[00:49] Is there a simple way find what packages depend on package foo? [00:51] lamalex: apt-cache rdepends package-name [00:51] apt-cache rdepends foo [01:00] ScottK: would it be too late to upload calibre 0.6.17 to karmic, if I were to prepare a debdiff? (0.6.17 fixes numerous bugs present since 0.6.13) === freeflyi1g is now known as freeflying === ripps_ is now known as ripps === TheMuso` is now known as TheMuso === TheMuso` is now known as TheMuso === norly is now known as ejat === lfaraone_ is now known as lfaraone === echidnaman is now known as JontheEchidna [02:13] hello? anyone here? [02:24] I want to fix a bug in a package for karmic [02:34] rmjb: great, go for it [02:34] my question is on the SRU [02:34] thanks for the response lamalex [02:34] anytime [02:35] it's this bug: https://bugs.launchpad.net/ubuntu/+source/backintime/+bug/409130 [02:35] Launchpad bug 409130 in backintime "launching backintime fails to start" [Undecided,Fix committed] [02:35] there's already a fix and branch linked, how do I get that into the karmic package? === TheMuso` is now known as TheMuso [03:07] would it be possible to get the opencore-amr codecs added to backports for various ubuntu versions? (jaunty mainly). or is this totally the wrong channel for asking this? [03:09] i guess i could just get the debs and install them and hope nothing breaks, but some official solution would be better [04:02] <_Andrew> In the .install file is there a way to prevent files from being installed in a package? [04:03] <_Andrew> Maybe something like "usr/include/*\n ^usr/include/*NOTWANTED" ?? [04:04] <_Andrew> So it wouldn't include any files with the name blahblahNOTWANTED ? [04:30] Hi there. My package needs updating from Debian testing. This is the right place, right? [04:33] The package's name is "mandos", and Ubuntu has version 1.0.11, and Debian testing has had 1.0.12 for a while now, which fixes several bugs. [04:42] Teddy_, See the email just posted by ScottK @ https://lists.ubuntu.com/archives/ubuntu-devel-announce/2009-October/000634.html [04:45] philwyett: Hmm, that seems to say that the deadline has *not* been passed. [04:47] Teddy_, Indeed. You can always subscribe to the motu list (https://lists.ubuntu.com/mailman/listinfo/ubuntu-motu) and ask how to get your update pulled. [04:48] philwyett: I was hoping to avoid subscribing to yet another mailing list and just ask here instead... [04:49] Teddy_, Wait for ScottK to appear and he will most likely be able to help you out. [04:49] philwyett: Sounds like a plan. [04:49] lol [04:52] Teddy_, Just to ask being that it is only a minor revision number bump, it is only a bug fix release with no new features and API changes? [04:53] philwyett: Yes, that is correct. 1.0.11 -> 1.0.12 [04:54] Teddy_, Cool, I can see no reason for it not to be pulled. :-) [04:54] philwyett: Me neither; I expected it to be done semi-automatically, but the date keeps creeping closer and I'm getting worried. [04:55] Teddy_, merge-o-matic does pull automatically up to a certain time before release. It's last run was early Sept. [04:57] philwyett: I see that the package entered Debian testing on Oct 1. I though it was less recent than that. Oh well, "missed it by *that* much". [04:59] Teddy_, Seems so. I only know the last run as I am subscribed to way too many lists. ;-) [05:00] Teddy_, To learn about motu, you can read all about it here: https://wiki.ubuntu.com/MOTU === shampavman is now known as wrapster [05:04] Teddy_, This section will be of particular interest https://wiki.ubuntu.com/SyncRequestProcess [05:05] philwyett: Oo, looks interesting [05:08] Teddy_, if it's a bug fixing only, and the package compile/install/run in Ubuntu, it's still possible to have it in KArmic [05:10] fabrice_sp: Yes, so I've come to understand. But the question still remains: How do I get the package to actually update/sync? [05:10] !sync [05:10] Sorry, I don't know anything about sync [05:10] :-/ [05:10] you can use requestsync in Ubuntu [05:11] fabrice_sp: Well, I don't actually *run* Ubuntu... [05:11] oh [05:11] you should :-D [05:11] what is the package name? [05:11] fabrice_sp: I'm happy with Debian. [05:11] fabrice_sp: The package is "mandos". [05:12] ok: I'll test build it, and if it works, I'll send it [05:12] if i have a pkg built only for 32bit.. how do i compile it for 64 as well? eg : libtspi-dev [05:12] where should i be looking and for what? [05:12] fabrice_sp: Oh wow, that would be great! [05:13] is there an upstream changelog somewhere? (just to check that it's a bug fixing only release) [05:13] fabrice_sp: Hmm, w8 [05:14] fabrice_sp: You'll have to use the VC browser: http://bzr.fukt.bsnet.se/loggerhead/mandos/release/changes [05:16] Teddy_, ok. I'll have a look, and in the meantime, I will testbuild/install it [05:18] fabrice_sp: Thanks a lot! [05:18] guys can anyone help me? [05:19] Teddy_, thanks to contribute with this app and Debian packaging :-) [05:19] wrapster, where are you compiling it? [05:20] #302330 the xmail bug in Hardy exist for a long time, and there's somebody have a patch with it [05:21] anyone can help to apply/commit it? :) [05:21] bug #302330 [05:21] Launchpad bug 302330 in xmail "package xmail 1.22-5 failed to install/upgrade: subprocess post-installation script returned error exit status 3" [Undecided,Confirmed] https://launchpad.net/bugs/302330 [05:22] fabrice_sp: on ubuntu 64 [05:22] Teddy_, the installation is ok, but it fails to run in a chroot, with a python exception on dbus. Is it a known issue? [05:23] BlueT_, subscribe Ubuntu Sponsors for Universe to the bug report, and put a debdiff [05:23] fabrice_sp: Well, the mandos server needs an avahi daemon running, which in turn need a d-bus daemon. I guess that could be it? [05:24] wrapster, so do I. How do you compile it? With a pbuilder? [05:25] fabrice_sp: i know that's the best way and I'd really like to do so for a long time :) [05:25] fabrice_sp: but before that... [05:25] fabrice_sp: no [05:25] fabrice_sp: D-Bus communicates over a Unix domain socket, which probably is not connected inside the chroot. [05:25] Teddy_, yes: I get a connection refused to socket /var/run/dbus/system_bus_socket [05:26] im very new to this .. so ,so far what i've been doing is to download the source.. do necessary modifications...(to suit my requirements) then build it... [05:26] fabrice_sp: You need a D-Bus daemon (and an Avahi daemon) running inside the chroot then. [05:26] wrapster, then, how are you building it? [05:26] but as an example ,i want libtspi-dev in 64 but not availbale.. so would like to compile for it. [05:26] Teddy_, shouldn't that be pulled by dependencies? [05:27] fabrice_sp: They should be, and they are, as far as I know. [05:27] wrapster, install a pbuilder, in amd64 flavor, and you will be able to compile it, if you have Ubuntu running the amd64 flavor [05:27] Teddy_, could be a problem of using a chroot. I'll try in a VM [05:28] fabrice_sp: Yeah, that should work fine [05:28] wrapster: it's probably possible to do something screwy with bindmounts to get the system's bus socket to show up inside the chroot (doing that instead of running a second chrooted system bus may or may not make sense) [05:29] wrapster: (specifically: a second bus is a nonissue, but if that seconds bus starts a second networkmanager confusing stuff might happen) [05:29] oh [05:29] err, do I have the wrong nick? [05:29] BlueT_, so, what are you expecting? :-D [05:30] mzz, I think it was for Teddy_ :-) [05:30] thanks [05:30] well, see above [05:30] and for myself (as user :-) ) [05:30] mzz: / fabrice_sp: im just starting off without understanding the issues ... so could you point out to any resource online that i can use to learn the nitty gritty issues of compiling for 64B [05:31] wrapster: are you doing your compiling on a 64 bit system? :) [05:31] mzz: yeah [05:31] i think i know that much at leat :D [05:31] wrapster: if you don't actually have a 64bit system it might be easiest to just use a ppa [05:31] !pbuilder [05:31] pbuilder is a system to easily build packages in a clean chroot environment. To get started with PBuilder, see http://wiki.ubuntu.com/PbuilderHowto [05:31] wrapster, ^ [05:31] wrapster: otherwise I'm a fan of sbuild's lvm support, which is a bit slow but very thorough [05:31] also, that [05:32] fabrice_sp: dunno how much time it would take for the MOTU application, and would like the patch to be applied if there's anyone can do so before me :p [05:32] it rocks with apt-cacher-ng! :-) [05:32] BlueT_, sponsorship is different than MOTU application [05:33] sponsorship is when you you get a debdiff or a patch uploaded by a MOTU [05:34] you have to get 'some' pathees/debdiff sponsored before being able to apply fo MOTUship [05:34] mzz: When is it important to be able to run in a chroot? My application just uses the Avahi libraries; should it be my responsibility to know that it's using D-Bus and provide some workaround for running chroot:ed? [05:34] Teddy_, I would say no [05:34] Teddy_: I wouldn't bother doing anything special unless you know many of your users are going to run it chrooted for some reason [05:34] that's why I have a VM (not updated since a long time, btw) [05:35] fabrice_sp: OK, that's good then. [05:35] mzz: No, there's no particular reason. [05:35] yes: I'm updating it (192 Mb to download) [05:36] Teddy_: especially because frequently you simply *cannot* offer a sane chroot-specific mode [05:36] fabrice_sp: I thought I *had* an Ubuntu image for QEMU, but it seems to have gotten corrupted somehow. I'm downloading installation media now to recreate it. [05:36] Teddy_, ok. It seems updating my vm will take 15 min :-/ [05:37] fabrice_sp: 24 min left to download the DVD here... [05:37] DVD?! [05:37] meh, dvds [05:37] you could have downloaded only the CD :-) [05:37] fabrice_sp: That's what was there for karmic beta [05:37] and no CD? Strange [05:37] then you found a weird download link [05:37] fabrice_sp: http://cdimage.ubuntu.com/releases/karmic/beta/ [05:38] yeah, don't use that, use the regular mirrors [05:38] http://releases.ubuntu.com/releases/9.10/ for example [05:38] mzz: Thanks! [05:39] well, might as well use the dvd if you've already grabbed most of it, I'm assuming it'll work [05:39] mzz: No, only got about 25% [05:40] mzz: 5 min remaining on CD download... [05:40] heh [05:40] BlueT_, the patch has not been seen because it's attached to a duplicate... And u-u-s has not been subscribed [05:41] still 8 minutes to update my vm [05:41] mandos is very qucik to compile, so it should quick after :-) [05:41] fabrice_sp: It'll take ages for me to install a new OS so I think you'll beat me. :) [05:42] I think so :-D [05:42] fabrice_sp: Yeah, the C programs are small and the server is in Python, so yeah, quick to compile. :) [05:43] When using xdm in Ubuntu, the logo used is Debian. It makes people think I'm using Debian and not Ubuntu. If I submitted a fix for it, would it be able to be in the repos or could there be issues in that? [05:59] fabrice_sp: trying to figure what should I do now [06:00] BlueT_, https://wiki.ubuntu.com/MOTU/Contributing [06:01] Preparing new revisions part [06:01] fabrice_sp: checking [06:13] * fabrice_sp is still updating his Karmic VM :-/ [06:14] * Teddy_ is still installing a new Karmic :) [06:14] lol [06:14] ..In a QEMU. [06:15] I'm using Virtualbox for virtualization [06:16] * hyperair updates his karmic installation which isn't in a VM :) [06:16] hyperair: may i PM? [06:16] go ahead [06:17] but i have to go soon so make it quick [06:39] KVM's nice [07:08] Teddy_, it installs fine in a VM. I'll request & ack the sync request [07:09] fabrice_sp: Wonderful. [07:09] fabrice_sp: Thanks a lot! [07:10] yw ;-) [07:23] good morning === YDdraigGoch is now known as Richie [07:47] <_Andrew> Anyone know why "dh_install ... -XCEGUI" includes CEGUI files in my package? === TheMuso` is now known as TheMuso [08:13] morning jono! [08:13] hey highvoltage :) [08:34] hi [08:35] is anyone using karmic with 2.6.31-14 official kernel? well, it gives me kernel panic cuz it's unable to mount rootfs (ext4) 2.6.31-11 works. [08:37] goshawk: seems to work for me [08:37] goshawk: (my root is ext4 in lvm on ide) [08:37] mzz: exactly the same here [08:38] but i've just installed karmic and did the update.. i should investigate more [08:51] goshawk: actually that sounds as if it's not mounting the initrd [08:52] uhm... i can force a initramfs rebuild [08:52] goshawk: do you have enough output when it panics to tell if the initramfs mounted? [08:52] no [08:52] if for whatever grub isn't feeding it that it won't boot at all if you're using lvm [08:52] i just see that line [08:53] i think initramfs is not mounted [08:53] cuz i don't see the usplash logo too [08:53] (which is in initramfs [08:53] I'd boot off something bootable and doublecheck the initrd is in the right place and /boot/grub/grub.cfg makes sense [08:53] i'm on 2.6.31-11 now, which works [08:53] that's weird, unless you're missing the initrd for 14 or grub.cfg is borked [08:54] running update-initramfs -k all -c [08:54] (your symptoms sound exactly right for grub trying to boot without feeding the kernel a working initrd) [08:54] yep [08:54] so i'm rebooting [08:54] and see what wil happen [09:08] mzz: solved [09:08] it works now [09:08] hmm, odd. Wonder how it broke [09:09] me too [09:09] hope i'm a odd case [09:09] time to go for me [09:09] see you === dholbach_ is now known as dholbach [10:51] hello [10:51] how do we request a package removal ? [10:54] joaopinto: file a bug specifying which package and why it should get removed and subscribe ubuntu-archive [10:55] check rdepends, check rbuilddepends, file a bug with any info you find (like no r(build)depends, debian removal bugs) and apply the usual sponsoring (if needed) [10:55] ok, gave up on grnotify [11:01] does it make sense to mark all the "package is broken" bugs as duplicates from the removal request ? [11:01] joaopinto: no [11:02] there's no point tracking bugs on a removed package [11:02] except maybe for sru [11:02] Laney, so what will happen to the open bugs ? [11:02] they'll stay [11:02] and what's the point of keeping bugs for a package which is no longer on the archive ? [11:03] history [11:03] srus [11:03] if it's ever reintroduced [11:04] that may represent introducing infite bugs, aren't closed bugs ketps for history already ? [11:04] also it gives the bug reporter some expectation that someone will work on the bug, which is not the case [11:05] joaopinto: just because you remove the package from the archive, it doesn't automatically disappear from the users systems so it might be good for them to still see the unfixed bugs [11:06] geser, the app does not work, and never worked [11:06] actually on karmic is not even installable [11:06] there's no reason we could not SRU a removed package [11:06] joaopinto: did it work in jaunty? or was it broken there too? [11:07] geser, yes, at least from my test in a chroot [11:07] it was a bad revu package [11:07] ubuntu only [11:08] you know [11:09] this might be good for a revu case study [11:09] anyway, enough effort about a broken package, removal request filled, done [11:09] how did such a broken package get in? [11:09] and why did the packager not take care of it? [11:10] in this case the package==upstream iirc [11:11] which is even more interesting [11:11] and why did it get in, when there are other packages sitting, like mine, pending due to an unclear license, from the review perspective :P [11:50] hi [11:50] is there a way to update only a specific repository? [11:55] walterl: only add that to sources.list? (that might of course hold back a lot of packages, or force removals of packages not in this repository depending on apt's resolver) [11:56] sistpoty|work: in my case i only want to update packages from a ppa. wouldn't removing the other sources cause the rest of the packages to "disappear"? [11:56] walterl: it would mean that apt doesn't have an idea about them, not that these will get removed on your system [11:57] sistpoty|work: k. so there's no way to tell update to ignore repo's? [11:59] walterl: not too sure actually. if it's a different release name, I guess you could use -t as apt-get parameter (however ppa's share the same release name as the distro) [11:59] walterl: maybe apt pinning can do what you want, however I don't have too much clue about that [12:00] * walterl googles [13:12] Hi, I'm about to go through the Packaging guide linked on your motu site and noticed that e.g debhelper version has been changed - willsomeone correct this locked wiki pack to e.g. use compat with 7 ? [13:15] and the hello source version has been change from 2.1.1 to 2.5 [13:15] sorry 2.4 [13:18] and I'm having problems locating postinst and prerm in the original source directory - may so give me an advise ? [13:19] benste, postinst and prerm shouldnt be located into the original source [13:19] they care called maintainer scripts [13:20] * are [13:20] plus I don't think you need them for 'hello' example [13:28] av`: I just came across them cause there are mentioned in the guide that they should be in the source which I copyied from archive.ubunt [13:29] benste, orig source contains upstream files only [13:29] benste, and debian/ dir contains maintainer scripts [13:30] if you did a dh_make all maintainer scripts got created [13:30] it's now up to you to keep the ones you need and remove the one you don't need [13:30] :-) now I now what you wanted to say me:-) - could you take a look at https://wiki.ubuntu.com/PackagingGuide/Basic#postinst%20and%20prerm [13:31] av`: in my selfmade /debian dir - maintainer scripts - I already have those [13:31] but now I should copy pre and post files but they don't exist [13:31] benste, maybe they got removed [13:31] possibly, so I can skip them? [13:31] - who'll update the wiki page ? [13:32] benste, to know that you should have a look at changelog [13:32] benste, the documents does not refer to a prristine source dir, it refers to a debian source, which is the original source + building diff [13:32] you should see if those files are mentioned and if yes why [13:33] joaopinto: what's prristine ? [13:33] benste, * Removed prerm and postinst, as info files are missing now. [13:33] benste, they got removed, the info file is no more there, then you can skip them [13:33] hum, maybe I was reading it wrong :P [13:34] benste: pristine (in this context) = unmodified; as you can download it from the upstream webpage [13:34] av`: so may someone be so kind to correct the guide for newbees like me ? [13:34] geser: joaopinto - thanks [13:34] benste, still, it was refering to copy from a debian source, not from a source [13:35] there is no debian/* on a source tarball :) [13:35] benste, I dunno who has the rights to change it :) [13:35] benste, and that page looks a bit outdated, yes [13:35] joaopinto: there is a hello_2.4-debhelper [13:35] dir in it which has maintainer scripts [13:35] joaopinto, depends [13:35] erm, except native packages [13:35] joaopinto, upstream may want to include it [13:35] which are not usuaul cases [13:36] which is way to bad [13:36] av`, that is a bad thing to do :P [13:36] yep^^ [13:36] but it happens [13:36] so usually it should have the -debhelper dir ? [13:36] no [13:37] sorry I forogt the not :-) [13:37] usually should have nothing debian-related [13:37] you should add debian-related stuff in it [13:37] it = source tree [13:37] to the original source, but also to the one I got through APT ? [13:37] benste, hello-debhelper_2.2.orig.tar.gz does not contain debian/* [13:38] benste, if you do apt-get source foo [13:38] benste, you are looking into a debian source, which is the orig source withe the building diff applied [13:38] benste, it will contain debian dir in it, yes [13:38] benste, apt-get source downloads orig and applies the diff.gz [13:38] av`: thanks [13:38] np :) [13:41] where will debuild -S look for my PGP key ? [13:42] benste, have you set DEBEMAIL ? [13:42] benste, it will verify if your secret key is available [13:42] it will use the standard gpg sign [13:43] joaopinto: I've set it to MYadress+motu@gmail.com [13:43] benste, if it fails either set a correct email with DEBEMAIL or use -kKEYID [13:43] benste, and did you create the GPG key for it ? [13:43] benste, try debuild -S -kYOURKEYIDHERE [13:43] sistpoty|work: thanks for your help :) [13:43] joaopinto: I've got a pgpkey for MYadress@gmail.com which includes some sub keys with the + [13:43] including a motu one [13:43] ok [13:44] possibly it can't find my gpg key cause of the sub key ? [13:44] I#ll try to search my key from CLI with gpg - usually I use seahorse [13:45] benste, yeah, maybe it's unable to use your subkey, I don't know, alwais worked for me as main key [13:46] benste, otherwise use the -kKEYID option and you are done [13:46] benste, does the email address shows at gpg --list-secret-keys ? [13:47] joaopinto: yes it does, including the right commend and name [13:48] av`: 91E4A5BE is key id ? [13:51] av`: Enter passphrase: gpg: gpg-agent is not available in this session - does this encounter any problems - with keyid it woun't accept my right passphrase [13:52] strange 5th try - accepted :-) [13:55] av`: - next question I'll read the whole paragraph before asking :-) [14:08] can someone give me advise about Bug #449349 ? is it right to request a sync at this stage in the release cycle? [14:09] Launchpad bug 449349 in bash-completion "regression for completing remote files/dirs over ssh" [Medium,Confirmed] https://launchpad.net/bugs/449349 [14:15] jbernard, I would decided that based on what was changed on the debian version [14:16] if there are non trivial changes, a specific bug fix patch would be safer [14:24] what's wrong with the first paragraph of the following changelog? - http://paste.ubuntu.com/294695/ [14:24] debuild -S woun't work cause of this [14:27] benste: the fact that it's not built for an ubuntu release? [14:27] (ie, unstable --> karmic) [14:27] i'm assuming, anyway, without the error [14:33] Hobbsee: but it worked using unstable with only the lower one [14:34] Hobbsee: http://paste.ubuntu.com/294706/ [14:34] still not working [14:36] benste: looks like your trailer line is badly formatted? (try dch -i and check the difference between the generated trailer line and yours) [14:39] sistpoty|work: I added the upper paragraph so I know the dif - am I allowed to change the changelog manually with nano ? [14:40] benste: sure, if you're careful [14:41] sistpoty|work: I guess I wasn't - I wanted to add a new entry in the exact formatting, with only one point [14:41] looks like it didn't work [14:41] sistpoty|work: - I changed the revision number, does this take an effect ?P [14:42] benste: yes, it will change the version. [14:44] checked again and looks like ; was missing in header - now working [14:45] sistpoty|work: do you know who I can solve this error "Enter passphrase: gpg: gpg-agent is not available in this session" so that I can use debuild without -kKEYID ? [14:46] benste: you can set DEBSIGN_KEYID in your ~/.devscripts.conf (or /etc/devscripts.conf) [14:46] sistpoty|work: not in .bashrc ? [14:47] benste: it's not an environment variable, but rather a config option of devscripts [14:49] can't I simply import my gpg private key or set seahorse as keyring in # A colon separated list of extra keyrings to read. [14:49] # DSCVERIFY_KEYRINGS="" [14:49] sistpoty|work: [14:59] benste: why would you want to do this? [15:04] sistpoty|work: for not entering the keys the whole time and auto signing with the key mentioned in the feils ? [15:04] files [15:04] benste1: how about using a gpg-agent? [15:05] I'm using seahorse, but it looks like there is a problem between seahorse and gpg which ned up not finding my private key [15:05] - but strangely it's listed with gpg --list-private-keys [15:05] or similar [15:09] benste1: but it does work with -k? [15:09] (using debisgn, i.e) [15:10] sistpoty|work: with -k it works fine [15:10] - with debuild - guess it's the parent app isn't it ? [15:10] benste1: then you can just set DEBSIGN_KEYID [15:11] and if I want to sign with anotehr key ? [15:11] sistpoty|work: I'll set up your solution looks like it's the easiest atm [15:12] benste1: I'm not too sure, but I guess that -k will override it then [15:12] simply adding at the first line of /etc/devscripts.cond [15:12] f [15:12] ? [15:13] benste1: no you have to place it exactly at line 42 :P [15:13] (just kidding) [15:14] :-) [15:19] sistpoty|work: looks like it worked, strange thing is that I still didn't get a .deb using pbuilder too [15:19] but I'll go on the Guide first , so maybe at the end .. - thanks for helping [15:19] benste1, check /var/cache/pbuilder/result [15:21] av`: lol my first deb - why wasn't it copied to the curent dir when building it ? [15:21] benste1, cause pbuilder uses that dir as default [15:22] benste1, if you give a dpkg-buildpackage you'll find it on the current dir [15:23] av`: but than it's build withouht fakeroot , on the installed system ? [15:24] strange - there is only rev 1 and not 2 [15:29] benste1, it's built with fakeroot on installed system as well [15:29] if not some targets on debian/rules would fail [15:30] benste1, dh_testroot tests if you are running it as fakeroot [15:30] if not fails [16:23] bdrung: imho eclipse failed on the buildds, because only i386 builds the arch:all package [16:25] sistpoty|work: that's a good point. thx [16:25] bdrung: I'm not too sure about the actual debian/rules invocation on the buildds, but I think it's binary-arch for non-i386, so common-install-indep might not be executed (?) [16:27] it's binary on the i386 and binary-arch on all others (see a build log of your choice) [16:31] thanks geser [16:52] eh, if libpurple-dev depends on libpurple0 (>= ${source:Version}) lintian claims: E: pidgin source: weak-library-dev-dependency libpurple-dev on libpurple0 (>= ${source:Version}) ; but if it depends on libpurple0 (= ${source:Version}) lintian __additionally_ claims: E: pidgin source: not-binnmuable-all-depends-any libpurple-dev -> libpurple0 [16:57] c_korn: is libpurple-dev or libpurple0 arch:all? [16:59] sistpoty|work: libpurple-dev is arch:all [16:59] c_korn: it should be arch:any, I guess (static library in there?) [17:00] c_korn: otherwise you'd need binary:Version instead of source:Version [17:00] (though we don't really have binNMU's here... yet) [17:02] sistpoty|work: eh, you are right. http://packages.ubuntu.com/karmic/all/libpurple-dev/filelist it has a shared library in it but it is arch:all. this is a bug [17:03] * c_korn wonders it has not been detected all these years [17:03] libpurple is not that old :P [17:04] c_korn: I guess that's just the symlink, so it could in theory be arch:all [17:04] c_korn: the .so are pretty sure just symlinks to the real lib [17:04] c_korn: makes me wonder though, why there isn't a static library [17:05] why should there? static linking is not really liked [17:07] it can be useful from time to time, so why deprive users of this ability? [17:07] geser, that's not a general purpose "dislike" [17:08] geser: you are right. they are just symlinks [17:09] (and the library packaging guide still mentions it, however the guide also still mentions libtool files, which shouldn't be in there any longer) [17:10] even libpurple-dev in lenny has no .a file [17:11] hm, maybe the upstream build system only builds a shared library? [17:12] anyway it's not like the world ends if there's no static library, it's just a nice to have [17:12] sistpoty|work: Actually Debian is pushing to have them removed. [17:12] ScottK: oh? hm... :( [17:13] There was a thread on debian-devel about it. [17:21] with libpurple0 (>= ${binary:Version}) the error is: E: pidgin source: weak-library-dev-dependency libpurple-dev on libpurple0 (>= ${binary:Version}) [17:22] with libpurple0 (= ${binary:Version}) the error is: E: pidgin source: not-binnmuable-all-depends-any libpurple-dev -> libpurple0 [17:23] c_korn: oh, yeah, my mistake. arch:all aren't part of a binNMU, so the version of libpurple-dev's dependency isn't increased [17:25] exactly, how is situation being solved then ? [17:25] making the -dev arch:any ? [17:25] yes [17:26]