] lamont: The incantation is a down arrow at gfxboot, I think. [02:49] Fujitsu: no. [02:50] that's the right answer after the iso generation code is fixed [02:51] * lamont grumbles at the installer for not allowing a blank password. [02:52] Fjodor: if we assume it's in the gnome-session source package (which is a leap), then https://bugs.launchpad.net/ubuntu/+source/gnome-session/ [02:53] and if it's not gnome-session, one of the gnome-heads will eventually move it to the right package, I expect... [02:53] but that's just a shoot-from-the-hip answer. [02:53] 1 → 50 of 506 results [02:53] yeah! [02:54] * lamont wonders if the queue will break 500 before he heads out in a bit [02:54] lamont: Ok, thanks. I'll load that url, and if it doesn't fit, I'll roll from there. Launchpad can just be somewhat confusing from time to time... [02:54] s/somewhat/very/ [02:55] s/from time to time// [02:55] you are in a twisty maze of pages, all different [02:56] You would be likely to be eaten by a grue, except that the green bar at the top is too bright. [02:56] Fujitsu: clearly you haven't played the original much. - [02:56] :-) [02:57] advent has no grues [02:57] That would be a correct assumption. [02:57] lamont: Thanks again. https://bugs.launchpad.net/ubuntu/+source/gnome-session/+bug/123078 seems to very acurately describe what I am seeing [02:57] Launchpad bug 123078 in gnome-session "System -> Quit takes a long time to appear" [Low,Confirmed] [02:57] * lamont beams with glee [02:58] I got one right!! [02:58] (the title doesn't cover it all, but the problems are all described there)... [03:02] dear installer. please consider saying something cuter than "Please wait...". maybe something like "please don't kill me, sir...". kthxbye [03:02] hrm... gender biased. [03:02] never mid [03:02] mind [03:03] lamont: i shoved kde4 down again on hppa [03:03] Hobbsee: heh. tahnks [03:03] although, apparently there's a neat LP bug that will rescore it to where LP thinks it belongs [03:03] for the win [03:04] hehe [03:04] I don't mind building kde. I mind building out-of-date kde [03:04] so if they'd just get it right before they uploaded, then I wouldn't have to think of ways to hurt them. [03:05] "Installation complete" [03:05] lamont: not really. i'd just like working wifi, thanks! [03:05] LongPointyStick: heh [03:05] Hobbsee: so why you running two clients? [03:05] lamont: one to log. [03:05] this is a laptop. i turn it off [03:06] --- [LongPointyStick] (n=mystery@ubuntu/member/hobbsee) : " " [03:06] such a mystery... [03:06] lamont: yeah, i know. it's used for other networks, when i mess with people's minds [03:06] hehe [03:06] It might take you ... seconds to solve [03:07] individuals of clue will note the similarities [03:09] kewl. console install of gutsy: /etc/event.d/ttyS* [03:09] that's with server install. [03:09] kewl. default was the right thing [03:09] (netboot) [03:09] Hobbsee: You are concerned with user experience, right (if I remember correctly)... [03:09] lamont: on non-cloaked networks it's useful. [03:10] Fjodor: that's oen of the things i'm concerned about, yes [03:11] Hobbsee: Thought so. Did you see my last two messages before I addressed you directly? [03:11] lamont: That build-score-clobbering bug should be fixed with 1.1.12. [03:11] Fjodor: about https://bugs.launchpad.net/ubuntu/+source/gnome-session/+bug/123078 ? [03:11] Launchpad bug 123078 in gnome-session "System -> Quit takes a long time to appear" [Low,Confirmed] [03:11] ...and I apologize again for singling out individuals to complain to, but my gf is really bugging me about this one... [03:12] Hobbsee: Yes, indeed [03:14] For the longest time, I just assumed that it had hung, and ctrl-alt-backspaced it all... [03:15] ...and try to explain that to a non-literate in the arts of linux... [03:16] so... when I remove the password from /etc/shadow and make it say :: there... why can't I log in anymore? [03:17] Fjodor: "It's like ctl-alt-del, only with the backspace key instead'. :-) [03:17] lamont: That build-score-clobbering bug should be fixed with 1.1.12. [03:17] Oops. [03:17] Wrong button. [03:18] lamont: indeed, but she just wants a machine that *works* [03:19] yeah. my wife occasionally asks to install windoze for that reason, and I look at her strangely until she hits me. [03:19] all in all, I'm not sure that counts as a "win' [03:20] hrm... init=/bin/sh seems to have not quite done what I wanted. [03:20] * lamont declares victory and heads home [03:20] Haha [03:20] It didn't do what you want, so you're victorious. I see. [03:21] Fujitsu: no. I declared victory [03:21] Well, per request I gave her OSX. It took 10 minutes till she cried "Give me back my Ubuntu" :-) [03:21] which is to say, I redefined the problem. [03:21] And that counts as a win for all you guys :-) [03:21] Fjodor: those are nice wins. [03:22] and time for me to run away. [03:22] weird [03:22] back in a few hours [03:22] kdmthemes overrides are broken [03:23] and irc doesn't make a good todo list. [03:23] whatever [03:23] just letting you guys know [03:23] Jay-Oh-En: Yes, but there's no guarantee anyone who cares is here now. File a bug. [03:24] persia: they wouldn't be on this channel anyway [03:24] holy cow [03:24] aptitude segfaulted [03:24] why wouldnt they [03:24] persia: well ill files a bug [03:24] Jay-Oh-En: Thanks :) [03:24] Hobbsee: quit being so uptight [03:25] because they're likely on #Kubuntu-devel, and a lot of people aren't awake at this time of day anyway [03:25] heh [03:25] if he thought *that* was uptight... [03:25] * Fjodor tries in vain to step in for Hobbsee an say that she isn't uptight... [03:25] if he files a bug with the same content as his irc message, i'm going to have savage pleasure setting it to NEEDSINFO, and waiting for LP to do it's job. [03:26] Janitor reappears this release, sort of. [03:26] i thought it was always on now [03:26] Does that mean another few hundred bugs to undo? [03:27] When I say `sort of' I mean `with a per-project off button, and only commenting, not closing, so we can see what it will do when it is turned on properly' [03:27] oh *classy* [03:27] Oh, btw., have I told you about this strange bug that I seem to be the only one to have on my office machine? [03:27] aptitude segfaults if you try to get a package on a changelog that doesn't exist [03:27] Does it? Nice of it. [03:27] yup [03:27] Hobbsee: That code is *horrid* [03:27] william@irranat:~$ aptitude changelog boom [03:27] E: Couldn't find a changelog for boom [03:28] * TheMuso is reminded of the aptitude fisco on Planet Debian recently. [03:28] Fujitsu: which version of aptitude? [03:28] Killing the apt transition, TheMuso? [03:28] Installed: 0.4.9-2ubuntu3 [03:29] Fujitsu: Yeah. [03:29] Fujitsu: strange. [03:30] What could cause gnome-update-manager to confuse it's buttons? I run it, but when I press "Install Updates", it acts as if I pressed "Check". And to make it stranger still, I can only reproduce it on my office machine... [03:32] something crackful... [03:34] Hobbsee: Indeed, but this machine is the one that was installed from scratch most recently... [03:34] ah ha. reprioritizing some of the stuff in gnome makes it start much quicker. [03:36] * TheMuso is shocked. Hobbsee still using GNOME? :) [03:37] Fjodor: the defaults look borked. [03:37] TheMuso: yeah, it appears so. [03:37] TheMuso: there's some really annoying stuff about it though [03:37] Hobbsee: Pointers? [03:37] TheMuso: kde sessions win over gnome's. although this priority stuff is nice [03:38] Fjodor: give the crucial stuff a priority of 40, then hit apply to each. [03:38] like, gnome-session-properties, gnome-volume stuff, etc [03:38] That's only because the KDE session information contains stuff as the last time you scratched yourself, where and for how long. [03:38] TheMuso: /me is using GNOME. I left the path of LFS/BLFS and WindowMaker some time ago, where you making a point? [03:39] Fjodor: Hobbsee is a KDE advocate. [03:39] * Hobbsee likes that you can you can start with an empty profile, start the stuff you want, and then hit "save current session" - which will restore things *exactly* how you've left them [03:40] which i haven't found out how to, under gnome. devilspie works somewhat, but dosen't do to tray [03:40] and doesn't appear to work if you have 2x2 windows [03:40] * Fjodor does not expect things to be as when he left them. Because then there must have been a power outage... [03:40] well, no, but i have a list of stuff that i want to start, every time. [03:41] (nm, gnome-do, pidgin, etc) [03:41] gnome-do? [03:41] TheMuso: Quiksilver for GNOME [03:41] see lp.net/gc [03:41] Hobbsee: Indeed, but does devilspie also place them in the correct workspaces? [03:41] StevenK: WHich is? [03:41] ...because then I might take alook [03:42] Fjodor: usually. when you run a 1 row thing, yes. when you run 2 rows, not always. [03:42] Hobbsee: Ok [03:42] devilspie is *nice* [03:42] TheMuso: Katapult is the KDE one [03:42] StevenK: What is quicksilver? [03:43] what [03:43] TheMuso: It's a quick launcher, hit alt-space, the first few letters, it'll try and match and hit enter to launch [03:43] I can't make fun of Hobbsee anymore for being a KDEhead? [03:43] StevenK: ah [03:43] bluefoxicy: i'm sure you can find something to make fun of, w.r.t. me. and i'm sure that i can ban you in return. [03:44] Well i could make fun of you for being a girl, but that's kind of childish in comparison to making fun of you for using KDE [03:44] Hey Hobbsee? Make me a bouncer, then I'll relay all messages and deride the senders in return :-) [03:44] * bluefoxicy gives Fjodor shades and a big stick. [03:44] You are now a bouncer. [03:44] * Hobbsee wonders why the words "castration" and "bluefoxicy" are coming to mind... [03:45] Badabum [03:45] And bluefoxicy: thanks :-) [03:46] Hobbsee: Wouldn't make much of a difference [03:48] How do I reset all of my gconf menu options to the factory-default, I remember there being an undocumented option something like gconf-editor --reset-defaults /location/in/gconf and I just need some help, thanks. [03:49] --recursive-delete I think [03:50] Hmmm, I was given shades and a big stick. Anyone in the mood for a joke? [03:51] (since nothing much else happens) [03:51] think not. [03:51] hodgson: Ok, sorry [03:58] damn [03:58] i can make an app dock with devilspie, but then it keeps docking === j_ack_ is now known as j_ack [04:22] hello everyone [04:22] wow..not really busy in here [04:22] is there anyone here? lol [04:23] oh, goody. --safe-upgrade for aptitude now behaves sanely, w.r.t new packages [04:23] at least, once mvo merges it across [04:25] Quiane: 22 seconds isn't a very large sample size. [04:27] * lamont does his happy dance. [04:28] 496 packages in needs-build hardy/hppa [04:28] lamont: That was a short departure. [04:28] * Fujitsu checks the number of failures. [04:28] 'twas the drive home [04:28] Not too bad, actually. [04:28] * Fujitsu is surprised at how quickly it is chewing through them, and the success rate. [04:28] 22 miles [04:29] gutsy was at >97% in gutsy-stage0 [04:29] * Fujitsu charges lamont with use of imperial units. [04:29] * lamont notes that he is imperial, after all. [04:29] * Fujitsu bows. [04:30] great. LinuxOLD boots, Linux doesn't. that's gonna take a little love [04:30] * lamont afk for a bit === asac_ is now known as asac [04:41] i'm here actually! [04:41] i was just talking in the other room [04:42] Ok, so I've got a decently well-directed question: Where are the terminal preferences stored? As in the right-click ->options->preferences on the panel [04:43] is it in gconfeditor, a conf file, or some other random esoteric gnome location that is plesently undocumented [04:45] hodgson: Which terminal are you using? Gnome-terminal? If gnome-terminal, then gconf. [04:47] oh shoot [04:47] not terminal preferences [04:47] panel preferences* [04:48] specifically the one top one, I think it is called the ''application-launcher'', has applications/places/systems by default [04:48] Oh then I'm not really sure. [04:51] anyone using raid-for-root [04:51] ? [04:52] fucking gnome [04:52] * lamont tries to remember the incantation to get initramfs to stop during each step [04:52] I swear to god it gets more shitty with each growing interation. [04:52] Ok get this -- ~/.config/menus/applications-menu [04:53] how the hell. [04:53] I swear last week it was it was in gconf-editor, and you could reset it with --recursive-unset, before that I remember it being in ~/.gnome, or the like. [04:54] !ohmy [04:54] Please watch your language and topic to help keep this channel family friendly. [04:54] where the hell did ~/.config come from === ScottK2 is now known as ScottK === cassidy_ is now known as cassidy [05:56] ah. mischief managed. [05:56] adding md-mod and raid456 to /etc/modules makes life much happier. [05:57] (root being raid5, it helps to have the modules needed to support that in the initrd) [05:57] something in initramfs should notice that [05:58] * persia thinks /etc/modules is currently entirely manually managed, and encourages a migration to /etc/modules.d/ [06:11] Hi all! [07:15] Good morning [07:15] Hey pitti. [07:16] pitti: So, why do libgsl0 and libglib1.2 show as FTBFS in the NBS list? They don't. [07:16] They changed library names and grew a suffix [07:18] StevenK: hm, good question [07:21] * pitti accepts l-b-m and linux-meta [07:28] StevenK: fixed, thanks; running again now [07:28] good morning [07:28] Hey dholbach. [07:29] heya Fujitsu [07:32] pitti: Thanks [07:35] z,cb 5 [07:35] Oops. [07:52] good morning [08:02] infinity: can you please commit your pkg-create-dbgsym changes to bzr? [08:33] hey seb128, hey Toadstool! [08:33] hey glatzor, jsgotangco! [08:34] hello dholbach [08:34] yo! [08:34] :-) [08:34] * jsgotangco currently at a domestic airport going back home [08:34] servus dholbach! [08:34] * dholbach hugs y'all [08:34] ho ho ho! [08:35] Hi dholbach et al. [08:35] hiya ion_ :) [08:35] I ordered a Waldorf Micro-Q synthesizer yesterday. ♥ [08:36] NICE [08:36] let me know when there are tracks I can download :) [08:37] I haven’t got pretty much any music at all made for a long time with this health, though. [08:40] hey blueyed_ [08:40] ion_: I'm sure you'll come up with something soon :-) === blueyed_ is now known as blueyed [08:45] slangasek: did you notice debian bug 456915? [08:45] Debian bug 456915 in boost "boost: FTBFS: build.sh: line 15: gcc-4.1: command not found" [Serious,Open] http://bugs.debian.org/456915 [08:47] tjaalton: heh, thanks for the heads-up [08:50] dholbach: I hope so. :-) [08:51] slangasek: np. What about flute-1.3-jfree that the new openoffice needs, will it be promoted to main? [08:54] tjaalton: not my department :) we're trimming OOo-base from the seed for alpha2 instead in the meantime [08:55] slangasek: ah, ok :) [08:56] * Fujitsu wonders if anybody uses OOo Base. [08:57] I've used it in the past [08:57] only for ODBC :) [08:57] hooray, the livecd generates a valid xorg.conf again :) [08:58] bryce: ^^ [08:58] tjaalton: Is input hotplug likely to deal with Synaptics touchpads sanely in the near future? [08:58] Fujitsu: yes [08:59] Fujitsu: synaptics was dropped from dexconf but sadly i-h is not there yet [08:59] tjaalton: excellent [09:01] hmm wait, need to double-check what actually happened [09:01] tjaalton: I'm planning on doing work on synaptics over Christmas [09:01] Upstream seem to have given up [09:02] mjg59: really? that's too bad, but nice if someone still cares :) [09:02] my X61 doesn't have a touchpad, and neither did the T23 [09:03] I've a couple of patches that never got merged [09:04] And the input hotplug stuff needs doing [09:04] As does sorting out a standardised config mechanism [09:04] hmm, seems that the livecd from this morning did all sort of crappy things, like ran xresprobe etc. :) [09:05] mjg59: I also get the feeling they're not paying sufficient attention to it [09:05] bryce: Well, any. Last update was months ago. [09:07] when do we have uptodate livecd's to test? [09:07] alpha-2 is scheduled for towards the end of the week [09:07] I know :) but surely there are prereleases before that [09:08] don't think so - just the usual dailies [09:08] but the one from this morning is actually old, uses the old kernel and xorg etc [09:09] maybe because current ones don't build? [09:10] tjaalton: that would be the usual reason for CDs to be old; check the build logs? [09:11] http://people.ubuntu.com/~ubuntu-archive/livefs-build-logs/ [09:11] The following packages have unmet dependencies: [09:11] libhsqldb-java: Conflicts: openoffice.org-base (< 1:2.3.1~m8) but 1:2.3.0-1ubuntu5 is to be installed [09:11] cjwatson: I guess they fail for the same reason my own build do, OOo-base is the remaining issue [09:11] touché :) [09:12] tjaalton: xubuntu, kubuntu should have working liveCD builds by this point; ubuntu, edubuntu are waiting for that OOo seed change [09:12] slangasek: ah, good point, I'll check them out === sjoerd_ is now known as sjoerd [10:04] hmm, kubuntu daily-live ends up in initramfs console, and xubuntu isn't published yet [10:08] ah, xubuntu daily-live is now available :) [10:20] ..but it fails just like kubuntu [10:26] hmm [10:35] why is gftp in main? [10:38] /j #ubuntu-x [10:38] oops :) [10:39] m/sg superm1 good luck at the parole hearing then, I guess! [10:39] oops :) [10:39] haha [10:39] /msg spads I love you too, honey. [10:39] Whoops [10:39] hi pitti [10:40] dholbach: hysterical raisins (since warty); is there a better replacement now? or is it abandoned upstream? [10:40] hey tkamppeter, how are you? [10:40] fine [10:41] pitti: it hasn't seen updates in years [10:41] indeed [10:41] dholbach: ideally, people would just use nautilus with ftp:// ? [10:41] yeah [10:41] which actually works just fine [10:41] and I doubt there's many people who use it [10:41] pitti, are you working on updating CUPS to 1.3.5? Then you could also include the patch of bug 177075, the reporter has even included a debdiff. [10:41] Launchpad bug 177075 in cupsys "CUPS 1.3.x lists network interfaces only at startup (regression)" [Undecided,Confirmed] https://launchpad.net/bugs/177075 [10:42] tkamppeter: Kenshi updated svn trunk to 1.3.5, I'll merge it soon [10:43] tkamppeter: right, that should just be committed to the debian svn before we merge [10:44] tkamppeter: thanks, I'll do that soon [10:44] tkamppeter: or, you can do it yourself if you want === bigon` is now known as bigon === pedro is now known as pedro_ [11:02] pitti, I have added the patch to the trunk now. So if you merge we will have it. [11:02] tkamppeter: thanks [11:10] bryce: re http://www.bryceharrington.org/drupal/node/18 feature requests> why not use the blueprints tracker? [11:11] sabdfl suggested that too [11:11] pitti: I'm just about to upload 46 source packages for fftw3-dev -> libfftw3-dev [11:11] tjaalton: thanks, I'm updating the seeds and meta packagees now [11:14] pitti: Those 46 are all the Build-Depends, so I was going to clear them out of the way and see what's left [11:15] StevenK: great [11:19] * StevenK throws 46 uploads at Soyuz. \o/ [11:21] pitti: Oh, thanks for the tasks promotion [11:22] no problem [11:29] pitti: There's all 46 uploaded, let's see what falls out [11:35] pedro_: ah, thanks for forwarding my tracker bug; I didn't do it, because so far jamiemcc read the LP bugs, too :) [11:36] StevenK: oh, that wasn't just a no-change rebild? [11:36] StevenK: in that case, wouldn't it have been easier to add a Provides: to libfftw-dev? [11:36] hello all, especially pitti [11:37] hey mhb, how are you? nice to 'see' you again [11:37] pitti: quite busy with school, exams are coming. [11:37] oh, good luck! [11:38] I'll find some time for the restricted-manager at the weekend (friday-sunday) [11:38] pitti: Unsure, I did what I thought was easiest [11:38] mhb: ah, great (but don't let that make you fail the exams :) ) [11:40] pitti: I'll try not to... ubuntu is more fun, though :o) [11:41] * pochu agrees with mhb [11:41] mhb: heh; please let me know if anything in the current design is unclear; let's keep it clean and swift this time :) [11:42] pitti: will do, thanks! === dholbach_ is now known as dholbach [11:57] pitti: ah sure, no problem :-) [11:58] pitti: i must sorry, i had no time for usplash, still for some days. [11:59] gaspa: no problem :) [12:00] pitti: for me it's a problem... I wished to see my stuff in usplash... :D [12:05] pitti: Can (or, rather, will) do. [12:27] * Hobbsee waves [12:27] hi Hobbsee [12:27] * Hobbsee stomps on Mithrandir's feet, as he forgot to levitate. [12:28] I'm sitting in a chair, with my feet crosslegged, so you failed. [12:28] awww [12:28] * Hobbsee coats Mithrandir in ice, then [12:28] mmm, ice. [12:29] * Mithrandir whips the ice so it becomes ice cream and eats it. [12:29] heh [12:33] think Mithrandir stole all the Ice, it was -10°C in Oslo when I checked and it's only ~0°C in Helsinki [12:33] sladen: it's +6 in Umeå [12:36] the tollef that stole christmas weather? [12:36] * Hobbsee steals christmas, then. [12:36] nah, I'm innocent. [12:36] it's just cold here, no snow. [12:37] Hobbsee: you can have it, what's the point without proper cold and snow? [12:37] Mithrandir: you're *never* innocent. [12:37] maswan: an australian christmas has point. [12:37] maswan: it's properly cold here all right, but it fails at the "snow" part. [12:37] Hobbsee: Shorts and beachparty? [12:37] Mithrandir: the little snow we have is failing rapidly [12:38] torkel: haven't been to Umeå yet, there's a ferry to try out... must swing by there some time [12:38] Hmm.. So where does one have to move to get real winters these days? And is there any intersection of this area with the areas you have real bandwidth? :) [12:39] maswan: beaches are busy, yeah [12:39] I'm sure there's snow somewhere in Australia (an inhabited territory of Antartica perhaps?) [12:39] ther'es snow in au [12:39] just, south [12:39] There's Perisher and Therdbo [12:39] Clarification: snow that doesn't become slush or melts between early december and early march [12:40] StevenK: There's still snow there during Dec? [12:41] maswan: middle of Norway seems to have snow. [12:41] (or the equivalent months for south of equator) [12:41] about 65cm in the wild. [12:42] Mithrandir: if you go a bit inland from the coast here you'll find snow too. there just aren't much other things than snow and trees there. ;) [12:42] realist: Not sure about that [12:42] maswan: what's wrong with only trees and snow? [12:42] maswan: depending on what you call real bandwidth, you can get at least decent DSL in the middle of Norway. [12:42] torkel: lack of bandwidth and possibly workplace. :) [12:43] StevenK: there's always ice in Antartica :-) [12:44] * Hobbsee hibernates in antarctica. [12:44] Good luck with getting bandwidth there though. [12:44] green aliens find it quite warm. [12:45] maswan: a shovel and an axe will keep you busy, so you don't need any bandwidth :-) [12:45] especially in -25°C, you'll be busy all winter. [12:47] I'm not sure it passes the "no slush" requirement though. :/ [12:47] But I guess I'll just have to investigate some more. :) [12:47] (or if it will in 10 or 20 years for that matter, it's changing quite rapidly) [12:55] good morning, all. i've filed bugs against unintallable binaries in xubuntu (#'s 177444, 177446, & 177448), and have pinged cody somerville and lionel le folgoc - is there anyone else i should ping about this? [12:58] j1mc: they should deal with it, and poke others if they need them. [12:58] j1mc: of course, if it makes the others uninstallable, then other people will be looking into it [12:59] Hobbsee: thanks [12:59] and thanks for your reply to my email - I wasn't sure who to contact. [12:59] j1mc: but they should be taking reponsibility for their own version, particularly somerville32, as he likes to control it a lot. [13:00] (the discussions in here have suggested that, anyway) [13:00] ok, I'll keep that in mind. at least the bug reports are out there now, and they are each aware of the issues. [13:00] did you subscribe them to the bug reports? [13:01] no, i'll go ahead and do that. [13:04] * Hobbsee wonders why raptor is in main [13:04] thanks, Hobbsee ... have a good day! [13:04] oh, it's not what iw as thinking of === tkamppeter_ is now known as tkamppeter [13:10] * lamont yawns, waves [13:12] heya lamont [13:12] Hobbsee: http://bld-4.mmjgroup.com/~wb/buildLogs/stats/hardy2-short.png [13:12] note where kde et al were uploaded. :-) [13:12] lamont: nice work! [13:12] it's nice that hppa only flattened last night, rather than down-dipping like the rest of the world [13:12] heh [13:13] * StevenK has been pondering reinstalling his hppa as Dapper [13:13] Then again, it's only a 712 [13:13] http://bld-4.mmjgroup.com/~wb/buildLogs/stats/hardy2.png looks even cuter. [13:14] StevenK: don't do it. gutsy is the hppa-loving answer [13:14] But, but, it's a server! [13:14] (as in, the hppa buildds in the data center are running a dapper userspace with a gutsy kernel because that's the kernel they need...) [13:15] and that doesn't work well... had to hack initramfs scripts to make that kernel work with that udev et al [13:15] very not pretty [13:15] full-bore gutsy is much happier [13:15] I didn't think we released Gutsy on hppa [13:16] lamont: It's only a 712, so maybe it isn't worth the bother [13:17] well, running a desktop on it would be kinda silly [13:17] server bits aren't bad though [13:17] * lamont briefly considers considering the advantages of breast enlargement, hits 'D' [13:18] heh [13:19] lamont: it could be a talking point. *shrug* [13:19] Hobbsee: I could discuss the pros and cons... then again, this is a family channel. :) [13:19] anyway, afk for a few minutes [13:19] heh === norsetto is now known as norsetto_limbo [13:36] libcurl4-openssl-dev: Depends: libssh2-1-dev but it is not installable [13:36] Bah === \sh_away is now known as \sh [14:03] StevenK: I thought we dropped that? or just the build dep? [14:03] pitti: That's while trying to install Build-Depends [14:04] The package is in main, and libssh isn't, I'm guessing [14:07] StevenK: we dropped the build dep because we don't want it; apparently the binary dep was just forgotting [14:07] s/ing$/en/ [14:07] pitti: Would you like me to fix it, or will you? [14:08] StevenK: I have a call now, can do later if you want to go to bed :) [14:09] pitti: Okay, thanks. If you're able when it gets published, give-back libofa [14:09] sure [14:09] Ta :-) [14:09] * StevenK runs to bed [14:27] pitti: Does all of NBS get regenerated each time you run it? I don't understand the build timing of a pair of packages (libparmetis depending on libmpich) and don't know whether to upload a build1 to force things. [14:30] persia: you don't need an upload to get a build retry [14:30] seb128: So I shouldn't be uploading build1 packages? [14:31] persia: what do you need the build1 package for? [14:31] if that's something which already built and need a rebuild you need an upload [14:31] if that's something which didn't build yet no need to do a new upload [14:31] seb128: In the case where the current binaries in the archive depend on a library that is no longer built by the source that built it previously. [14:32] you need a source upload in this case, ubuntu doesn't do bin-nmu uploads [14:32] seb128: In this case, it built. Specifically, I uploaded a parmetis 3.1-4build1 for the libmpich transition last month, and then there was a sync, which rebuilt, and yet the NBS list still shows the package for i386 & amd64, which confuses me. [14:33] ah [14:33] persia: let me look [14:34] seb128: Thanks. [14:37] persia: NBS? [14:37] persia: there is a depends on "libopenmpi1 | libmpich1.0c2", maybe the | case is not handled correctly by the nbs code or something like that, better to ask to pitti about this one [14:38] lamont: http://people.ubuntu.com/~ubuntu-archive/NBS/ [14:38] lamont: packages not built from source but which still are used by something [14:38] ah, ok [14:38] lamont: or that should be cleaned ;-) [14:38] lamont: An excellent target if you're bored and feel like a few uploads :) [14:39] heh. /me is kinda hip deep in getting ready to take 2 weeks off work [14:40] seb128: Thanks for investigating that. [14:42] persia, seb128: re [14:43] persia: NBS is regenerated automatically three times a day [14:43] pitti: Thanks. When chasing those should foo | bar type dependencies be manually updated as well, or are they safe to leave? If left, how does that impact the script? [14:44] seb128, persia: right, it doesn't play well with |'ed deps [14:44] persia: don't worry about the script; if the actual dependencies are correct now, we can just drop it [14:44] s/it$/the NBS lib/ [14:45] pitti: sounds like a bug to fix, eh? [14:45] pitti: OK. I'm not sure about all the dependent packages yet. Is there a good way to report that a specific package is a false positive, or is that easy for you to detect? [14:45] * lamont hates it when his code doesn't do what he meant. :-) [14:45] lamont: one of these things which are quite hard to get right with just some dumb lines of shell :) [14:45] persia: to be concrete, which NBS lib are we talking about? [14:46] persia: usually I check it, see that it's fine, and remove the old package anyway [14:46] pitti: Specifically parmetis depending on libmpich1.0c2 [14:46] OK. I won't worry about them if I know they are fine then. Thanks. [14:47] lamont: a sensible implementation would involve DB queries, not repeatedly grep-dctrl'ing the entire set of package indices :) [14:47] hi, does anybody know how I can make sure that a restricted module is not loaded without using the /etc/default/linux-restricted-modules-common. The idea is that of modifying either the /sbin/lrm-video or the lrm-manager. [14:47] blacklist it? [14:47] persia: http://people.ubuntu.com/~ubuntu-archive/NBS/libmpich1.0c2 ? [14:47] echo blacklist module >> /etc/modprobe.d/my-blacklist [14:47] Keybuk: without blacklisting it [14:48] what's wrong with blacklisting it? [14:48] pitti: Yes, although I've not yet looked at illuminator (next on my list) [14:48] pitti: so read the packages/sources into a linked list of structs once. :-) or query the db.. :-) [14:48] something like: if this file exist, don't load this module [14:48] blacklisting is *how* you prevent modules from being automatically loaded [14:48] if you want to do fancy checks, use an install rule [14:48] persia: right, libparmetis3.1 is fine [14:49] Keybuk: blacklisting will keep it from loading, but it'll still get linked. restricted-manager is the way to deal with not even linking it [14:49] e.g. /etc/modprobe.d/lrm-video runs a program that will load the module or not [14:49] Keybuk: can't I edit the lrm-video or the lrm-check? [14:49] what's wrong with linking it? [14:49] well, if someone is silly enough to call their modified module the same name, the may not want it stomped on. [14:49] persia: so as soon as the other rdepds are solved, I'll kill it [14:50] lamont: / [14:50] lamont: ? [14:50] Keybuk: it sounded like they wanted to modify the driver... dunno [14:50] pitti: Great. I'll let you know. [14:50] pitti: was it you or keescook that made it so that having a null password is fatal these days> [14:50] drivers in one directory can override another [14:50] that's how stable driver updates work [14:50] ah, even better. [14:51] Keybuk: that and r-m links into a ramdisk, so it's not like it'll be overrwriting something [14:52] lamont: might be pam 0.99? anyway, not my feature/fault/praise/blame :) [14:52] Keybuk, lamont: I working to integrate Envy in Ubuntu as dholbach knows [14:52] here's the link: [14:52] https://wiki.ubuntu.com/EnvyNG [14:52] envy would need an LRM-a-like anyway, no? [14:53] Keybuk: what do you mean? [14:53] well, Envy has the same problem that caused lrm to exist [14:53] binary code linking to GPL code [14:53] sigh. and why does init=/bin/sh not work for me, i wonder. [14:53] BenC: just reading l-r-m changelog -> this affected nvidia, too, did you bump that as well? or just reformatted the version numbers? [14:53] * lamont blames pitti and keescook, just to be thorough. [14:54] Keybuk: what is not clear to me is how I can simulate what happens when the driver is blacklisted in the /etc/default/linux-restricted-modules-common [14:55] do you know what happens when a module is blacklisted there? [14:55] yes [14:55] lrm skips it [14:55] read /sbin/lrm-manager, it's just a shell script [14:55] pitti: nvidia shouldn't have been affected [14:56] Keybuk: yes I tried to edit it but the module wasn't skipped [14:56] tseliot: probably a bug, I don't think anyone's ever used that feature before [14:57] if I was going to do this, I would probably extend LRM to check whether the module existed in a different place, and if so, link that module instead [14:57] e.g. if doing nvidia, check whether /lib/linux-restricted-modules/$uname/extra/nvidia exists [14:58] and if it does, link that one and put that in volatile, not the one above [14:58] Keybuk: I'm using DKMS [14:58] (or /var/cache in fact, for dkms compatibility) [14:58] then pick up the module from there and link it to the kernel [14:58] rather than picking up its own [14:59] cjwatson_: as I sit here conversing with d-i, it occurs to me that when there is more than one nic, it'd be really cool in network conf (esp when netbooted) if the MAC addr were printed next to the name we chose for the interface. or have we finally made it so that when booted from network, eth0 is always what we booted from? [15:00] related to bug 56679 [15:00] Launchpad bug 56679 in netcfg "provide a method to use a specified MAC-address as the installation device" [Wishlist,Confirmed] https://launchpad.net/bugs/56679 === cjwatson_ is now known as cjwatson [15:01] Keybuk: do you think that I should only modify these lines? [15:01] if [ "$CURRENT_MODULE_DISABLED" = "false" ]; then [15:01] if type ld_static >/dev/null 2>&1; then [15:01] ld_static -d -r -o /lib/modules/"$KVER"/volatile/$module.ko $module/* || true [15:01] fi [15:01] fi [15:01] lamont: somebody needs to work on netcfg :-) [15:02] Keybuk: those lines belong to lrm-manager [15:02] tseliot: dunno, not without looking carefully [15:02] * lamont is reminded again that 9600 baud _S_U_C_K_S_. and this machine is loud [15:03] brb - reboot [15:03] * lamont finishes touching the partitioner, invokes that favorite of d-i steps "let me tell you what you're gonna do..." [15:08] ~ # mkdir /target [15:08] ~ # mount /dev/md0 /target [15:08] ~ # chroot /target [15:08] feh [15:08] it should say something after that. [15:10] lamont: /bin/sh not found, maybe? [15:19] Treenaks: that'd be consistent with init=/bin/sh failing [15:19] * lamont pushes d-i back to the top of the hill [15:26] yea!! back at a busybox prompt again [15:27] well... it'd be nice to know why that's not happy. === \sh is now known as \sh_away [15:40] sed is love. [15:40] * persia hands lamont awk [15:43] persia: awk is love too. [15:43] today, sed is more love. [15:49] BenC: then it will fail again [15:50] pitti: Hi, please give-back: lua-bit lua-cgi lua-copas lua-curl lua-doc lua-expat lua-filesystem lua-graph lua-logging lua-posix lua-rings lua-soap lua-sql lua-svn lua-xmlrpc lua-zip. Thanks :) [15:50] geser: done [16:06] BenC: it did fail again; can you please bump the nvidia versions as well, or re-juggle the version formatting, so that it is >> the current version? [16:07] can't understand how that could happen [16:07] pitti: will do [16:09] 1.0.9639 << 96.39 :) [16:09] ArneGoetje: my fonts appear to have all changed recently [16:09] do you know of any recent changes that could have caused that? [16:09] BenC: I guess an epoch bump will do, we don't have to keep it in sync with Debian or so [16:09] Keybuk: fontconfig [16:09] specifically it looks like Sans and Times have become different fonts entirely [16:10] what was the change? [16:11] Keybuk: don't know yet... didn't have time to take a lok at it... [16:11] fontconfig was uploaded a while ago [16:11] pitti: right, that's all I did for fglrx [16:11] the change is much more recent than that, ie. last couple of days [16:11] I noticed it in epiphany, so it may be a gecko change? [16:13] Keybuk: presumably since we switched from firefox to xulrunner? [16:13] Keybuk: epiphany since yesterday? that would be the switch to xulrunner1.9 [16:14] xulrunner? [16:14] Keybuk: is that a question? [16:14] hmm, about:config says it's using "Times" and "Helvetica" for fonts [16:14] instead of "serif" and "sans-serif" [16:14] Keybuk: xulrunner = libgecko [16:15] Keybuk: if you use the xulrunner version you will have noticed, the buttons look like GTK ones on the webpage [16:16] BenC: seems that you didn't use the latest l-r-m for 2.6.22 when you made .24? [16:16] so fglrx diverts libGL again? [16:16] seb128_: epiphany is currently held though [16:16] I'm on 2.20.2-1ubuntu2 [16:17] (the new epiphany-gecko breaks epiphany-extensions) [16:17] so I'm not on xulrunner [16:17] ok [16:17] some people don't care about epiphany-extensions and remove it apparently so I was mentioning it in case ;-) [16:18] BenC: I spent quite a lot of effort for that, and asked on u-k list for comments before uploading it [16:18] and got none [16:19] sorry, meant the kernel-team list [16:20] uh, make that u-d [16:21] lamont: bet you the shell is actually there but just producing no prompt [16:21] lamont: try 'id' to see if it's alive [16:21] lamont: I've noticed that bug a couple of times, but not tracked it down [16:33] Keybuk: do you know which script calls lrm-manager? [16:35] $ grep lrm-manager /var/lib/dpkg/info/*.postinst [16:35] /var/lib/dpkg/info/linux-restricted-modules-2.6.22-14-generic.postinst: lrm-manager --kver=2.6.22-14-generic [16:35] also an init.d script [16:35] /etc/init.d/linux-restricted-modules-common [16:37] I have modified the lrm-manager so that it can print the output of [16:38] a sentence to /var/log/envytest [16:38] but it doesn't do it [16:42] BTW /etc/init.d/linux-restricted-modules-common calls lrm-manager with [16:42] only the "--quick" parameter [16:45] therefore I just can't understand where the lrm-manager took its parameters [16:45] for the "for" loop (line 44) of /sbin/lrm-manager [16:45] any ideas? [16:46] or who should I bug about this? [16:47] there's nobody particularly maintaining lrm right now [16:47] who created lrm? was it fabbione? [16:48] i.e. Fabio Massimo di Nitto [16:48] ? [16:53] tseliot: check the changelog [16:54] azeem: the changelog of which package? [16:54] linux-restricted-modules? [16:55] the package containing the source you're looking at, I guess [16:55] pitti: please give back: moon-buddy turbomail musixtex-slurps price.app . Thanks [16:55] geser: done [16:56] geser: moon-buddy doesn't exist [16:56] moon-buggy [16:56] ah; given-back [16:57] azeem: I'm not dealing with a package. I'm just looking at /sbin/lrm-manager [16:58] and /sbin/lrm-manager is not in a package? [16:59] yes, I just found out it's in linux-restricted-modules [17:06] there are a few names which have to do with lrm-manager [17:06] mdz: do you know anything about this? === Keybuk_ is now known as Keybuk [17:18] tseliot: i didn't touch lrm in ages.. [17:19] tseliot: at least one year or so [17:19] it might have undergone plenty of changes [17:21] thegodfather: can you see if you know what happens in lrm-manager (line 44), please? [17:21] tseliot: version? [17:21] what version... [17:22] on gutsy [17:22] for module in *; do ... blah .... ;done [17:22] ogra: right [17:22] tseliot, thats like ls [17:22] tseliot: what's not clear about it? [17:22] it wlks over all the files in the dir [17:22] *walks [17:22] that reads like: for all files in that dir do;... ; done [17:23] yes, I know but where does it take its $* ? [17:23] so that it can check whether $1 is "nv", etc. [17:23] ? [17:24] another script must provide such parameters [17:24] but /etc/init.d/linux-restricted-modules-common seems to pass lrm-manager [17:24] Hi all! [17:25] only "--quick" [17:26] thegodfather, ogra: I hope it's clear what my doubt is. [17:28] the rest of the script is quite clear (case, shift, etc.) [17:28] any ideas? [17:29] tseliot: i am still downloading the source [17:29] thanks, I'll wait [17:30] tseliot: set -- $DISABLED_MODULES [17:30] this is the command that sets the new $@ [17:30] so $1 would be the first module in DISABLED_MODULES list [17:30] etc. [17:30] this is basic bash operation [17:30] s/bash/shell [17:31] and where is DISABLED_MODULES? [17:32] . /etc/default/linux-restricted-modules-common [17:32] sourced from there [17:32] ok, I see [17:33] and what does line 72 do? [17:34] tseliot: sorry but you are really asking basic shell stuff.. it's not related to this channel [17:34] did you find a bug in the script? [17:34] or something you are trying to debug? [17:34] thegodfather: I'll tell you what I'm trying to do [17:35] I would like to make sure that a module from the lrm is not loaded [17:35] tseliot: specify that module in /etc/default/linux-restricted-modules-common [17:35] that's all you need to do [17:36] without actually putting such module in the /etc/default/linux-restricted-modules-common [17:36] for example, if a certain file exists then the module won't be loaded [17:37] 'if [ -f /path/to/file ]; then DISABLED_MODULES="$DISABLED_MODULES nv"; fi' after sourcing /etc/default/blah [17:37] would do it [17:40] cjwatson: thank you very much [17:40] and thanks to all of you for your time [17:40] line 72> 'help type' [17:41] actually, 'man bash' and look for the section on the type builtin, since that tells you the relevant fact about its exit status [17:42] cjwatson: I'll RTFM ;) [17:43] might any python exports know why this doesn't work? "return str.__cmp__ (other.get_reason (), self.get_reason ())" [17:44] Small question, is the usplash still the default system for kubuntu/ubuntu, is the uswsusp beeng now the default, is a conflict between usplash and uswsusp fixed? [17:45] Riddell, the spaces in front of teh brackets ? [17:45] MaximLevitsky: there's no conflict between usplash and uswsusp? [17:45] Thre is , (at least here on kubuntu fiesty) [17:45] MaximLevitsky: it's better in 7.10 [17:45] System fails to resume (hangs) if usplash is active [17:46] ogra: no, the test case is 'str.__cmp__("hi", "low")' [17:46] And I know that this is connected to vt switching race [17:46] MaximLevitsky: does it work if you press alt-sysrq k after a little bit? [17:47] Mithrandir, just clears the screen [17:47] MaximLevitsky: hm, that makes it resume happily for me. [17:47] it's a kernel bug, really. [17:48] Riddell: str doesn't have a __cmp__; objects don't need __cmp__ if the rich comparison operators are defined [17:48] http://docs.python.org/ref/customization.html [17:49] when you call str.__cmp__() I think you actually get the __cmp__ method on the str class object which is totally not what you want [17:49] (which is why it expects 1 argument rather than 2) [17:50] why not just return cmp (other.get_reason (), self.get_reason ()) ? === ScottK2 is now known as ScottK [17:50] (or munge those into strings in advance if you need to [17:50] ) [17:53] cjwatson: well its comparing strings so __lt__ seems like a good substitute [17:53] __cmp__ is a three-way comparison, __lt__ isn't [17:54] does get_reason return a string? [17:54] if so, cmp is the right substitute IMO [17:54] you shouldn't really be calling __cmp__, __lt__, etc. directly [17:54] as in __lt__ (foo, bar) is an overcomplicated way to write foo < bar [17:57] cjwatson: yes it returns strings [17:57] < also works fine [17:58] tkamppeter: do you know where system-config-printer svn is these days? [18:25] greetings :) [18:25] i wanna ask for help... [18:26] i wanna develop a gui of "reportbug" written in python [18:26] can someone give advices or links for a book can help me? [18:37] cjwatson: If I launch lrm-manager the hack you suggested seems to work but it doesn't work at boot [18:38] for example something like this: [18:38] if [ -f /etc/default/envy-dkms ]; then [18:38] DISABLED_MODULES="$DISABLED_MODULES nv" [18:38] echo "nvidia">/etc/default/envytest [18:38] creates a /etc/default/envytest file only if I type "sudo lmr-manager --quick" [18:39] but, since lrm-manager is executed at boot it is weird that it doesn't work === ScottK2 is now known as ScottK [18:39] and that this file is not created [18:42] Riddell. system-config-printer SVN has recently moved. It is now at svn.fedoraproject.org/svn/system-config-printer/trunk A web interface for the SVN is currently not available. [18:43] Riddell, what did you change on s-c-p to make it working under Kubuntu? Did you pass all GTK to Qt? Or can Kubuntu ship GTK apps? [18:44] Riddell, I think we should whenever possible have only one s-c-p package in the distro, which works on all flavors: Ubuntu, Kubuntu. Xubuntu, Edubuntu. [18:46] Riddell, did you contact Tim Waugh and talked with him about your work on s-c-p? [18:49] tkamppeter: so far I've ported the applet to Qt, there's a lot of duplicated code [18:50] tkamppeter: I have contacted Tim Waugh but he doesn't seem too keen to have me work in his repository which would be the way to avoid much of the duplicated code [18:51] (and to keep it all in one package) === macd_ is now known as macd [18:53] Riddell, is Qt really necessary? Does Kubuntu ship libgtk? Or not? [18:53] tkamppeter: no, we don't use gtk [18:57] Riddell, so a whole distro Qt-based, which means that all config tools are different to Ubuntu? [18:58] tkamppeter: yep, same as ubuntu won't ship with hplip config tools since they use Qt [18:58] tkamppeter: and of course in many places its only different backends, much of the code is shared [18:59] but that seems to be something new to Fedora [19:04] Riddell, Fedora is an install-only distro on one DVD or 3-4 CDs, so they ship açways both desktops (where GNOME is their preferred one). So they never had to think about providing one and the same app with two GUI frontends. AFAIR they have even a theme which makes GTK and Qt apps look the same. [19:05] tkamppeter: the newest fedora release also have 1 CD "spins" for gnome and for KDE. the KDE one has to include a bunch of gtk apps which takes up a notable amount of space [19:05] so I'm hoping I can persuade him it'll help fedora too :) [19:08] Riddell, perhaps he could modularize the printing/CUPS magic and the GUI. Then you could write an additional GUI module. Perhaps he would host this structure upstream with the Qt part written by you. [19:10] tkamppeter: yes, that would be nice, but it needs tim waugh to let me work in his svn repository [19:13] If you can persuade him that a dual GUI will facilitate RH live CDs, but problems are probably that it will only help if all config tools get dual-GUI (needs a lot of man power) and that he has perhaps to do also Qt stuff when adding features or fixing bugs ... === bigon is now known as bigon` === \sh_away is now known as \sh === bigon` is now known as bigon === Zdra_ is now known as Zdra === bigon is now known as bigon` === bigon` is now known as bigon [21:16] Dec 19 06:25:33 j6700 udevd[2653]: add_to_rules: unknown key 'RUN{ignore_error}' in /etc/udev/rules.d/80-drivers.rules:5 [21:16] ew [21:17] oops. [21:17] Package: * [21:17] Pin: release a=hardy [21:17] Pin-Priority: 500 [21:17] how did that get to be 500, I wonder... :( [21:18] * lamont downgrades udev [21:20] * lamont runs the machine back down to gutsy, downgrading 193 packages. [21:30] hm, did gnome-gpg stopi working for anyone else? [21:36] * slangasek wonders if keescook has been spending a little too much quality time with gdb [21:37] hehe [21:37] keescook: I just use the regular gpg. :-) [21:38] I do too now, it seems. :P [21:38] heh. gnupg-agent + pinentry-gtk(?) is love [21:39] will the recently released perl 5.10 make gutsy+1 [21:39] (just today) [21:39] pinentry-curses is not so much love. [21:39] join #ubuntu [21:40] lamont: where do you config the pinentry bits? [21:40] (i just get a gui prompt if I add 'use-agent' to my options. [21:40] gpg-agent with --enable-ssh-support is even better. :-) [21:41] and having a OpenPGP card is even more fun :) [21:42] or two! :-) [21:42] keescook: afaik it uses /usr/bin/pinentry which points in the end to /usr/bin/pinentry-gtk-2 here [21:42] keescook: uh... that's configurable??? :-) [21:42] lamont: dunno, I didn't have pinentry-gtk installed, and it worked the same. :P [21:43] ah, ok. with pinentry-curses, I had wierdness [21:43] In the past, gnome-gpg would cause the gnome-keyring goo to prompt. now it says "Couldn't search keyring (code 9)" :P [21:43] mind you, I haven't checked in on it since sometime in feisty(?). [21:43] keescook: is that related to gnome-keyring not accepting ssh-add? [21:44] slangasek: I did notice some hiccups there too, but I always forced gnome-keyring to ignore ssh, so I'm not sure what the "right" behavior is. [21:47] how do you force gnome-keyring to ignore ssh? [21:47] because if it were ignoring it properly, at least then I could theoretically get ssh-agent to start [21:47] slangasek: I just always clicked "Cancel" ;) [21:47] oh :P [21:47] then ran ssh-add [21:48] and it'd stop prompting [21:48] the problem here is, gnome-keyring exports SSH_AUTH_SOCK and then can't actually accept keys via ssh-add [21:48] so because SSH_AUTH_SOCK is exported, ssh-agent doesn't start [22:10] slangasek: I just have a : unset SSH_AUTH_SOCK in my .bashrc, works fine :) (I would really like to have a better way to turn off the ssh part though :)) [22:18] stgraber: well, this is a new failure mode for me [22:35] Anyone on irc.canonical.com mind pinging pkl for me? [22:35] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=410095 [22:35] Debian bug 410095 in xscreensaver "xscreensaver: Provide .desktop files for gnome-screensaver" [Wishlist,Open] [22:38] * tormod wonders why that popped up - hyperactive touchpad probably... === \sh is now known as \sh_away === Martinp24 is now known as Martinp23 === asac_ is now known as asac [23:55] slangasek, ping