[00:05] Hi guys! [00:06] I'm having issues with gambas2 packaging [00:06] lol, i forgot some commas [00:07] Forget it [00:17] figli di puttana [00:18] ops [00:18] wrong window [00:19] that's what they all say [00:19] I am highly offended sir [00:19] * tacone turns red :-) [00:19] ;) [00:19] Pueribus Impudis [00:20] Anyway, enough butchered latin for one night. I'm off to bed [00:22] hhahah [00:22] poor tacone :) [00:22] not my fault. world's against me :-D [00:26] Hi, all. [00:29] I have a question about mentorship program. Could someone help me? [00:34] kunigami: don't ask to ask :-) [00:38] Any hint from where to start? I read a lot of things in the wiki, but I'm a bit confused. [00:44] did you read https://wiki.ubuntu.com/MOTU/GettingStarted ? [00:46] Heya gang [00:51] vorian, good :) how ya doing!? [00:59] Well, I tried to do so, but it seems a bit confusing for me =( [01:17] kunigami: to me too. You could just trying to solve a bug then ask here for help when you get some issue [01:31] Ncommander, you around? [01:32] RoAkSoAx: howdy [01:32] vorian, good good, how ya doing? [01:32] great [01:33] vorian, so what's new? i've been giving a quick review to the debian new maintainer's guide and i have setup this wikipage: https://wiki.ubuntu.com/4nDr3s/Mentorship [01:33] * vorian looks [01:34] neuromit, yup [01:35] you got some time to talk to me about compiling multiple debs from a single source package? [01:35] I'm home now and I've ssh'd into my work machine [01:35] hello MOTUs [01:35] neuromit, yeah, I can talk [01:36] moin [01:36] neuromit, its pretty straightforward; you have multiple binary entries in the control file, and then you use *package*.install files to move files from debian/tmp to each package folder [01:37] NCommander: i'm guessing that the file names of the .install files has the be the same as the packages in control? [01:37] neuromit, yup [01:38] neuromit, its pretty straightforward if you know how to use dh_install [01:38] so I've never used dh_install by hand before... [01:39] dh_install is called by dpgk-buildpackage right? I mean unless its commented out in the rules file right? [01:41] Can some one take a look at my REVU please? http://revu.ubuntuwire.com/details.py?upid=2900 [01:46] NCommander, does dh_install take an install file as an argument? === nenolod_ is now known as nenolod [01:47] neuromit_, All the debhelper scripts have man files :) [01:51] cody-somerville: do you have time to check out a revu upload? [01:53] so when I build a package in the form of a single binary a .install file isn't created.... if I create one will it disrupt the building of the deb? [02:04] so I've been reading the man pages for dpkg-buildpackage, dh_install, dh_icons.... and I can't quite seem to figure out how to package an icon with my debian package... I've included the icon in the debian dir, but I'm not exactly sure what to add to the rules file to include it in the installation [02:06] neuromit_: I would put an entry for it in your *package*.install file [02:06] Or do you mean like a .desktop type file? [02:08] no I mean a icon file. So when I followed the ubuntu packaging guide I thought I didn't need a *package*.installl file if I was compiling a single binary, so I never created one, if I create one will it disrupt the compiling of the deb, if all that is in it is the icon file? [02:11] neuromit_: No, not at all. [02:12] ok one more question... (sorry for flooding ;-p ) but will dh_install automatically install all of the *package*.install files or do I have to specify them in the rules file? [02:12] hiya nxvl [02:13] hi! [02:13] neuromit_: Reading 'man dh_install' will tell you that; the answer is that it will automatically read $package.install (with some fairly minor assumptions). [02:15] heya nxvl [02:15] hi [02:15] :D [02:21] nxvl: you worked with anything that had a weird upstream version number? [02:23] or, anyone have the right advice for dealing with upstream versions with hyphens in them? [02:23] a lot [02:23] :D [02:23] * NCommander returns [02:24] nxvl: Can you tell me what to do when upstream is "1.4.3-2"? all the debian utilities think I'm making a debian native package because of the hyphem [02:24] tbielawa: 1.4.3-2-1 [02:25] You mean to add a -1? [02:25] tbielawa: package and upstream link [02:25] so in the end it would be bibus-1.4.3-2-1-0ubuntu1 ? [02:28] heh, to continue the question spam, if anyone has the URL to something that can clarify this for me I'd appreciate it. the deb policy manual wasn't clear enough fo rme [02:28] tbielawa: No, bibus 1.4.3-2-0ubuntu1 [02:29] tbielawa: debian policy [02:29] tbielawa: can you send me a link to the upstream page [02:29] wgrant: lintian and all deb helper scripts refude to belive i'm making anything but a native debian package. when I build it it refuses to see the .orig.tar.gz [02:29] nxvl: link coming [02:30] tbielawa: How have you named the .orig.tar.gz? [02:30] http://sourceforge.net/project/showfiles.php?group_id=110943 [02:31] wgrant: bibus-1.4.3-2.orig.tar.gz [02:31] tbielawa: bibus_1.4.3-2.orig.tar.gz [02:31] actually -2 is the debian version [02:32] nxvl: No it isn't. === leleobhz is now known as metilfenidato [02:32] nxvl: The part after the last - is the Debian version. === metilfenidato is now known as leleobhz [02:32] Not if it's in the upstream version. [02:32] Oh. [02:32] I see. [02:32] You're right - it's not part of the upstream version at all. [02:32] this is where all my problems are coming from. He debianized it, wayyy out of spec [02:32] wgrant: are you here -> http://sourceforge.net/project/showfiles.php?group_id=110943&package_id=151072&release_id=600703 [02:33] is that package in debian? [02:33] tbielawa: Use the real 1.4.3 tarball. [02:33] yep [02:33] nxvl: I see now - I was taking tbielawa's word that the upstream version was 1.4.3-2. [02:33] -2 is the upstream-debian version [02:33] which we don't care of [02:33] wgrant: oh ok :D [02:33] there is none? [02:33] the upstream version we care about is 1.4.3 [02:34] rather @wgrant: there is no version as just 1.4.3 [02:34] tbielawa: Then you complain at upstream for being ridiculous. [02:34] * tbielawa remembers talking with LaserJock about why no one wanted to touch this package [02:34] tbielawa: take a look at the link you first post [02:35] tbielawa: it has a debian and a windows version [02:35] tbielawa: the windows version doesn't has the -2 part [02:35] so we can assume it's not part of upstream [02:35] but from upstream packaging [02:35] nxvl: I see what you mean. [02:35] so we don't care about it, because we know upstream packaging goes from bad to horrible [02:35] I need to get a clean un debianized 1.4.3 from upstream [02:36] tbielawa: Correct. [02:36] tbielawa: just take the .orig.tar.gz === RoAk is now known as RoAkSoAx [02:37] nxvl: take it from where? I've got this up on REVU almost done save for this rediculous versioning problems [02:37] nxvl: There isn't one. [02:37] * nxvl rechecks [02:37] nxvl: Upstream has packaged it natively. There is no 1.4.3 tarball. [02:37] so I've created my *package*.install file, i uncommented dh_install in the rules file, I added the appropriate dirs to the dir file, and I added my icon file exist under the debian directory, but when I run dpkg-buildpackage -rfakeroot, I get cp: cannot stat `./usr/share/icons/*icon*.png': No such file or directory I can't figure out what i'm doing wrong [02:37] Which is plain wrong. [02:38] and in the end, this winds up breaking my watch file [02:38] well [02:38] hes using his sourceforge file serving part as his apt mirror [02:38] yes, but, there is a .tar.gz [02:38] nxvl: But it's a native Debian packaged one. [02:38] so you can download it, remove debian/ and you got orig.tar.gz [02:38] or something [02:38] ok [02:39] what would i do in this particular case: [02:39] ping upstream and ask for a clean tarball [02:39] I agree [02:39] thanks [02:40] this has been hell getting through. it's two months in now [02:40] I just want to get it finished in time to get it into intrepid [02:41] neuromit_: can you pastebin us your .install file and you .dirs file? [02:41] http://pastebin.ubuntu.com [02:43] http://pastebin.ubuntu.com/29169/ [02:44] neuromit_: when working with the .install file this helps me remember whats going on [02:44] imagine it's like using the copy command and that it's being run from the main directory of your package [02:45] so you'll want something like this in your .install file: [02:45] soma-tspikes.png usr/share/icons [02:46] ok.. I wasn't sure I could do that as none of the examples of in man page of dh_install were like that [02:46] * tbielawa nods [02:55] tbielawa, Thanks! that did it! [02:56] neuromit_: you're very much welcome [02:57] I just sent upstream an email about his wacked out version scheme. He's been very responsive in the past so this should get resolved soon :-) === philwyett_ is now known as philwyett [04:35] NCommander: You should close bug 247782 in Intrepid [04:35] Launchpad bug 247782 in mono "Ubuntu mono patch dont_check_proc_self_exe causes random segfaults in mono" [Undecided,Confirmed] https://launchpad.net/bugs/247782 [04:35] StevenK, not if the bug needs to be SRUed [04:36] NCommander: Well, do you think it requires one? [04:36] Random segfaults in hardy building stuff would probably count to me. [04:36] As serve regression [04:37] Only on a Live CD, unless I'm misremembering? [04:37] No [04:37] It effects everything [04:37] THe original patch allowed for Mono to work on a LiveCD [04:37] But it causes random segfaults all over the place. evolution-sharp's FTBFS on amd64 was caused by this [04:37] The bug was discovered by NCommander while working out why evolution-sharp built in sid chroots but not intrepid chroots. [04:38] RAOF, in addition, it did the same thing on Hardy [04:38] (But just fine in a non-chroot environment) [04:38] RAOF, no, it FTBFSed on me in a non-chroot environment [04:38] * StevenK approves the Hardy/Intrepid bits [04:38] NCommander: Oh, even better :). Worked for me out of teh chroot :) [04:38] * StevenK unsubscribes u-m-s [04:39] It was a rather inconsistant bug .... [04:39] NCommander: Now, fiddle with the bug, closing the Intrepid task :-) [04:39] * NCommander does what StevenK wishs [04:39] I don't think the patch is in dapper, I should probably see if this braindeadness is in gusty or feisty [04:40] I'm reconfirming evo-sharp FTBFS on Hardy right now === asac_ is now known as asac [04:42] hi all, I am wondering if there are some helpful souls here that might mentor me in fixing a simple bug [04:42] bug 250681 is simple enough, I apt-get source irssi [04:42] Launchpad bug 250681 in irssi "first run help link outdated" [Undecided,In progress] https://launchpad.net/bugs/250681 [04:42] bug 250681 is simple enough, I apt-get source irssi [04:43] found the string that needs changing [04:43] changed it, but now what should I do? I think I need to make a deb diff or something right? [04:43] Awsoonn, you need to modify the changelog to create a new ubuntu revision, and possibly update the maintainer, dpkg-buildpackage, and then debdiff the old .dsc and the new .dsc and post [04:43] THat's pretty much it in a nutshell [04:44] * Awsoonn is really new and only kinda understands [04:45] You need to build an updated package [04:46] jcastro: around? [04:46] If it already has ubuntu specific changes, you just need to add a new changelog entry to debian/changelog [04:46] well, accually, before that the source came with a diff should I apply that to the source, then make my change [04:46] start with sudo apt-get build-deps irssi [04:46] or build-dep, can't remember [04:47] no s [04:47] in any case it is done [04:47] k now, is there anything in /debian/patches? [04:47] there is no such folder [04:48] hmm, how big is debian/rules? [04:48] StevenK, I don't have a reproducible bug for Hardy just yet [04:48] there is no debian folder at all [04:48] what? [04:48] whoa [04:48] coppro: depending on what you use [04:49] there is no debian folder? [04:49] coppro: I think I need to first apply this diff to the source [04:49] oh, yeah [04:50] from the top, run dpkg-source -x [04:50] that will extract the source and apply the diff [04:51] bingo~ now we're rolling [04:51] ok, now, anything in debian/patches [04:51] lots [04:52] what extension? [04:52] * wgrant points out that we are not Windows. [04:53] yes, but that doesn't mean we can't have file extensions [04:53] * NCommander points out that most patches files do end in .patch or .dpatch ;-) [04:53] coppro: for the patches? there are some .h files and the rest have none [04:54] try running "file [04:54] 00list.UBUNTU: ASCII text [04:55] ok, I'm just gonna go download the source package myself, one second [04:55] :) [04:55] Thank you for helpnig me coppro [04:55] helping even [04:56] no problem [04:56] ok, while I figure out the patch system, you should make a changelog entry [04:56] ok [04:56] from the top-level srcdir, simply run dch [04:57] that should set up a new log entry [04:58] Awsoonn, you may need devscripts [04:58] (and ubuntu-dev-tools has the all helpful update-maintainer script which you may or may not need) [04:58] dch seemed to work [04:58] Cool [04:59] It might be a litlte late for this, but are you working in a clean source tree? [04:59] Or have you already applied your fix to this source tree? [04:59] I have already applied it. :) [04:59] you'll need to undo it [04:59] Ok, that's a bad thing in this case [04:59] * NCommander sees if irssi has a patch system [04:59] it was a one word change, so its no big deal to start over [05:00] NCommander, it's quilt [05:00] Ew [05:00] Ok, you can explain to him the patch system :-) [05:00] Awsoonn: first step on fixing a bug is confirming it [05:00] I'll try [05:00] Awsoonn: that bug is invalid [05:00] oh? [05:01] https://bugs.edge.launchpad.net/ubuntu/+source/irssi/+bug/250681/comments/1 [05:01] how is it invalid? [05:01] Launchpad bug 250681 in irssi "first run help link outdated" [Undecided,Invalid] [05:01] ah [05:01] sweet, [05:02] sooooooo I shall look for a new bug to eat [05:02] I'm sorry guys [05:02] no worries, you've learned something [05:02] which was the goal, after all [05:02] A lot I recon [05:03] Awsoonn, it helps if your working on bugs either to run intrepid, or at least have an intrepid chroot sitting around [05:03] I leared to set up an intrepd chroot [05:03] yes that... [05:03] you should also have an intrepid pbuilder [05:04] something I don't understand about chroots [05:04] sudo pbuilder --create intrepid [05:05] pbuilder is awesome [05:05] if I am running a chroot, can I run X? [05:05] nope [05:05] yes, in theory [05:05] no, actually you can't [05:05] why not? [05:05] because you will already have one X server running on :0 so, the 2nd one would not find it available [05:06] for running X you will need a virtual machine [05:06] virt-manager is nice and easy [05:06] right, but if I close my current X session, then start one in a chroot, it'll work right? [05:06] or you can use that one app that can run X inside X, forget what it's called [05:06] if you have you /dev mounted, yes [05:06] your* [05:07] actually i haven't tried it that far, i tried once to run a graphical app and didn't work, so i move to vm [05:07] so can I run a gnome app from a chroot in my current X session then? [05:07] and now that we have ubuntu-vm-builder is even easier [05:07] Awsoonn: "Sort of" [05:08] :) [05:08] isn't there some X server that runs inside another server though? What's that called? [05:08] Xnest [05:08] * Awsoonn enjoys the constant state of confusion [05:08] coppro: PITA [05:08] that's not the one I was thinking of [05:08] Xephyr [05:08] sounds more like it [05:09] i still call it PITA [05:10] let me see if I understand now, he chroot is only useful for testing CLI apps basicly and if I want to work on Rythmbox for example I need a VM [05:10] or another system [05:10] not necessarily [05:10] there are a few ways around that [05:10] a chroot is usefull for building packages on a clean environment [05:10] like closing down X on hardy and running it from the chroot [05:11] and to be sure you are using clean and standard packages and not 3rth party ones [05:11] yeah [05:11] I don't use it for development [05:11] well, from the chroot, it is by default clean no? [05:11] Awsoonn: Rythmbox is harder -- you want to test the full stack, like DBUS and such, and you can't easily have the Rythmbox in the chroot talking the dbusd on the host [05:11] yes, that's the point [05:12] StevenK: so if I wanted to do that, you would recommend a VM? [05:12] I would recommend setting up a nested X server, but that's just me [05:13] Awsoonn: kvm, if your machine supports it [05:13] <3 kVm :) [05:13] how do I use kvm? [05:13] how do I get udev to work in a chroot? [05:14] coppro: You bindmount /dev instead. [05:14] use virt-manager ? [05:14] Awsoonn: https://help.ubuntu.com/community/KVM [05:14] it's involved [05:14] :0 so am I... [05:14] is there a way to set it up in fstab? [05:14] Awsoonn: but simple once you get the hang of it. We took a week and implemented a system in my workplace to automate the process [05:15] my cpu doesn't suport kvm it seems [05:15] so I'm stuck with vbox methinks [05:15] qemu [05:16] use ubuntu-vim-builder [05:16] just apt-get it [05:16] ubuntu-vm-builder* [05:16] * Awsoonn is getting [05:17] pbuilder finished btw [05:17] is there anything I should do to it? [05:19] what should I do with ubuntu-vm-builder as well? [05:19] read the man page [05:19] https://wiki.ubuntu.com/PbuilderHowto [05:19] awsoonn, also read this: http://wiki.mandriva.com/en/Development/Howto/Chroot#Using_Xnest [05:19] :) Should have seen that comming [05:19] that's an easy way to test graphical stuff from within a chroot [05:20] (replace the urpmi command with "sudo apt-get install xnest") [05:44] * NCommander returns from duty [05:45] any MOTU's around? [05:46] kostmo, I'm not, but if you would like a review on a package, I'm glad to help [05:47] NCommander: you may have seen this one already ;) http://revu.ubuntuwire.com/details.py?upid=2940 [05:47] thanks tho === beuno_ is now known as beuno [05:48] oh yeah [05:48] I already did some reviewing on it [05:48] I revised the package based on another round of comments - just hoping for some official approval [05:49] I have a CDBS-based short debian/rules that doesn't do what I want, can someone take a look please? http://pastebin.ubuntu.com/29204/ [05:49] jmarsden, did you get googlegears to work? [05:50] And uh, what's wrong with it? [05:50] NCommander: I get errors from lintian about the config.sub/guess files being too old. [05:50] googlegears? Was I doing anything with that?? :-) [05:50] er, wrong person [05:50] sorry, a tad slow tonight [05:51] Post the lintian output the -v option [05:54] NCommander: http://pastebin.ubuntu.com/29209/ [05:55] jmarsden, I'm not an expert, but I'm going to surmise that lintian is detected the original config.* from the source tarball, and not anything you've changed [05:56] jmarsden, ask an MOTU, but I'm going to guess your going to need an override [05:56] NCommander: OK... any MOTU's listening who can help? :-) [05:56] * NCommander kicks cody-somerville === beuno_ is now known as beuno [06:01] I'll attach a .debdiff to the relevant bug #236140 and see if that triggers help from someone reading LP bugs... the modified package seems to build and run fine, the bug is fixed... it's just those pasky lintian errors in the way, as far as I can see. [06:01] Launchpad bug 236140 in koverartist "New .desktop file for KoverArtist" [Wishlist,In progress] https://launchpad.net/bugs/236140 [06:16] Anyone here running Ubuntu PPC I can pop a question off of? [06:18] NCommander: Try asking on #ubuntu-powerpc maybe? [06:18] Or even on plain #Ubuntu ? [06:24] NCommander: TheMuso had at least one PPC machine, at least at some point in the near past. Maybe ask him? [06:24] Well, I requested the user post his sources.lists file [06:24] The generated errors suggests it got managed [06:25] * NCommander is working on his five bugs a day ;-) [06:39] StevenK, ping [06:39] NCommander? [06:40] StevenK, care to looking into sponsoring a main package (it replaces a Recommends to a Suggests because the Recommends adds a GNOME library) [06:40] https://bugs.edge.launchpad.net/ubuntu/+source/policykit/+bug/250619 [06:40] Launchpad bug 250619 in policykit "[intrepid] policykit pulls in policykit-gnome due to Recommends" [Undecided,In progress] [06:40] * NCommander has now hugged his five bugs for the day [06:43] \sh: re debescan yes, but IIRC the resolution was to wait a bit if someone comes up to fix it eventually. which didn't happen [06:47] * StevenK tries to figure out why this package runs configure during make install [06:48] It didn't do that for me O_O; [06:48] * NCommander updates his pbuilder jail though [06:49] NCommander: Where this package is one that isn't currently in Ubuntu [06:49] StevenK, oh, I thought you were referring to the bug I linked you to ;-) [06:50] NCommander: Nope, sorry. :-) [06:50] StevenK, upload to your PPA or REVU and I'll look at it with you [06:51] It's sort of on REVU [06:51] "sorta"? [06:51] It's on REVU, but I've updated it a new upstream version and have not uploaded that [06:52] ah [06:52] * NCommander scurries in his kitcehn for food [06:53] * warp10 moins [07:16] * StevenK works it out, kicking CDBS [07:16] makebuilddir -> apply-patches -> configure. I need something between apply-patches and configure [07:18] StevenK, try using the oldstyle rules? [07:18] vs CDBS? [07:18] And, can you do me a quick favor, and see if you can access: http://nemesisnetworks.com/revu? [07:19] NCommander: Loads [07:20] NCommander: Sure, I could rewrite debian/rules. [07:20] Thank you [07:20] Damn CDBS. [07:20] StevenK, want a hand packaging it? [07:20] (if you promise to upload it that is ;-)) [07:20] or advocate on REVU [07:20] I'm pretty sure there's a CDBS magic rule for what you're after. [07:20] So I am, I just can't find it. [07:20] makebuilddir is too early. [07:21] Might be worth poking CDBS's source ... [07:21] There isn't something nice like "post-configure? [07:21] Forcing makebuilddir to run apply-patches makes cdbs do ... odd things [07:21] Heh. [07:21] post-configure is too late [07:21] I've tried that before :) [07:21] * RAOF obviously meant 'pre-configure' :) [07:21] * StevenK chuckles [07:22] what are you trying to do specifically? [07:22] * RAOF guesses: apply a patch to the build system, then touch files so that configure/autogen isn't re-run. [07:22] StevenK, try prebuild, it comes after makebuilddir, but before patches [07:22] er, pre-build [07:22] I need it after patches [07:22] post-patches? [07:23] RAOF: Almost. I need to run ./autogen.sh before ./configure, and one of the patches touches configure.ac [07:23] (And upstream doesn't ship ./configure, damn his black heart) [07:23] Do you need to modify configure.ac before or after ./autogen.sh? [07:23] Before [07:24] StevenK: And ./autogen.sh doesn't call configure? You could just make CDBS call autogen instead of configure. [07:24] SO why can't you use post-patches or update-config? [07:24] Just trying post-patches now. [07:24] pre-configure isn't actually called as a target. [07:24] I got that looking at CDBS's source ;-) [07:24] RAOF: So, you lose. :-P [07:25] Aww. [07:25] Do I win? [07:25] NCommander: See -devel, by the by [07:25] NCommander: Just checking if you do. [07:25] StevenK, I'm looking at gmail [07:25] Don't see anything really of interest that pops out at me [07:26] I meant #ubuntu-devel [07:26] <\sh> siretart: what do you think about the suggestion...if we can fix debescan on ubuntu and using our own resources... [07:26] StevenK, lol, I've been in #d-* too long [07:26] Right, NCommander loses too. [07:26] Man, dh7 is the right way to do an automated source package. [07:27] StevenK, what's wrong with post-patches? [07:27] (what patch system is your package using?) [07:27] StevenK: And autogen doesn't call configure? [07:27] RAOF: Nope. [07:27] RAOF, in my expereince autogen won't do that out of the box [07:27] ./autogen.sh is shipped upstream that runs autoreconf. [07:28] NCommander: Depends on where it's from; GNOME stuff always calls configure. [07:28] I can hack that to run ./configure $@ [07:28] StevenK, what was wrong with post-patches? [07:28]