[00:44] ScottK: Can I just make a straight upload of r-base since it's not a new package (but rather a replacement for the older one) or does that need a freeze exception too? [03:16] YokoZar: It's a new upstream release that's not just bugfixes, so it needs to be included. [03:17] ScottK: so attached to the bug report then [03:17] Yes. === micahg1 is now known as micahg === krusaf|bnc is now known as krusaf === qiyong_ is now known as qiyong === zooko_at_work is now known as zooko` === zooko` is now known as zooko_home === nixternal is now known as handbag_henry === handbag_henry is now known as nixternal [05:59] Can anyone see why lintian gives me this? http://lintian.debian.org/tags/manpage-has-bad-whatis-entry.html on http://paste2.org/p/402746 [06:00] I don't understand what's not right- it looks fine t me [06:23] good morning [06:23] :) just barely [07:18] so can anyone help me fix this manpage? [07:22] lamalex, on line 27 should there be a "."? [07:24] binarymutant: I dont think that would create the warning lintian gives, would it? [07:25] i guess ill try it and see.. [07:26] nope, still get it [07:27] lamalex, no clue, it could be lacking escape characters or something. srry :/ [07:28] binarymutant: no problem [07:31] ahha! got it [07:32] the name should be the binary name, not the application name [07:32] tictactoe instead of Tic Tac To === dholbach_ is now known as dholbach === warner is now known as warner_sleep [11:17] Hi there, Ive got a build error of which I don't know what to do with: http://launchpadlibrarian.net/30927603/buildlog_ubuntu-jaunty-amd64.ika_0.62~21_FAILEDTOBUILD.txt.gz [11:18] It seems the amd64 version wants to use a uint from , how do I prevent that [11:18] ? [11:24] hello everyone. I got a problem with one of my packages: http://launchpadlibrarian.net/30930407/buildlog_ubuntu-intrepid-amd64.mpg123_1.7.2-3ubuntu1~ppa2~intrepid1_FAILEDTOBUILD.txt.gz . here is the file which it refers to: http://pastebin.com/d5534de35 [11:27] does cupsddk replace cupsddk-drivers? [12:07] hi, I need some help for create a .deb (exactly with signing gpg) [12:08] I run dpkg-buildpackage -rfakeroot and get this: [12:08] gpg: [stdin]: clearsign failed: secret key not available [12:09] but gpg --list-secret-keys show me a key [12:09] what I need for signing my .deb? [12:10] shakaran: I think you need to specify which key you want to sign with via the -k option [12:10] shakaran: however, it is usually better to pass the -us -uc options to dpkg-buildpackage in order to skip signing [12:10] and do the signing later on using debsign [12:11] how to use debsign? [12:11] a guide? I search on google? [12:11] it has a manpage I think [12:11] ok, I search now, thanks [12:20] I am doing wrong. [12:20] debsign -k 1DEEE63D myprogram_0.0.1-0ubuntu1.dsc [12:20] debsign: Only a .changes, .dsc or .commands file is allowed as argument! [12:22] shakaran: remove the space between the -k and your keyid [12:23] oh, was easy! thanks again === k0p_ is now known as k0p [12:51] another question, I have in the file rules a command for copy my binary to /usr/bin [12:51] But the command dont work [12:51] cp myapp $(CURDIR)/debian/usr/bin/myapp [12:51] if I put with sudo dont work [12:55] I need more privileges for copy in /usr/bin? [12:59] some help please? [13:00] what do you mean with dont work? [13:01] do you get an error message? [13:04] $ myapp [13:04] bash: /usr/bin/myapp: No such file or directory [13:04] shakaran: where is myapp locatet? [13:04] *located [13:05] I made a shell script for run in /usr/share/myapp/myapp.py [13:05] /usr/bin/myapp invoque to /usr/share/myapp/myapp.py [13:05] shakaran: just copying things under $(CURDIR)/debian is not enough, you need to make sure it is actually getting into the .deb at the end [13:06] shakaran: yeah and where is your shell script myapp currently? [13:06] on myapp-0.0.1/myapp [13:07] shakaran: cp $(CURDIR)/myapp $(CURDIR)/debian/myapp/usr/bin/myapp [13:07] shakaran: look into dh_install to install it properly so thata it ends up in the .deb [13:08] oh, maybe is for $(CURDIR) [13:08] My actual command is: [13:08] cp ./myapp $(CURDIR)/debian/usr/bin/myapp [13:09] shakaran: your commend can't work. $(CURDIR) = myapp-0.0.1 and you have to use debian/myapp/usr/foo because debian/usr doesn't exist [13:10] I do this before: [13:10] mkdir -p $(CURDIR)/debian/usr/bin [13:10] then /usr exists no? [13:11] shakaran: please don't do such weird stuff. For creating folders there is debian/dirs and it won't be in the archive. You want /debian/myapp/usr/bin [13:12] ok, I gonna try [13:15] yeah, it works ;) thanks [13:15] shakaran: installing the shell script should be done by makefile/python install though if it's in the upstreams tarball [13:16] ok === keylocker is now known as leleobhz [14:31] hi folks [14:31] Hey. [14:32] hey sistpoty|work [14:32] hi DktrKranz and iulian === moya_ is now known as moya [14:46] hi, I forgot how to build a source package even if I can sign it (it's a packge from another dude) [14:46] -kyou [14:46] e.g. debuild -klool [14:46] To disable signing, -us -uc [14:46] qnix: Check the dpkg-bp and debuild man pages for the available options [14:47] ha, that's it. I had the -uc, but the -us was missing. [14:47] thanks a lot [14:59] Heya gang [15:00] hi bddebian [15:00] Hi sistpoty|work, congrats! :) [15:00] hm? what are you congratulating me for? [15:01] Oh nevermind, wrong person, I'm not awake yet. I was thinking you just got married but I was confusing you with someone else :) [15:01] heh, well, I know whom you confused me with ;) [15:01] :) === ian_brasil is now known as IAO === ScottK changed the topic of #ubuntu-motu to: Karmic Feature Freeze is in effect now! | Want to get involved with the MOTU? https://wiki.ubuntu.com/MOTU/Contributing | Sponsor queue: http://is.gd/2y76G | http://qa.ubuntuwire.com/ftbfs | http://people.canonical.com/~ubuntu-archive/NBS/ | http://qa.ubuntuwire.com/debcheck [15:04] Is it safe to rebuild the intrepid gcc 4.3.2 package for hardy, and install it ? Or it could be a few packages that could have be built against the old version ? === Tonio__ is now known as Tonio_ === korn_ is now known as c_korn === ripps_ is now known as ripps [16:25] Is there an in-depth guide to python-support (for apps) or a mailing-list available somewhere? [16:33] IntuitiveNipple: internals guide, or "How to implement stuff with pysupport" ? [16:35] I think 'internals' - I've got a package built using pysupport and the deb looks good and it installs fine. The problem is, python-support compiles to /var/lib/python-support/pythonX.Y// and adds symlinks to the source in /usr/share/python-support/, but it isn't linking any non-python supporting files. In this case, it is the application's Glade interface definitions in a sub-directory of one of the module directories. [16:37] it does so for python only files, I see no reason for glade files to be under python directory [16:37] IIRC there's a specific directory for glade files [16:37] I'm wondering if the way to do it is have the glade files installed to /usr/share/ Locations the app expects to find things are hard-coded relative to the app's main() module [16:38] hmm [16:38] indeed! :D [16:39] I've been humming for the last 24 hours trying to find documentation that describes this kind of scenario. [16:39] http://svn.debian.org/viewsvn/collab-maint/deb-maint/python-support/trunk/README?revision=13985 [16:39] these are pysupport internals, but code is easy enough to be fully understandable [16:40] Yes, I've read the README several times but it doesn't provide any guidance for this kind of issue. [16:41] If there were something saying "python-support" enabled packages will separate python and non-python files to different paths, so the application may need to be patched, I at least know what is required :) === Milyardo_ is now known as Milyardo [16:41] and you won't find in pysupport as well [16:42] but wait [16:42] is it a standalone app? [16:42] e.g. no public modules, or extensions? [16:42] define 'standalone' ? It's called "openshot" (a non linear video editor) and it depends on the MLT framework [16:42] it's an application? [16:43] so no, nothing public or extensions and it is an application [16:43] does any other part depend on one of its components? [16:43] I've already added a python-mlt package to 'mlt' which openshot depends on [16:43] (part of Ubuntu archive) [16:43] no, openshot is new. I'm doing the packaging for the very first time [16:44] It depends on several mlt/ffmpeg/etc/ packages that are in my PPA since they need newer upstream releases than are in the archives so far [16:44] good, so it can be self-contained in a private directory [16:44] Yes, it can [16:45] so, things are getting easier [16:45] does it use distutils? [16:45] I've created a distutils setup.py for generic cross-distro installations and added debian/unbuntu specific stuff in debian/openshot.{install,postinstall,postrm} [16:46] yes, it does... that bit and pysuport with CDBS works nicely [16:46] you can instruct distutils to install things under /usr/share/pkgname [16:46] and then provide a symlink in /usr/bin pointing to /usr/share// [16:47] and you're in business! [16:48] hmmm... ok, how to tell distutils the install location when using CDBS? I've been looking at the CDBS web doc and it isn't clear there's a CDBS_ variable for that [16:49] The only one I thought might be it is "DEB_PYTHON_PRIVATE_MODULES_DIRS" [16:49] but the description isn't too precise on its intent [16:49] mh... look at distutils.mk, it has something to pass to setup.py install [16:50] I don't remember variable name, though [16:50] do you mean "python-distutils.mk" ? [16:50] to install a module in a private dir? [16:50] pochu: to install the entire app in a private dir :0 [16:51] try --install-lib [16:51] if you're using CDBS: DEB_PYTHON_INSTALL_ARGS_ALL = --install-lib=usr/share/emesene [16:52] Does the 'lib' affect the fact it is an application in any way? [16:53] looking at python-distutils.mk I think I'd need to set DEB_PYTHON_INSTALL_ARGS_$(cdbs_curpkg) [16:54] because the CDBS script sets DEB_PYTHON_INSTALL_ARGS_ALL already [16:54] e.g.: DEB_PYTHON_INSTALL_ARGS_ALL = --no-compile -O0 --install-layout=deb [16:55] it is DEB_PYTHON_INSTALL_ARGS_ALL += --install-lib=/usr/share/ --install-scripts=/usr/share/ [16:57] ahhh, I see the way you're thinking :) [16:57] I was forgetting about delayed evaluation of the variables in the makefile [16:58] I'm testing the build/install/execution now [16:59] you have to provide symlinks as above, or adjusting sys.oath [17:00] *path === goshawk_ is now known as goshawk [17:10] Fab guys, thanks! the .deb shows the new locations... I just have to alter setup.py now to install the glade files in a sub-dir and it ought to start. [17:19] k zooko_work [17:20] IntuitiveNipple: check if your package has a .private file [17:21] I don't see one listed via Gdebi... let me check in the installed directory [17:25] DktrKranz: No .private file under /usr/share/openshot [17:34] * sistpoty|work heads home... cya [17:45] IntuitiveNipple: it's not in /usr/share, do dpkg -c *.deb | grep private [17:46] DktrKranz: No, there is no 'private' file [17:51] DktrKranz: Thanks for all your help... the app is now partially starting although needs some changes to more path assumption in looking for files. I'll revisit the 'private' indicator another time. I assume that should be set by distutils ? [18:17] I see a pair of tickets which are duplicates. === micahg1 is now known as micahg [18:17] Do I have the power to link them together? [18:17] If not, what would it take to give me that authority? [18:19] zooko, you need to be member of bug-control LP team [18:19] zooko, which requires full understanding of how bugs reports works / priorities etc [18:19] Sounds like a lot of authority. Is there a way to give me such authority but only for bugs related to my projects? [18:19] no you don't [18:20] Projects of which I'm the maintainer? [18:20] anyone can duplicate bugs can't they? [18:20] zooko, I guess yes [18:20] How doI do it? [18:20] I explored the UI but didn't see it. [18:20] james_w, not really sure [18:20] "Mark as duplicate" under the bug title [18:20] just above the package name [18:20] james_w: thanks. [18:21] How can I see what are the other duplicates already makred as duplicated of this ticket? [18:21] at your right side [18:21] "Duplicates" list on the right [18:21] yeah [18:21] Thanks! [18:21] yw [18:22] Here's another question. This package of which I'm the upstream maintainer -- pycryptopp -- was uploaded to REVU, but then it was instead synced from Debian. [18:22] How do I mark the REVU item as "duplicate" of the please-sync ticket? [18:22] what do you mean with REVU item? [18:22] http://revu.ubuntuwire.com/p/pycryptopp [18:23] you should ask to have it archived [18:23] done [18:24] james_w: you archived the REVU page for pycryptopp? [18:24] yeah [18:24] Thanks. [18:24] np [18:28] geser: do you know what changed in the LP API to break getSourcePackage? [18:29] bug 418802? [18:29] Launchpad bug 418802 in python-wadllib "requestsync crashed with ValueError in validate_param_values()" [Critical,Triaged] https://launchpad.net/bugs/418802 [18:30] I think I'm just going to re-upload the older python-wadllib [18:31] ValueError: Invalid value '"Published"' for parameter 'status': valid values are: "Pending", "Published", "Superseded", "Deleted", "Obsolete" [18:31] that error [18:32] is that the same thing? [18:32] yes [18:32] ok [18:32] * Laney grabs the older wadlib then [18:35] james_w: is the old python-launchpadlib also affect by the problem that keeps python-oauth out of main? as it embeds a copy of it (according to a comment on the MIR) [18:36] which version is the most recent unbugged one? [18:36] current - 1 [18:36] makes sense [18:36] * Laney should check timestamps [18:36] well, the old launchpadlib is in main [18:36] was, sorry [18:37] so python-oauth *could* be promoted based on that [18:37] however, ubuntuone also embeds a copy, and so we want to fix that [18:37] but they have some issues with it [18:37] so we're at a bit of an impasse [18:37] can't they continue to use their bundled one? [18:38] not for release [18:38] I mean, if that's what they're currently doing then there's no change right [18:38] I can understand it but isn't the problematic code already in main (as part of python-launchpadlib (the old version as the new doesn't build currently))? [18:38] it's currently in main *twice* [18:38] * Laney has forgotten what he was initially doing [18:39] so having it in as a separate package is fine from that point of view [18:44] and if the code is already in main (as embedded copies) it shouldn't block the MIR of a separate package for it IMHO [18:46] well, the issue is that we probably don't actually want the package in main by release [18:46] does putting it in now make that harder? [18:46] probably not [18:47] How can I find a list of all packages that depend on python-foolscap? [18:47] if the code is that problematic should launchpadlib be moved to universe instead till it's fixed? [18:48] it's not problematic for launchpadlib's usage [18:48] james_w: could you please archive this REVU page, too: http://revu.ubuntuwire.com/p/zfec [18:48] rdepends: python-apport [18:48] geser: ^that's why [18:49] how long after a package is accepted into debian unstable will it make it into the ubuntu repositories [18:49] zooko: it is already archived [18:49] james_w: okay, thanks. [18:49] slayton: currently it won't get in until 10.04 opens for development at the start of November [18:50] currently main has a source version of python-launchpadlib not matching the published binaries, this should be probably also fixed before release [18:50] it will be [18:50] is 10.04 a lts? [18:50] dobey is currently working on an alternative to python-oauth that he hopes to get everyone using in karmic [18:50] Here is the package web page for python-foolscap, and I don't see any mention of other packages that depend on it: http://packages.ubuntu.com/karmic/python-foolscap [18:50] Does that mean that there are none? [18:50] that could well be the way we go [18:51] % apt-cache rdepends python-foolscap [18:51] python-foolscap [18:51] Reverse Depends: [18:51] ipython [18:51] zooko`: ^ [18:52] james_w: thanks! [18:55] geser: ok, I have permission to put it in main and track it for release [18:55] geser: thanks for talking sense in to me :-) [18:57] np :) === zooko` is now known as zookowork [19:32] Hi, I have a question. I've got a failedtobuild error and I don't know how to solve it. -> http://launchpadlibrarian.net/30927603/buildlog_ubuntu-jaunty-amd64.ika_0.62~21_FAILEDTOBUILD.txt.gz [19:32] It's for PPA [19:33] I want to put something on PPA [19:34] and for some reason the program I'm trying to put on doesn't want to be built in amd64 because of a duplicate uint. Problem is, I don't want to use usr/include/sys/types.h at all [19:35] Another thing is, for reason the program itself doesn't end up in the pacakage. [19:37] I suspect getting rid of that (indirect) sys/types.h include would be much harder than getting rid of the conflicting definition in the package, but c++ (or c for that matter) really isn't my strongest language [20:30] !help register [20:30] Sorry, I don't know anything about help register [20:50] crack... hard-coded dependencies against libraries... ffmpeg transition isn't as easy as I thought [20:50] * sistpoty admits to have missed dvdauthor's hardcoded one :( [20:51] dvdstyler even [20:52] btw.: if anyone wants to help out: apt-cache rdepends libavutil-unstripped-49 [20:53] anything in that list needs to get rebuild to pick up the new shlibs dependencies. You can see if it worked out if there are no more -unstripped dependencies in the resulting binary [20:53] instead it should be "... | ...-extra" [20:53] * sistpoty walks the list from top to bottom right now, and I'm at zoneminder [20:54] oh, and if s.th. ftbfs, or the dependencies don't go away, please inform me, I'll add that to the trouble candidates lists [21:13] !help packaging [21:13] Sorry, I don't know anything about help packaging [21:15] !help they're after me! [21:15] Error: I am only a bot, please don't think I'm intelligent :) [21:15] okay :) [21:15] Hi, I have a question. I've got a failedtobuild error when tryng to upload my PPA and I don't know how to solve it. -> http://launchpadlibrarian.net/30927603/buildlog_ubuntu-jaunty-amd64.ika_0.62~21_FAILEDTOBUILD.txt.gz [21:16] For some reason the program I'm trying to put on doesn't want to be built in amd64 because of a duplicate uint. Problem is, I don't want to use usr/include/sys/types.h at all [21:22] sluimers: then don't include it? (or did a dependency include types.h?) [21:23] I don't know how it gets included [21:24] could be a dependency yes [21:25] If I only knew where it was included [21:25] sluimers: I assume looking at ./common/port.h:44 might shed more light? [21:27] It's just a bunch of typedefs [21:28] maybe grep for types.h? === vorian is now known as stevie [21:39] thanks, not sure if it'll work by turning "types.h" into "common/types.h" (because the program I'm trying to upload to the PPA also has a types.h), but I'll test it. === IVBela1 is now known as IVBela [22:04] what does dpkg-source -x do?? [22:05] eXtracts a source package [22:05] ok, so it necessary to do before we do fakeroot debian/rules? [22:07] yes [22:20] ScottK (or others): If I have a package that I uploaded a couple weeks ago but was rejected due to an incomplete copyright file, do I need a feature freeze exception to upload it? [22:20] (I've since fixed it) [22:26] YokoZar: if it was in the new queue before FF, you don't need an FFe [22:26] YokoZar: just make sure to find an archive admin to review it again [23:21] I add a line 'm4_include([m4/intltool.m4])' in aclocal.m4 but after running autogen.sh the line is disappeard. any idea why ? [23:21] I need that m4 file because a Makefile.am uses it [23:22] c_korn: afaik aclocal.m4 is autogenerated [23:22] c_korn: you could try acinclude.m4 though [23:24] sistpoty: this file does not exist. I wonder why this m4 file is not included although it is required [23:26] c_korn: hm... actually on a second thought, the configure.{ac|in} should get this straight, given that you b-d on intltool [23:27] b-d? build-depend ? [23:27] yep [23:28] intltool is installed, yes [23:28] hm... [23:29] is a line like "IT_PROG_INTLTOOL([0.35.0])" required in configure.ac to get it right ? [23:31] c_korn: that macro doesn't sound too wrong to me (maybe except for the version)... but I must admit that I'm a little bit clueless myself right now [23:32] let me look at the autoconf documentation for a second ;) [23:32] hm, I added it. now I get this error http://pastebin.com/d5e22d5c3 [23:32] there is no po directory actually [23:36] but there's a directory with translations? [23:51] I think I messed something up [23:51] all I want to have is some code in Makefile.am to install a gconf schema [23:54] hm... sorry, can't really help you there, haven't done that myself yet [23:55] ok, thanks anyway