[00:00] Lutin: I don't know much PHP either, but you should be able to kinda of copy the LP link code [00:00] hmm, although you will need the first letter of the package name [00:01] LaserJock: and even worse, the first four if it starts with 'lib' [00:01] yeah [00:01] definitely too hard for me :p. going to bed. I'll ask Adri2000 tomorrow [00:01] k [00:01] no biggie [00:01] k :) [00:01] it's just one of my pet peeves with MoM [00:18] hi folks [00:18] how can i tell what file was actually used to install a package [00:18] like.. say i installed it from a deb [00:18] via dpkg -i [00:18] for testing [00:21] johnny_, you want to find the deb name given a file name ? [00:21] or what? [00:21] i installed a deb and i want to some proof that it got used [00:22] dpkg -l packagename [00:22] it will show you the installed version [00:22] it's the same version [00:23] not what you expected ? [00:23] well i built the package with the same version [00:23] because i didn't know any better perhaps [00:23] oh, then it's difficult to verify [00:24] johnny_: would looking at the contents of your .deb help? [00:25] what i wanted was a way to say that /root/ldm2_5.0.39.deb was used to install vs the ubuntu one [00:25] johnny_: well, if you dpkg -i it and it installs then it worked [00:25] You could do md5sum checking. [00:26] johnny_: and you don't need to package as root [00:26] johnny_: what are you patching in ldm? [00:26] auto-login stuff? [00:28] i didn't package as root [00:28] i just put it in the i386 chroot [00:28] so i could install it [00:28] ok.. [00:28] there we go [00:29] yes.. LaserJock you know .. cuz you're in that other channel :) [00:29] aha.. i figured it out [00:29] to prove to myself [00:29] i added some .... to one of the installed shell scripts and they appear [00:30] since i obviously couldn't look at the source [00:31] since i'm installing a binary [00:31] i come from the gentoo world.. so all the requirements to build packages is kinda confusing to me [00:31] i can built gentoo packages very easily [00:32] we use ubuntu at our coffeeshop, so i've had to learn a lot more of deb stuff [00:32] good for you :-) [00:33] johnny_: I haven't been tracking the auto-login stuff much, would your package be generally useful? [00:34] not really [00:34] ogra has told me that the patch much be simpler [00:34] and if it is, it will be packages for inclusion [00:34] packaged* [00:34] francis already has one [00:34] ok [00:35] but the debs for the thin client are wrong [00:35] err deps [00:35] it wants a version of pango that isn't in the thin client [00:35] it might be related to post gutsy changes [00:35] where the thin clients are supposed to have a newer version of pango [00:36] my system pango is one version higher than in the chroot [00:36] ah [00:37] so i might just bug francis to fix up his packages and make the actual patch smaller [00:37] i just need something working, and an excuse to figure out packaging [00:37] our entire business is almost completely based off open source stuff [00:37] and using ubuntu [00:38] we do have a few macs involved for illustrator and photoshop purposes [00:38] but otherwise.. we even use gnucash for finance [00:38] cool [00:38] we're a non profit store for the most part [00:38] it's pretty great env we're setting up now [00:50] darn, axiom died [01:17] window level all [01:18] oops :( [01:18] ;) [01:56] sudo has changed in Hardy so it does not forward env?? e.g. "FOO=1 sudo su -c 'echo $FOO'" does not work anymore. [01:56] I guess that's the reason my .pbuilderrc breaks for DIST=xxx [01:58] See this recipe: https://wiki.ubuntu.com/PbuilderHowto#head-1be378ab60d3bab23eefabce49cf7df927d46f81 [01:59] blueyed: yes, according to the changelog this was done in version 1.6.9p6-1ubuntu1; the Ubuntu sudo package's behavior is now consistent with the defaults upstream and in Debian [02:00] slangasek: so it would make sense maybe to change the "sudo pbuilder" call to "sudo -E pbuilder"? [02:01] that appears to be correct [02:04] PBUILDERROOTCMD="sudo -E" - should it be the default? [02:04] Heya gang [02:04] Hi bddebian [02:04] Hello blueyed [02:09] bug 175776 [02:09] Launchpad bug 175776 in pbuilder "PBUILDERROOTCMD should default to "sudo -E"" [Undecided,New] https://launchpad.net/bugs/175776 [02:09] Good night. [02:10] Gnight [02:12] Heya persia [02:12] heya bddebian [02:13] persia, hi [02:13] http://stompbox.typepad.com/blog/2007/12/prism-is-neat.html [02:13] 1st fan [02:13] Ubulette: :) [02:53] persia: Do you have an unstable machine? [02:53] heh [02:53] bddebian: Yes, but only a chroot for sid. [02:54] persia: Do you have problems with libsdl1.2-dev atm? [02:54] Heya LaserJock [02:55] bddebian: No, but I'm guessing you're encountering audio issues related to competition between libsdl-1.2debian-alsa and libsdl-1.2debian-pulseaudio [02:55] (or, not competition between the libraries, but between the audio outputs) [02:55] persia: No, I can't even install libsdl1.2-dev [02:56] bddebian: Odd. I'll take a look... [02:56] In my sid pbuilder [02:56] I've got the weirdest problem with axiom [02:56] it looks like we can sync it [02:56] but it FTBFS in my hardy pbuilder [02:57] I can't figure out why it would do that [02:58] What's the error? [02:58] here's the relevant section of the build log: http://paste.ubuntu.com/2655/ [02:59] I checked to make sure it built on the Debian buildds and it did [03:00] Whacky [03:00] yeah [03:00] I tried on two different machines [03:01] I'll try it with my sid pbuilder, but the thing takes a long time to build [03:01] bddebian: Debian Bug #455490 [03:01] Debian bug 455490 in libsdl1.2 "libsdl1.2: Uninstallable due to dependency on libdirectfb-0.9-25" [Grave,Open] http://bugs.debian.org/455490 [03:01] Does configure miss anything? Maybe debian pulls in a build-dep that we dont? [03:01] persia: Grr, thx. I checked the bts but I obviously missed it [03:02] bddebian: yeah, I'll try that. It's weird though cause it's not an error I'd expect if it worked in Debian [03:02] it seems to be failing on cp not on actual building [03:03] LaserJock: Is this a repeatable error? [03:03] persia: I built it on two different machines [03:03] * persia finds the answer in backscroll, and withdraws the question [03:03] * persia is, as usual, slow [03:03] although I think I may have scp'd the base.tgz to the other machine [03:03] so the pbuilders are "related" [03:04] LaserJock: This is building sid source against hardy? [03:04] Is axiom one of those that takes forever to build? [03:04] bddebian: yep, 40MB source and builds latex and emacs stuff [03:04] persia: yep [03:05] * persia tries with a different build syste [03:05] m [03:05] I'm pretty sure this is a sync [03:07] * bddebian tries a "quick" build :-) [03:18] persia: boswars should be coming to Debian *soon* [03:18] bddebian: Just waiting on libsdl1.2-dev :) [03:18] Heh, no shit :) [03:20] dang, I still have 10min left of downloading deps before I even get to start compiling axiom [03:21] LaserJock: lol :P [03:21] local mirror FTW! [03:21] Hobbsee: :D [03:22] heya! [03:23] well i am going to retire for the evening [03:23] Goodnight All [03:23] Gnight joejaxx [03:24] Hobbsee: yeah, but I can't mirror everything [03:25] who said anything about who's mirror it was? [03:25] LaserJock: get UNR to put up a mirror :) [03:26] imbrando1: I should, but they don't seem to be very Linux friendly from what I've heard [03:27] I should ask though, maybe they do have a mirror and I don't know about it :-) [03:28] heh [03:30] LaserJock: I get the same error [03:31] bah [03:31] well at least my pbuilder isn't borken [03:40] * somerville32 is downgrading his internet connection :( [03:40] downgrading? [03:40] as in slower? [03:41] Yea [03:41] 7Mbs Down/256Kbs Up -> 1Mbs Down/128Kbs Up [03:42] What sort of crazy world gives you 128 kbps up? [03:42] Is 128kps up that bad? :S [03:43] I have 1.5 down/5** down I think [03:43] bah 5** up [03:44] either that or 3** up [03:44] I think mine's 786 up. [03:44] I can't remember [03:44] The next step up is 8 up and 1mb down [03:44] * TheMuso has 1.5 down, and 256 up. [03:45] I'm downgrading to save money after they lied to me about a promotion [03:45] Sadly, they don't have a good intermediate package so now I'm moving to "Light" at the end of the month [03:45] I just get the cheapest DSL I can find [03:46] I should have stuck with the express [03:46] It is only like $10 more anyhow :/ [03:46] * RAOF got the ~$50/month adsl2+ [03:46] Mine is cable [03:47] LaserJock: I get a different error: http://paste.ubuntu.com/2656/ [03:47] Heh, 22mbit down, 1mbit up. :) [03:47] * victor_ got the 150$/month 4 MBit/s down 384 kbit/s up. [03:48] : O [03:48] persia: hmm :( [03:48] Wow. Somewhere worse than .au :) [03:48] bddebian: so you got the same as me? [03:48] LaserJock: yessir [03:48] Bah. .us & .au (& .nz & .ca) need to upgrade the common infrastructure. [03:49] 4 MBit/s is enough to stream tv/movies [03:49] persia: that was -10? [03:49] Omg, I feel so stupid. I shouldn't have downgraded :/ [03:49] somerville32: why? [03:50] * imbrando1 has 6M/384K [03:50] LaserJock: axiom_20050901-10.dsc [03:50] I just calculated it, and it is only like $10 difference [03:50] * LaserJock stabs a straw into imbrando1's computer [03:50] not bad, mine's only about 6M/800K [03:50] persia: well that's weird [03:50] (or thereabouts) [03:50] persia: common infra? isnt that against what the internet is about, diveristy hehe [03:51] who has ever said that it's about diversity? that's only what 4chan wants you to believe! :) [03:51] somerville32: so call them back [03:51] :) [03:51] imbrando1: Maybe, but those are all tiny bandwidth numbers. Ethernet is preferable. [03:51] imbrando1, It sounds like it was painful for them to switch it over and credit me crap when I did it :P [03:52] imbrando1, I might make them upset if I call back tomorrow and say "I changed my mind, sorry" :P [03:52] somerville32: thats why you pay then its their job :) [03:52] I guess I'll wait and see if I even notice that big of a difference [03:52] It is unbearable, I'll call them back [03:53] you'll feel like your on isdn :) [03:53] hehe [03:53] I never did isdn [03:53] me neither [03:54] I could always go to ultra lite - $20 a month, 64kbs up and down [03:54] lol [03:54] heh i had 2400, then 14.4 then 33.6 then 56.x and then bonded 56k then isdn then cable now dsl [03:54] my parents have 28.8 [03:54] it's sooooo painful [03:55] If I can use POP3 or IMAP, which should I pick? [03:55] i skipped that one hehe, its not even good nuff to check modern mail [03:55] somerville32: imap IMHO if you have enough storage server side [03:55] imbrando1, google :P [03:55] imap [03:56] i love thier imap [03:56] or use prism-google-gmail :) hehe /joke [03:56] there is no gears support yet [03:57] * TheMuso waits for the day when debmirror syncs those important pieces of the archive, such as indices, and installer-$ARCH files. [03:57] Until then, I have my script. :p [03:57] TheMuso: i can hack that into apt-mirror pretty simply [03:58] imbrando1: apt-mirror doesn't suit my needs. [03:58] ok, kinda dumbfounded on that one , but okies [03:59] imbrando1: However, if it did support that stuff, I would re-consider it. I probably didn't give it enough of a look for what I needed to do [04:00] TheMuso: i'd be intrested in hearing your use case , i'm not doubtign it, but i have failed to see a case where apt-mirror doesnt do what debmirror does and more+better [04:00] doubting* [04:01] imbrando1: Well, to build disks, you need the indices files, as well as the dists/$dist/main/installer-$ARCH files. [04:01] right but neither does debmirror, i ment more of your first statement [04:01] And dists/$dist/$section/debian-installer files. [04:01] imbrando1: Yeah I know debmirror doesn't, which is why I've scripted it for now. [04:01] The virtual package libgoffice-0-dev doesn't work anymore since libgoffice version has been updated to 0.5.3-1ubuntu1, can someone here fix this? [04:01] this is in Hardy [04:02] But debmirror allows the use of rsync. [04:02] Afaik apt-mirror doesn't. [04:02] * persia declares perl superior to ruby for parsing multimegabyte XML files [04:02] !doesntwork > saivann [04:02] ahh no it doesnt TheMuso [04:02] imbrando1: And my local mirror offers rsync, so I'm using it. [04:02] it would be tricial to add though [04:03] Yeah likely enough. [04:03] trivial* [04:03] persia: ? [04:04] Sorry, I mean that the virtual libgoffice-0-dev package doesn't point to latest libgoffice version in hardy [04:04] imbrando1: for the same input, and the same processing library, a one-line perl script does XML -> YAML in ~800MB of working space vs. ~4GB for a one line ruby script. [04:04] 12.7K e-mails since I opened my gmail account, hehe :/ [04:04] persia: yea, i ment more what it pertained to hehe [04:05] somerville32: nice you have about ~50% of what i do :) [04:05] imbrando1: My suddenly not having my IRC client swapping out :) [04:05] hehe === imbrando1 is now known as imbrandon [04:05] ahh better [04:05] Anybody have a (modern) LaTeX resume/CV document class? [04:05] Oh wait [04:05] I guess that was only bug report e-mails, lol [04:06] somerville32: i was gonna say thats kinda low [04:06] What ever it is doing now it says 33.8K [04:09] just checked i have about 36k stored [04:10] thats after a massive cleanup [04:10] And thats for your gmail account? [04:10] Why would you delete stuff in gmail? [04:11] yea, because it was not needed [04:11] most of it is old list archives [04:11] already archived on the web [04:12] I'm a data pack rat [04:12] Never know when you'll need it :] [04:16] Does somebody knows if the fact that libgoffice-0-dev doesn't exist anymore in Hardy is normal? libgoffice-0-5-dev seems to replace it, but I want to be sure [04:17] saivann: There's a transition underway, not currently complete. As expected, hardy is broken in several ways. [04:17] persia, LaserJock: Just to be contrary, I get a *different* error to both of you when trying to build axiom. [04:18] * persia suspects axiom to be a troublesome package [04:18] * somerville32 suspects that persia is correct. [04:19] persia : Gnucash dependencies is currently broken because of that package, do you believe that I should ajust gnucash dependencies or wait to see if libgoffice-0-dev package come back? [04:19] saivann: The APIs differ. Could be ported. Depends on how motivated you are. [04:20] persia : Own, you mean that gnucash have some chance to not work with libgoffice-0-5-dev ? [04:22] saivann: Yes. [04:25] persia : In your point of view, what's the better idea? Trying to make gnucash working with the new libgoffice version or port old libgoffice version from ubuntu gutsy in order to make sure gnucash will work? [04:26] persia : I'm ready to work on this but I need some guidance because I'm really new to this [04:27] the latter is part of the transition eventualy there will be a -0-4-dev iirc, as for doing something NOW, porting to 5 is the only viable option [04:27] saivann: I wouldn't do the former without checking mailing list archives, especially upstream. The latter is pointless, as it won't "make sure" gnucash will work in hardy: it'd be better to understand why it was dropped, and see what the library maintainer has planned for the transition. [04:28] persia : Great, I'll take a look at this, thanks for your help === asac_ is now known as asac [04:33] saivann: we have old goffice as of today [04:34] saivann: Thanks for looking into it. Good luck. [04:34] LaserJock : What do you mean? [04:34] saivann: we got goffice 0.4 today [04:34] persia : Thanks, I will have fun with this :) [04:35] saivann: I'm assuming that's what you're needing [04:35] LaserJock : Will it be available in Hardy? [04:35] LaserJock : Yes, exactly [04:35] saivann: yes, as of today :-) [04:35] we had a bit of fun upstream with goffice [04:35] LaserJock : Great, that means that I can sleep tonight :) [04:35] which caused goffice to go to 0.5 [04:36] hehe [04:36] but 0.4 is the stable version [04:36] so gnumeric uses goffice, and stuff that needs 0.4 should use the new goffice0.4 packages [04:36] LaserJock : I'm glad to hear this since no update will be needed for Gnucash [04:37] we'll have to make sure it uses the right stuff, but goffice 0.4 *is* in Hardy today [04:37] LaserJock : I assume that the packages names won't change? Will they? [04:37] yep [04:37] we can't have multiple versions with the same name [04:37] imbrandon, 40.5K all together [04:38] saivann: actually let me check the -dev real quick [04:38] LaserJock : I only see the 0.5 versions in Hardy from now, is it normal? [04:38] saivann: may depend on your mirror [04:39] * persia notes that it may also take a while to compile recently imported libraries [04:40] saivann: libgoffice-0-5-dev is for goffice 0.5 and libgoffice-0-dev is for goffice 0.4 [04:41] LaserJock : Thanks, it's still not available from now "libgoffice-0-dev has no installation candidate" but I'll wait few days [04:41] saivann: is gnucash having problems right now? why did it come up? [04:42] LaserJock : It doesn't install because it depends on libgoffice-0-4 which doesn't exist in repositories [04:42] ah [04:42] but it *does* have that dependency? [04:42] Yes [04:43] hmm, our package is kinda old, I'm surprised it has the dependency [04:43] are you talking about the current Debian or Hardy package? [04:43] current Hardy package [04:44] but it's not that old, it's the latest upstream version 2.2.1 [04:44] merged from debian [04:44] yeah [04:44] but goffice 0.4 is about 1 month old [04:45] LaserJock: libgoffice-0-4-dev is about 1 month old. libgoffice-0-4 is older. [04:45] well ... [04:45] only in experimental [04:46] as a part of goffice, not goffice0.4 [04:46] LaserJock: Yes, as a part of goffice, but the binary package dependencies might not know the difference, due to the magic of shlibs (not that I've actually investigated this) [04:47] yeah [04:47] if gnucash in Debian was last built from experimental [04:48] see [04:48] this is why it's nice when people only package a library for one package [04:48] :) [04:49] gnumeric uses goffice 0.5 and *everything* else uses 0.4 [04:49] LaserJock: What? We don't import Debian builds. [04:50] persia: sorry, I see it now [04:50] I thought we missed goffice version 0.4 [04:50] LaserJock : Will the package *named libgoffice-0-4, libgoffice-gtk-0-4 and libgoffice-dev-0* came back or? [04:51] well [04:52] LaserJock: No, just a source transition, from a single recommended version of the library to two supported versions. I don't like these because it leads to things like wx2.4 + wx2.6 + wx2.8 or db4.* or automake-*, but it can significantly reduce (or rather, defer) the porting effort required for the transition. [04:52] we have: libgoffice-0-4, libgoffice-0-4-dbg, libgoffice-0-common, libgoffice-0-dev [04:53] from goffice0.4 source package [04:53] persia: well, if some people didn't insist on doing certain things this wouldn't be a problem [04:53] persia: automake is a specific case, since you want to use exactly what version upstream is [04:54] LaserJock: Hrm? How do we control which version of libraries upstream links against? (Although you may be speaking about upstream). [04:54] StevenK: Yes, but wouldn't it be nifty if all upstreams used the same version? [04:55] persia: no, I'm speaking of package maintainers who package unstable versions of libs for 1 app leaving the other packages hanging [04:55] LaserJock : Great, so I just want to be sure of this : I don't have to make any modification on the Gnucash package? [04:55] saivann: I hope not [04:55] LaserJock: Ah. In that case, I'd suggest they should be agressively chasing NBS to port everything. [04:55] NBS? [04:56] LaserJock : Can I contact you about this if libgoffice-0-dev is still not available after a few days? [04:56] LaserJock: http://people.ubuntu.com/~ubuntu-archive/NBS/ - if someone starts a library transition, they should work to get everything working with the new library. [04:56] saivann: yes, please do, mantha at ubuntu.com [04:57] LaserJock : Thanks a lot [04:57] persia: yeah, this is part of why I'm starting to hate libraries [04:58] LaserJock: Usually the porting isn't so bad, and transitions can be fun: what other time do you have the excuse to upload 30 or so fixes simultaneously :) [04:58] persia: well, actually the porting was deemed "umm, no" by upstream [04:59] hrm "w00t" is getting added to the Webster dictionary, and coined word of the year 2007 [04:59] ugh [04:59] LaserJock: Ah. That's a case where it'd be nice to have an annoucement about the transition, and documentation of the package split (like for WX). [05:00] persia: yes, we got a bunch of Debian QA bugs filed yesterday [05:00] and we had to explain why our package dep'd on a lib in the oldlibs sectoin [05:00] *section [05:01] Then there is things such as libglib1.2 -> libglib1.2ldbl [05:01] Where there is no actual API change, just a package name transition. [05:01] LaserJock: That's a good thing: at least someone cares. For a couple of the apps I really should be porting some day, the dependency isn't even considered oldlibs, as nobody really cares. [05:01] so not only was their a change in libs but the Debian archive admins sent it to oldlibs [05:01] TheMuso: Wasn't that also an ABI change? [05:02] persia: I'm not really sure, but packages were uninstallable because they couldn't find libglib1.2 [05:03] TheMuso: I think it was a redefinition of a long double from 64-bit to 128-bit. (but I'm not sure) [05:04] ah right. [05:11] wow, axiom is *still* building here [05:29] I am completely puzzled as to why axiom builds in Debian but not Hardy :( [05:31] What is the error? [05:31] well, 4 people got 3 different errors [05:32] Ok, I'll try [05:32] clearly, these errors are themselves not axiomatic [05:33] * RAOF pokes slangasek with a halibut [05:33] * persia doesn't understand the poking. Shouldn't it be more of a whallop? [05:33] fish are food, not friends [05:34] I agree with persia, how do you poke somebody with a fish? [05:34] Talking fish can be friends. [05:34] maybe a swordfish [05:34] LaserJock: gingerly [05:34] LaserJock: Freeze it first, or use a stiff fish. [05:35] * RAOF is surprised to notice the emacs22 build dependency of axiom [05:35] yeah? [05:36] my guess is it's compiling something [05:36] * persia wonders if there's really a significant speed gain for compiled elisp [05:36] it depends I think [05:37] considering how long it takes to compile it I often wonder [05:38] cody-somerville@RAOF:~/packages/gfceu$ sbuild -d hardy gfceu_0.6.0-0ubuntu1.dsc [05:38] chroot_mode is obsolete at /usr/share/perl5/Sbuild/Conf.pm line 163. [05:38] How do I fix that? :P [05:39] nvm [05:39] somerville32: Edit your schroot.conf [05:40] Or ~/.sbuildrc, IIRC. [05:40] (or maybe sbuild.conf: I forget: it's a completly opaque error with such a trivial solution it's hard to remember) [05:40] When I use sbuild, is it suppose to just say "Build started at ...\n****..." and then nothing? [05:41] somerville32: No, it should give a build log. [05:41] RAOF, It hasn't finished [05:41] Although it may be a bit slow, because we're contending for resources :) [05:41] oh [05:41] sbuild is weird :/ [05:42] How do I get it to compile something arch all? [05:42] "gfceu_0.6.0-0ubuntu1.dsc: amd64 not in arch list: all -- skipping" [05:42] somerville32: Either use the i386 builder, or pass --arch-all [05:42] * LaserJock hugs pbuilder [05:43] ich auch. [05:43] :S [05:43] Hobbsee, at which point did you learn german? [05:44] DarkMageZ: only some of it, and in school. [05:44] * persia notes that pbuilder is different than sbuild, and there are a number of cases where something builds in pbuilder and fails on the buildds [05:50] woots [05:51] Finished? [05:51] Something happened? :) [05:53] It compiled. [05:53] And now I shall have a patch for persia to review in a minute [05:53] somerville32: -ENOCONTEXT [05:54] persia, To take my mind off the death of my friend, I decided to start working on the watchfile project [05:54] somerville32: Ah. Thanks for working on that, although I'm certainly not the only person who can review the results. [05:55] persia, I'm quite certain of that too [05:55]