[00:09] Is this an appropriate place to ask for help on azureus from karmic giving me an error on "debuild -S" of: "You must specify a valid JAVA_HOME or JAVACMD!" Google hasn't helped. [00:10] Impressively queit for 220 people. [00:15] persia, TheMuso: ping [00:16] Darxus: do you have a java jdk installed ? [00:18] Darxus: you could go to #debian-java and ask blackxored [00:19] hi andv [00:19] Huh. Maybe not. Thanks. [00:20] Nope, I do have the jdk installed. [00:36] However, I had not done "apt-get build-deps azureus" :/ [00:36] Er, build-dep. [01:35] Uploaded my first debdiff and subscribed ubuntu-universe-sponsors! [01:37] Darxus: what bug ? [01:44] c_korn: https://bugs.launchpad.net/ubuntu/+source/azureus/+bug/428514 [01:44] Launchpad bug 428514 in azureus "Azureus crashes both openjdk and sun java." [Undecided,New] [01:46] It's a bug in both java implimentations, but my patch avoids the bug by using a command line option to skip JIT compiling of the problematic part. [01:47] Darxus: does the current package in karmic have a patch system ? [01:48] (run what-patch in the source tree) [01:49] c_korn: Thanks, I was wondering how to tell - the relevant docs don't say. [01:50] Darxus: oh, forget about it [01:50] c_korn: ? [01:50] I thought the wrapper was outside the debian directory [01:50] so you do not need a patch anyway [01:51] but your changelog file is wrong. [01:51] :/ [01:51] How? [01:51] 1. explain what you change and why you did it (not only tell it fixes a bug) [01:51] Ah. [01:52] I was trying to mimic other changelog entries. [01:52] 2. Use (LP: #428514) to indicare that the bug was closed [01:52] the Closes statement is used by the debian BTS [01:53] Just those two changes to the changelog? [01:54] lintian may also complain that Darxus is not your full name (not sure about this one. it does at least complain if the maintainer is not a full name) [01:54] I saw mention that lintian was running. [01:54] With no error. [01:55] That's an annoying requirement. [01:56] -XX:CompileCommand=exclude,com/aelitis/net/udp/uc/impl/PRUDPPacketHandlerImpl\$5,runSupport [01:57] How should I refer to that? Something other than "Commandline to prevent compliation of problematic thing." [01:58] Darxus: why does the application crash without that line ? [02:03] c_korn: It's a bug in the java implimentations. [02:04] When they try to JIT compile that bit, they crash. [02:04] It's not a bug in azureus. [02:04] It's not azureus that crashes. It's java. [02:05] Darxus: ok, I do not know how detailed the motus want to have the changelog entry. [02:05] but it should at least mention that. [02:07] Mention that it's a bug in java? [02:07] What I want is a word to replace the word "thing". [02:07] What is "com/aelitis/net/udp/uc/impl/PRUDPPacketHandlerImpl$5"? [02:11] Workaround to prevent a crash caused by a bad java implementation. The workaround [02:11] grr, pressed enter too early. [02:12] so [02:12] wearing a MOTU hat [02:12] I'd want to see a bug report linked there [02:12] much more useful than handwaving [02:12] e.g. [02:12] 'Workaround JIT crash in openJDK6 ' [02:13] Looks like the word I was looking for was "class". [02:14] Darxus: the crash is not limited to openjdk6 but also occurs in suns implementation, does it ? [02:14] lifeless: I submitted a bug report to sun, they didn't reply. I also submitted a bug against open jdk (which shares the problem), and that's linked to in the bug that I'm going to mention as "(LP: #428514)". [02:14] Okay? [02:14] openjdk is suns implementation :) [02:14] c_korn: Yes. [02:14] Um. [02:14] bug 428514 [02:14] Launchpad bug 428514 in azureus "Azureus crashes both openjdk and sun java." [Undecided,New] https://launchpad.net/bugs/428514 [02:15] lifeless: you know what I mean :) === a1g_ is now known as a1g [02:17] It crashes the java in both packages openjdk-6-jre and sun-java6-jre. [02:17] ok [02:18] so bug 380378 claims that sun-java6-jre doesn't crash it? [02:18] is that a stale comment or something? [02:18] Launchpad bug 380378 in openjdk-6 "vuze crashes after upgrade on ubuntu 8.10 -> 9.04" [Undecided,New] https://launchpad.net/bugs/380378 [02:19] ok, so good work on the bug report in openjdk [02:19] I think you may want a similar one on sun-java6-jre [02:19] and they probably need to be forwarded upstream, but that is a separate problem [02:19] [one I encourage you to tackle] [02:20] lifeless: Does it actually say that (that sun-java6-jre doesn't crash it)? [02:21] I think it says I how I tried switching to sun-java6-jre, but not that I ever confirmed it worked. Because it didn't. [02:21] Darxus: https://bugs.edge.launchpad.net/ubuntu/+source/openjdk-6/+bug/380378/comments/3 [02:21] Launchpad bug 380378 in openjdk-6 "vuze crashes after upgrade on ubuntu 8.10 -> 9.04" [Undecided,New] [02:22] lifeless: Ah yes. I don't know. They both crashed on me. [02:22] RoAkSoAx: Not yet. Sorry. [02:25] lifeless: I did submit a bug directly to sun via the url included in the crash log. [02:25] I'll add a bug against the sun-java6-jre package. [02:27] Darxus: did you get a url for the sun bug? [02:27] or was it some fire and forget thing ? [02:28] lifeless: I got nothing from them :( [02:28] the main point is to let someone else pick it up and run with it later [02:28] so if there is no url then they have to find you and ask to figure out whats going on [02:29] Yeah, well, I don't even have a way to follow up on it with Sun. [02:29] Somebody in #bittorrent said you basically need a contract with Sun for that kind of support :/ [02:34] what url did you file the ticket at ? [02:35] lifeless: What ticket? [02:35] the one at sun that you got no response for [02:35] # If you would like to submit a bug report, please visit: [02:35] # http://java.sun.com/webapps/bugreport/crash.jsp [02:35] I submitted it there. [02:36] It's in all java crash logs. [02:36] Huh, azureus does use quilt. [02:42] Finally uploaded a debdiff with the improved changelog entry. [02:43] did you end up at http://bugreport.sun.com/bugreport/start_form.jsp ? [02:48] lifeless: No, I clicked I am an end user, not a developer, so I ended up at http://bugreport.sun.com/bugreport/hotspot_enduser_form.jsp [02:50] if i have added a ppa repos from launchpad, is it possible to remove that and go back to the package versions you had before updating? [02:51] nicklas_: Yes. Everything is possible. [02:51] Darxus: ok; so is there perhaps a bug report in the azureus bugtracker? [02:52] lifeless: Perhaps. [02:52] or that the package on that you plan to forward upstream? [02:52] Darxus: you dont happen to know how? [02:52] nicklas_: I don't think this is an appropriate channel. [02:52] Darxus: oh [02:52] nicklas_: I could be wrong. Download the .deb from the archive and dpkg -i it. === zooko` is now known as zooko [02:53] lifeless: What? [02:53] Darxus: ouch, then you have to downgrade every package manually :-S [02:56] nicklas_: I didn't say that. Check out apt pinning. I think you should be able to coax it into doing a massive downgrade to only what's on the official archives. [03:01] Good evening folks. (UTC-6) [03:03] How do I add a package that a bug applies to on launchpad? [03:04] Doh, found it. [03:06] "There is no bug supervisor for Sun Java. This means that there is nobody upstream we can notify about this issue." hah [03:13] Oh, hmm, I need an "also affects package" not "also affects project". [03:14] Ah, the "also affects distribution" link works. [03:17] So I added sun-java6 to the openjdk bug I created, attached the sun java crash log, and then decided that was totally the wrong thing to do :/ [03:17] But It's my bed time. [03:17] Thanks for your help. [04:05] ( . Y . ) === jdong_ is now known as jdong === mrpouit is now known as mr_pouit [06:39] good morning [06:43] hi dholbach ! [06:44] hi mruiz [06:47] I'm getting the following error during the building process: configure.ac:1335: error: AC_SUBST: `libdir exec_prefix prefix' is not a valid shell variable name . Ideas ? === micahg1 is now known as micahg === micahg1 is now known as micahg [07:03] Can anyone help me debug this error: [07:03] dh_builddeb [07:03] dpkg-deb --build debian/freeswitch .. [07:03] dpkg-deb: building package `freeswitch' in `../freeswitch_1.0.4+repack4-0ubuntu14925.3_amd64.deb'. [07:03] dpkg-deb: conffile `/opt/freeswitch/conf/vars.xml' does not appear in package [07:03] dh_builddeb: command returned error code 512 [07:06] quentusrex: You still should have used pastebin - with the whole log [07:06] ;) [07:09] Hi dholbach. Hi mruiz [07:09] fabrice_sp, hi [07:09] hey fabrice_sp! [07:09] mruiz, did you tried to regenerate the autoconf files? [07:09] fabrice_sp, can you guide me ? [07:10] run aclocal, autoconf and automake [07:10] fabrice_sp, under the src directoy ? [07:10] MTecknology: there is a simple explaination... [07:10] the rest of the build process works... [07:11] part of the builddeb isn't finding something... [07:11] quentusrex, do you try to install a conf file? [07:11] quentusrex: details help though :) - I can't really help you - just suggestions [07:11] mruiz, in the root directory of the package (in general) [07:14] fabrice_sp, http://paste.ubuntu.com/275687/ [07:14] yes, I'm trying to install conffiles. [07:14] I have that particular conf file listed in both the *.install file and the .conffile [07:15] quentusrex, you should have it only in .conffile [07:16] mruiz, it seems like a broken autoconf... It sometimes happens [07:16] thanks fabrice_sp [07:17] yw [07:18] fabrice_sp, then what I should do ? [07:19] mruiz, directly patch the configure.ac file... Do upstream provide an update? [07:20] sorry: no better idea [07:21] have to run... Bye [07:38] What is the largest package you guys can remember seeing? [07:38] debian/ubuntu package... [07:39] 300MB? [07:49] <_ruben> quentusrex: -rw-r--r-- 1 root root 595M 2009-08-07 13:05 ./universe/i/ia32-libs/ia32-libs_2.2ubuntu18.1.tar.gz [07:54] _ruben: I'll have that beat soon... [07:55] I'm at 450M with only 3 of the language prompts packaged, and there are 12 language sets [07:55] for this phone system... [07:55] I'm actually working hard to get it built into much smaller packages... [07:55] <_ruben> jikes [07:56] about 105M per language set... [07:56] plus about 75M for music on hold... [07:56] check out 'freeswitch' [07:56] <_ruben> wouldnt one lang per package not be nicer? i doubt most ppl would need more than 1 or 2 languages [07:56] it's a great telephony system. [07:56] <_ruben> i know it by name [07:56] awesome... [07:56] do you use it? [07:57] <_ruben> we use asterisk at work [07:57] aah [07:57] does that work for you? [07:57] <_ruben> usually ;) [07:58] <_ruben> its configs are nasty though [07:58] take a look at the freeswitch configs... [07:58] <_ruben> luckily i dont have to mess with them that often (mosty done by a cow-orker) [07:58] lol [07:58] any idea what could cause this packaging issue? [07:59] dh_builddeb [07:59] dpkg-deb --build debian/freeswitch .. [07:59] dpkg-deb: building package `freeswitch' in `../freeswitch_1.0.4+repack4-0ubuntu14925.3_amd64.deb'. [07:59] dpkg-deb: conffile `/opt/freeswitch/conf/vars.xml' does not appear in package [07:59] I've tried searching for the issue online... [07:59] but haven't found any possible solutions... [07:59] I know the file is in debian/tmp/opt/freeswitch/conf/vars.xml [08:00] but can't figure out why it is refering to /opt/.... [08:00] <_ruben> relative versus absolute paths was one the things i was gonna mention [08:01] yeah... but I don't see anything pointing to that... [08:01] I just did a: grep -r "/opt/freeswitch" ./ [08:01] and it turned up nothing... [08:01] <_ruben> probably make install isnt honoring INSTALLDIR variable properly or something similar [08:01] hmm [08:02] well, there is nothing in /opt/ [08:02] on my build machine... [08:03] I g2g [08:03] 'night [08:03] I'll check back in the morning if you happen to think of something. [08:06] <_ruben> g'night [08:28] persia, TheMuso: ping === dholbach_ is now known as dholbach === asac_ is now known as asac [10:02] ttx is doing patch/package reviews in #ubuntu-reviews [10:05] dholbach: can you help me joining ubuntu-universe-sponsors? [10:05] If a MOTU Release member has the time to spare, Bug #434078 needs another m-r ACK [10:05] Launchpad bug 434078 in gmameui "[FFE] Karmic: Please sync gmameui 0.2.11-2 with Debian unstable" [Undecided,New] https://launchpad.net/bugs/434078 [10:06] bdrung: no, I'm sorry - I'm not an admin [10:06] persia, themuso: uus membership of bdrung please? :) [10:06] bdrung: you can still review and sponsors stuff, it's just that you can't unsubscribe the team [10:07] bdrung: if you need somebody to unsubscribe uus from a but, please just yell in here [10:07] (for now) [10:09] dholbach: then bug 390580, 402174, 432564 [10:09] Launchpad bug 390580 in gwget2 "[FFe] Please update gwget to 1.0.2 version" [Wishlist,Fix released] https://launchpad.net/bugs/390580 [10:09] Launchpad bug 402174 in fsl "fileno implicitly converted to pointer" [High,Fix released] https://launchpad.net/bugs/402174 [10:09] Launchpad bug 432564 in openerp-server "Please sync openerp-server 5.0.3-0-2 (universe) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/432564 [10:10] bdrung: done - but usually you don't need to unsubscribe once they're uploaded [10:11] I mostly just unsubscribe when I say something like "patch is not ready, unsubscribing sponsors" [10:11] ... "please resubscribe when ready" :) [10:11] or for sync requests where there's nothing left to do for us [10:12] dholbach: yes, i know, but there are open debian / upstream bugs [10:12] that's why I use http://people.canonical.com/~dholbach/sponsoring/ :-) [10:13] it uses https://bugs.launchpad.net/ubuntu/+bugs?field.subscriber=ubuntu-main-sponsors so you don't get open upstream tasks :) [10:20] hi [10:21] so i've got these patches for bug #410267 and i'd like to learn to apply them the correct way so I can republish the package on my PPA [10:21] Launchpad bug 410267 in xorg-server "Pressure sensitive tablet: stylus paints off-cursor in Gimp and Inkscape" [Medium,Invalid] https://launchpad.net/bugs/410267 [10:22] if i remember correctly I shouldn't touch the original sources [10:23] dholbach: do i need a ffe for bug 430658? [10:23] Launchpad bug 430658 in lintian "Please merge lintian 2.2.16 (main) from Debian unstable (main)." [Wishlist,New] https://launchpad.net/bugs/430658 [10:23] should I paste the patches at the end of the .diff file ? [10:27] bdrung: try just asking in #ubuntu-devel - somebody from the release team should be able to tell you [10:33] strider: usually you'd just add the patch in debian/patches [10:33] strider: what's the name of the source package you need to patch? [10:34] libgtk2.0-0 [10:35] 27M, gonna take a while to dl... [10:35] dholbach: no response there === ropetin_ is now known as ropetin [10:38] then ping specific members of ~ubuntu-release :) [10:39] what's wrong with just subscribing them to the bug and asking for input? [10:40] that works too [10:40] would be a solution, too [10:40] I see a lot of *-release pinging that probably isn't so urgent [10:40] takes people away from what they are doing === TheMuso` is now known as TheMuso [10:42] strider: there's a patches directory in debian/, sot it should be simple [10:44] strider: the packaging uses quilt, so it should be straightforward. Just apt-get source libgtk2.0-0, then add the patches with quilt, dch -i update the changelog, then debuild/lintian/pbuilder and test [10:46] strider: https://wiki.ubuntu.com/PackagingGuide/PatchSystems#quilt%20(example%20package:%20xterm) [10:47] ok thanks i will try this [10:48] strider: document the patches afterwards, there's a spec for Ubuntu, that uses different keys than Debian, but I can't get a hold on it atm, forgot to bookmark it [11:00] strider: https://wiki.ubuntu.com/UbuntuDevelopment/PatchTaggingGuidelines for documenting the patches, and ping me if you need help at any point [11:16] ok, changelog updated, patches documented, trying to build the package === doko_ is now known as doko [11:47] TheMuso: thanks [11:48] np [11:49] is loic-m_ same as lool? [11:50] nope, whois would have told you ;) [11:51] well, it is never accurate... and lool is loic minier... sorry to ask. I hope I didn't ofend any of you :D [11:52] and I usually check via /nickserv info... === andv is now known as av` [12:51] reading from launchpad : For example: you're creating an experimental version of the myapp_1.0-1 package. Your PPA package would be named myapp_1.0-2~ppa1. [12:52] do i only have to rename the source folder or is there a file to edit ? [12:57] just add a new changelog entry to debian/changelog describing your changes [12:57] you can use "dch" (from devscripts) to do it (dch helps with the formating) [13:00] yes that's done [13:00] i change the version number here ? [13:00] strider: dch -i will put 2.17.11-1ubuntu2 for you, you can modify it to 2.17.11-1ubuntu2~ppa1 for example [13:00] yes, the version number of the package is the top-most entry in debian/changelog [13:01] ok got it [13:32] on lp, can i change the bug status and comment in one step? [13:33] ah, never mind, found it === jrib1 is now known as jrib === jrib is now known as Guest27195 === Guest27195 is now known as jrib [13:46] strider: your patches look good. The changelog entry shouldn't have "New upstream bugfix release:" though, since there's no new upstream tarball [13:47] oh right [13:47] it's a bit late now that the package's uploaded on Launchpad but i'll keep that in mind [13:47] you don't need the same Closes: 3 times, just put it on the first line decribing the fact you added the 3 patches [13:48] strider: don't you want to get it in the repos? [13:48] I just noticed that https://help.ubuntu.com/community has tabs for 7.04 and 7.10, but not 8.10 and 9.04. Who should I report this to? [13:48] well yes it would be even better [13:49] strider: debdiff gtk+2.0_2.17.11-1ubuntu1.dsc gtk+2.0_2.17.11-1ubuntu2.dsc > gtk+2.0_2.17.11-1ubuntu2.debdiff and you attach that to the bug report (just fix the changelog first, and be sure not to have ~ppa1 in the changelog) [13:50] jdstrand: #ubuntu-doc [13:51] slytherin: thanks [13:51] does this look ok ? : http://paste.ubuntu.com/275841/ [13:52] Closes: #410267 > (LP: #410267) AFAIK [13:53] Rest looks good, but 080 patch name ends without finishing the sentence... if-? [13:55] l. 2, maybe describe more what the 3 patches are for, like "Fix off-cursor position in gtk applications when using a Tablet" or shtg like that [13:55] yes i noticed that too, Gnome's bugzilla might limit the number of characters in an attachment [13:57] maybe Always_report_XI_events_to_the_grab_window_if_any ? [14:03] NCommander, that was a brief appearance! [14:04] directhex, it was a wrong channel thing [14:04] tsk [14:15] loic-m_, does this look ok ? http://strycore.com/files/gtk+2.0_2.17.11-1ubuntu2.debdiff [14:16] oh wait i forgot to save my new changelog [14:18] here, fixed [14:19] can I get a list of packages I've sponsored easily? === porthose is now known as porthose|afk [14:28] bdrung_: in what authority are you ACKing bug 432564 ? [14:28] Launchpad bug 432564 in openerp-server "Please sync openerp-server 5.0.3-0-2 (universe) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/432564 [14:29] as a MOTU? [14:29] MOTU [14:29] Joined on 2009-09-10 [14:30] Riddell: yes, as motu :) [14:30] strider: sorry, disconnection. Looking at your new debdiff, ll. 6, 13 trailing spaces, l. 11 : instead of / ; but more important l. 8 remove completely "Closes: #410267 > ", sorry I wasn't clear earlier [14:31] strider: "Closes" is IIRC for Debian BTS [14:31] bdrung_: and you're checking that such syncs don't have any new features? [14:32] closes: is debian BTS [14:32] lp: is launchpad [14:32] strider: loic-m_: 'Closes: #xxx' is Debian format. for Ubuntu 'LP: #xxx' is the correct format. [14:32] strider: and ll. 7, 12, 14, Like/Block better not uppercase [14:32] Riddell: yes, it is only a -1 -> -2 [14:32] anyone from motu-release around? [14:34] strider: btw the patch work, I'll comment on the bug report too when you'll put the debdiff. Thanks a lot! [14:34] bdrung_: groovy, thanks [14:41] Well Carlos Garnacho did all the hard work, he's the one who should be thanked :) [14:42] strider: indeed, but porting it to Ubuntu is a good thing too [14:42] hyperair: what's up? [14:45] yes and I learned a few useful things [14:57] sistpoty|work: not you. you already acked remuco ;-) === cprov is now known as cprov-afk [15:10] * RainCT wonders why the archive rebuild pages have Colin's face on the top :P [15:12] RainCT: he watches us :) [15:12] it's because colin is the copy archive owner [15:12] RainCT: Because he requested the rebuild? [15:12] jpds: so? [15:14] Riddelll: thanks for syncing tangerine [15:14] RainCT: bug 427051 [15:14] Launchpad bug 427051 in soyuz "Copy archives show their owner too prominently" [Medium,Triaged] https://launchpad.net/bugs/427051 [15:14] «I call this high priority, as it means Colin is staring at me far too much. Scary.» LOL [15:15] ok, thanks [15:15] * hyperair rofls [15:16] strider: if you want the debdiff reviewed, you should subscribe ubuntu sponsors for main. [15:16] Then you might want to ask on #ubuntu-desktop since it's one of their packages, to check if you filled the bug well (might need to set the status to Confirmed or Triaged, I don't know what they use). [15:18] * hyperair grumbles about firefox daily only working after installing debug symbols [15:20] hmm looks like it was starting before xulrunner got upgraded [15:20] oh well [15:20] hyperair: They do not want you to run firefox daily without contributing useful bug reports. [15:21] slytherin: heh [15:21] fta: http://pastebin.com/f1c2300b7 [15:21] fta: there you go. full backtrace [15:21] hmm no wait, i can get a fuller one i think [15:23] loic-m_, ubuntu sponsors for main is a restricted team, I can't subscribe directly [15:23] fta: http://pastebin.com/f618b552f [15:23] strider: I mean subscribe them to the bug ;) [15:24] you can probably get better feedback on #ubuntu-desktop though [15:29] uhm.. what's the fix for C code using an "int_32" type but not declaring it? "typedef int int_32;" or is there sth better? [15:31] int32_t [15:31] hyperair: thanks [15:32] it's C++, i'm not sure about C [15:32] or rather, c++0x [15:32] uhm, the package has: buildlib/inttypes.h: typedef int int32_t; [15:32] ah [15:33] then just #include it? [15:33] yeah, but just including won't do (as it's using "int_32" in some places) [15:33] ok, thanks hyperair [15:33] np [15:34] hello team, there's anyone running karmic which can reproduce this: https://bugs.launchpad.net/ubuntu/+source/azureus/+bug/428514 ??? [15:34] Launchpad bug 428514 in azureus "Azureus crashes both openjdk and sun java." [Undecided,New] [15:34] RainCT: find -type f -print0 | xargs -0 sed -i -e 's/int_32/int32_t/g' [15:35] azureus crashes *with* openjdk and sun java? or it seriously crashes both of those? [15:35] also i'm not an azureus user [15:36] Heya gang [15:43] Hi all, is there a way to get to revu packages when revu is itself down? (looks like there's no db connection there). [15:51] noodles775: no and checking into it now [15:52] imbrandon: ok, thanks! === rockstar` is now known as rockstar [17:16] How would I backport a fix from karmic's nautilus to jaunty? [17:16] Or would it be best to just bite the bit and upgrade before the release? [17:18] !sru|doctormo [17:18] doctormo: Stable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates [17:18] and [17:19] !backport| doctormo [17:19] doctormo: If new updated Ubuntu packages are built for an application, then they may go into Ubuntu Backports. See https://help.ubuntu.com/community/UbuntuBackports - See also !packaging [17:19] Thanks fabrice_sp, I guess I'll just wait. [17:19] lol. ok [17:23] doctormo: depending on how big the change in, it can be painful to isolate the change well enough to backport the fix [17:23] change *is* [17:25] jdong: Aye, I figured nautilus would not be isolated enough to really bother with, I can wait. [17:41] Can someone help debug a build issue? http://pastebin:freeswitch@pastebin.freeswitch.org/10450 [17:46] quentusrex, you still have this conf file issue? [17:46] yup... [17:46] is there a place where I can download your package? [17:47] (REVU or similar) [17:47] umm... it won't build... [17:47] aah, you mean the source package... [17:47] yeah [17:47] the source, I mean [17:47] just a sec [17:47] yes :-) [17:47] I'll build a source package and upload it to launchpad.. [17:47] ok [17:47] your ppa is ok [17:47] it will be easier to debug your issue :-) [17:48] Hey all, I've got a FTBFS due to a package shipping empty PO files which pkgstriptranslations doesn't seem to like. But simply deleting the files doesn't show up in a diff.gz. Any idea how to best deal with this? [17:49] Delete them in rules or is there something better? [17:51] fabrice_sp: I'm going to start a new +repack of the package, then I'll upload that to launchpad.. [17:51] I'm sooooooo close to having this working. [17:52] asomething, I would add a delete of empty po files in the rules file, or speak with upstream :-) [17:52] quentusrex, np :-) [17:52] hi. [17:52] can anyone confirm that deluge is busted on karmic? [17:53] freeswitch_1.0.4+repack5-0ubuntu14925.0 https://launchpad.net/~pbxbuntu-drivers/+archive/ppa [17:53] fabrice_sp: give it a few minutes to finish uploading... [17:53] LLStarks: mine works [17:53] jcastro. broken on fresh a6 install. [17:53] quentusrex, ok [17:54] https://bugs.launchpad.net/ubuntu/+source/deluge/+bug/434749 [17:54] LLStarks: no problem here either, what's your issue? [17:54] Launchpad bug 434749 in deluge "Deluge freezes upon opening in Karmic" [Undecided,New] [17:55] LLStarks, could you please re-report that with "ubuntu-bug deluge" so that it adds version and dependency info, ect... [17:55] fabrice_sp: ok it's accepted. [17:56] persia, heya! can I be added to u-u-s please? [17:57] * fabrice_sp is downloading freeswitch [17:58] geser: aaarrrrggghh. http://launchpadlibrarian.net/32265767/buildlog_ubuntu-karmic-i386.cdk_1%3A1.0.2-2ubuntu1_FAILEDTOBUILD.txt.gz [18:00] https://bugs.launchpad.net/ubuntu/+source/deluge/+bug/434757 [18:00] Launchpad bug 434757 in deluge "Deluge freezes upon opening in Karmic (v2)" [Undecided,New] [18:01] LLStarks, thanks! [18:03] sebner: I don't understand why. I've checked in my i386 karmic chroot that installing the build-dependencies as listed in the apt-get line from the build log works [18:03] fabrice_sp: just a heads up. it takes about 20 minutes to build... :) [18:03] geser: Yeah, As I told you, I testbuilt even in my PPA without problems [18:03] oh, that's the main archive? [18:04] geser: hmm? [18:04] quentusrex, ok: Let's see how long it lasts in mine :-) [18:04] fabrice_sp: and I haven't been using the make clean portion. I just delete all of the data, and extract the tar again. [18:04] sebner: :) libjgrapht0.6-java | 0.6.0-8 | http://archive.ubuntu.com karmic/multiverse Packages [18:04] quentusrex, by the way: the version is ugly [18:05] :-) [18:05] sebner: go hunt an archive admin [18:05] geser: omg. [18:06] fabrice_sp: what do you mean about the version being ugly? [18:06] fabrice_sp: can you think of a better way to do it? [18:06] sebner: I guess it's because the old version of that binary package was in multiverse [18:06] geser: "Sind die denn komplett deppert?" xD [18:06] quentusrex, ugly because of ubuntu14925.0 [18:07] sebner: the new upload didn't probably need to pass NEW as it's a known binary [18:07] fabrice_sp: I couldn't think of a way to add the svn revision to the version, in such a way that wouldn't effect the orig file... [18:07] only the -link packages needed NEWing [18:08] geser: it did go through NEW, jdstrand managed that for me [18:08] but would also let me have multiple shots at getting a particular revision packaged more than once... in case I run into issues. [18:08] hmm [18:09] quentusrex, something like freeswitch_1.0.4+svn14925+repack5-0ubuntu1 (if repack still apply in that case) [18:09] fabrice_sp: that'd be cool... but wouldn't that require a new orig for each svn revision? [18:10] geser: It's a horror. The same problem again [18:11] sebner: you just need to find an archive admin who fixes the component [18:11] quentusrex, each time you get the svn version, yes. How are you doing it now? [18:12] geser: yep [18:12] * fabrice_sp is downloading the build dependency of freeswitch [18:14] fabrice_sp: I'm looking to build a nightly package for this one. [18:15] but use the same orig for in between 1.0.* releases [18:15] so that's why I put the svn where I did... [18:15] it is ugly, but it works for my needs... [18:15] the non dev packages don't have the svn version.. [18:19] * hyperair wonders if anyone from motu-release can look at bug #432900 [18:19] Launchpad bug 432900 in ubuntu "[FFe] Please sync remuco (universe) 0.9.1.1+dfsg-1 from Debian Unstable (main)" [Undecided,New] https://launchpad.net/bugs/432900 [18:20] quentusrex, as you want. The firefox team is doing it in another way: https://launchpad.net/~ubuntu-mozilla-daily/+archive/ppa [18:20] package building [18:20] I'll take a look [18:21] geser: /me retries cdk build [18:22] fabrice_sp: is it starting to build for you? [18:22] sebner: wait on the next publisher run else you probably get an upload error [18:22] hyperair: Approved. [18:22] ScottK: thanks. [18:22] geser: at every full hour? [18:22] You're welcome. [18:23] sebner: Starts at :03 and finishes at ~:45. [18:23] sebner: yes [18:23] kk [18:23] thx [18:23] hyperair: I'll make MOTU review [18:23] sebner: thanks =) [18:25] quentusrex, yes [18:28] quentusrex, compilation got frozen :-/ [18:30] what? [18:30] Anyone know of a way to set up pbuilder to use pkgstriptranslations so that it acts more like the buildds? [18:30] it takes a while for some of the files [18:30] asomething: Install (IIRC) packagebinarymangler into your pbuilder chroot. [18:31] quentusrex, has to upgrade/update my chroot [18:31] asomething: install pkgbinarymangler in your pbuilder and enable the parts you want in the config files [18:32] geser, ScottK: thanks. It will run on its own? Doesn't need a hook? [18:33] asomething: it dpkg-diverts the standard tools [18:36] hyperair: what version of ffox do you get the backgrace? [18:36] asac: 3.7 daily [18:36] fabrice_sp: darn it... [18:37] asac: the backtrace can be gotten from xulrunner 1.9.3 later than 20090914. [18:37] hyperair: forcing extensions? [18:37] I build it for karmic... it should have been jaunty [18:37] hmm extensions eh.. [18:37] yeah [18:37] should i try safe mode? [18:38] fabrice_sp: I'll reupload.... [18:39] hmph how strange. it works again [18:39] okay, it's intermittent [18:39] fabrice_sp: I'm uploading the replacement... [18:40] hmm bad adblock [18:42] quentusrex, you mean in the changelog? It's no problem (I'm building with karmic, anyway) [18:43] fabrice_sp: ok... [18:43] I uploaded a new one anyways with the correct release... [18:43] asomething, I've just seen now that you are the one that put a comment on bug #416262. FYI, I tried to clarify the situation, if you want to have a look at it [18:43] Launchpad bug 416262 in aptoncd "Sync aptoncd 0.1.98+bzr112-1 (universe) from Debian unstable (main)." [Wishlist,New] https://launchpad.net/bugs/416262 [18:45] fabrice_sp: ok, i though that was the case, but wanted to make sure. [18:46] asomething, I have to admit it was a mess :-) Thanks for asking :-) [18:56] hyperair: ok. so did you force adblock? [18:56] where is the current feature freeze for universe policy? [18:57] * asac checks topic [18:57] asac: https://wiki.ubuntu.com/FreezeExceptionProcess#Exceptions for Universe/Multiverse [19:02] in what part of a package I'm creating a user when packaging a server? [19:08] quentusrex, I've the error :-) So now, I have to check what is installed where [19:09] fabrice_sp: ok. thanks. [19:12] fabrice_sp: I'm trying to set things up so that the configs are in a separate package... [19:12] because some people don't want the configs, since they have already made their own. [19:12] and these configs are more like sample configs. [19:13] <_ruben> such configs tend to go under /usr/share/doc//(examples/) [19:13] _ruben: that's where we'd like to move them to [19:14] but for now we are leaving all the files where they are now, and just getting it to package separately [19:14] quentusrex, so in this case, you won't need .conffile [19:14] <_ruben> ah ok, didnt read scrollback :) [19:14] just install them as example [19:15] quentusrex, but what about someone that install the pacakge for ths first time? [19:15] thanks for acking it asomething :-) [19:15] fabrice_sp: np =) [19:15] quentusrex, I think the /opt prefix is wrong === porthose|afk is now known as porthose [19:16] ok... but where is '/opt' set? I did a grep and didn't see it in the package at all [19:16] they are all 'opt/' [19:16] in debian/rules [19:16] ./configure --prefix=/opt/freeswitch --host=$(DEB_HOST_GNU_TYPE) --build=$(DEB_BUILD_GNU_TYPE) ${FEATURES} [19:16] that's wrong for a Debian pacakge :-) [19:17] s/pacakge/package/ [19:17] fabrice_sp: how should it be? === keffie_jayx_ is now known as keffie_jayx [19:20] fabrice_sp: I think I found the error.... [19:21] quentusrex, --prefix=/usr [19:21] 8I think) [19:22] oh, that kind of stuff will be fixed later... [19:22] I just need it to build... [19:22] This is why I'm still putting this in a development ppa... [19:23] quentusrex, about the conf file. Do you want to install it as an example or as a conf? [19:26] I'm changing it to be where it should be [19:26] in opt/ [19:29] hmmm, not sure opt is the right place [19:34] fabrice_sp: opt isn't the right place [19:34] but that is where the software expects it to be [19:34] so I need to put it there for now [19:35] pochu, ScottK, ScottK-desktop: I uploaded the phatch 0.2 release and POX is now busy with it. I will be back in one and half hour [19:36] pochu or stani: This will need an FFe approved. [19:37] quentusrex, so for the moment, use only .install files [19:38] ScottK: I know [19:38] stani: OK. Please ping me when the FFe is ready for review. [19:42] * POX gave phatch 0.2 branch a hard time, that should count as "+1" for FFe ;) [19:43] geser: cdk successfully build and accepted \o/ thanks for you help :) [19:55] POX: We'll get it in. Thanks for all your work on it. [19:56] stani redirected me to nadia yesterday (sick of my bug reports probably ;) === ripps_ is now known as ripps [20:40] any MOTU SRU about? === yofel_ is now known as yofel [20:58] * fabrice_sp just discovered that dbs is not a typo of cdbs. How many build systems are there ?! [20:58] a few ;-) [21:00] :-) === nxvl_ is now known as nxvl [21:04] does anybody know what the status of a bug should be if it's waiting on MOTU SRU to approve it? [21:08] I have a server I want to package as .deb [21:09] It has a folder with sub-folders containing library files and script [21:09] it is all compiled so I only need to copy it to the files system [21:09] I also understood that it is best practive to create a system user for this server [21:10] I've been looking at db-helper but it seems it is more suitable for source packages [21:10] LaserJock, New [21:10] (if I understand correctly https://wiki.ubuntu.com/FreezeExceptionProcess) [21:12] Davedan, debhelper is a serie of helper functions to build binary packages. The building part is done using the app building system [21:12] so it should fit your needs [21:13] fabrice_sp: hmm, I guess I could try it even though this is for an SRU and not a freeze exception [21:13] LaserJock, arghhh: I must be tired: I read FFe :-/ [21:18] fabrice_sp: what file do I need to edit to copy my server files to the file system? [21:18] what is the correct place on the file system for the server library files? [21:20] Davedan, you can have a look at http://www.pathname.com/fhs/pub/fhs-2.3.html [21:21] but by packaging, you won't copy it to your file system. You will copy it in a temporary place, that get 'packaged' in a deb file [21:22] fabrice_sp: how does the deb file knows where to put my server files on the file system? [21:23] and what file of the template files created by debhelper do I need to edit to create a system user that will run the server [21:24] Davedan, the deb file contains 2 gz file: one with controls files and the other with the packaged files. This second gz has the same directory structure as your 'sytem' [21:24] Davedan, for the user, I would say postinst after install to create it and postrm, to remove that user on removal [21:25] anyway, if you don't know how packaging works, it will be hard to package that 'complexe' app [21:25] so if I want my server to be placed under /usr/local/myserver I should put it under temp_deb_folder/usr/local/myserver ? [21:26] Davedan, yes, except that packages never installs on /usr/local. They install in /usr directly [21:26] my app is simple. I doesn't require compiling. just need to place a folder on the file system and create a user [21:27] thanks. I'll try that [21:27] Davedan: you want to use the debian/install file to gett he files where you want them [21:27] Davedan, except that this folder is binary content. It should really be compiled in some way [21:28] Davedan: and debian/postinst and debian/prerm to add and remove the system user respectively [21:28] LaserJock: what is debian/install? is it part of the files created by deb_helper? [21:28] Davedan: yep [21:28] Davedan: it tells debhelper what files go where [21:28] Davedan: but in the system context [21:29] fabrice_sp: it's an erlang application that has a VM so it is self consistent [21:29] LaserJock: so I don't need to touch the rules file? [21:29] so if you have file foo/bar and you want it in /usr/share/foo then you would put foo/bar usr/share/foo in debian/install [21:30] Davedan: you can if you want, but I like using debian/install better [21:30] that's the part I was looking for [21:31] LaserJock: what permissions will the foo/bar file get by default? who will be the owner? [21:31] root I believe for all files in a .deb [21:31] do I need to set permissions for the new user just on the script that runs the server? [21:32] that would be my guess but I don't know for sure [21:33] thank you two. that realy helps [21:33] Davedan, if you need more hellp, you can have a look at the packaging guide [21:33] !packaging [21:33] The packaging guide is at http://wiki.ubuntu.com/PackagingGuide - See https://wiki.ubuntu.com/UbuntuDevelopment/NewPackages for information on getting a package integrated into Ubuntu - Other developer resources are at https://wiki.ubuntu.com/UbuntuDevelopment - See also !backports [21:34] I read it but now I have some basis for reading it again and understanding more [21:34] :-) [21:38] Howdy folks. Yay Karmic! :-) [21:38] ScottK: would it help FFE if I assign all fixed bugs in https://bugs.launchpad.net/phatch/+bugs to the phatch(ubuntu) package? [21:39] ScottK: or does that create only more unnecessary work? === mdomsch is now known as mdomsch_linuxcon [22:01] will new source packages be automatically synced (after opening lucid)? === Tonio__ is now known as Tonio_ [22:03] bdrung_: yes [22:03] sebner: thanks [22:04] sebner: will they go through ubuntu's new queue? [22:04] bdrung_: afaik yes [22:13] MOTU team. Just filed needs-packaging request 434883. I know it's crunch time for Karmic and my needs-packaging request is a bit late. [22:14] I'd like to help out and make the package, for a PPA or something. Any pointers to where I can get started? [22:14] irvingpop: I don't think it's possible to get a new package in karmic at this point. [22:14] But, there's... [22:15] irvingpop: Start here: https://wiki.ubuntu.com/SponsorshipProcess [22:15] Yeah, I know. The thing doesn't even compile on Karmic, I really need to get it running on Jaunty [22:15] so I was thinking I need to make a PPA. I was told this channel would be the best place for me to get some help [22:15] Other than that, the MOTU page seems to have lots of good pointers. [22:16] irvingpop: Well, what help do you need? Building a package? Uploading it to a ppa? [22:17] Yeah. I've been packaging for other OSes for a long time, but I've never created a .deb package. This tool (Flashcam) has no packaging infrastructure at all right now [22:17] irvingpop: Have you checked out https://wiki.ubuntu.com/MOTU ? [22:17] Right now I'm just using "checkinstall" to package it for my own purposes. But I think it's useful to others so I want to share [22:18] I actually don't know what checkinstall is. [22:19] I just uploaded my first debdiff yesterday. [22:19] It's cheesy. Instead of running "make install" you run "checkinstall" instead, and it watches all of the install targets and makes a simple package for you. [22:19] https://wiki.ubuntu.com/UbuntuDevelopment/NewPackages is conspicuously missing a link on how to actually create a package. [22:19] irvingpop: Ah, nice. [22:19] stani: It's not needed. Just list the bugs in the FFe. [22:19] Yeah, so that's where I'm looking right now [22:20] I'm trying to find a simple template or something for new packages [22:20] stani: They ought to be in your changelog anyway, so a copy of that in the bug should do. [22:20] http://blog.mypapit.net/2006/02/create-you-own-debianubuntu-deb-package.html [22:20] ScottK: ok [22:21] !MOTU | Darxus [22:21] Darxus: motu is short for Masters of the Universe. The brave souls who maintain the packages in the Universe section of Ubuntu. See http://wiki.ubuntu.com/MOTU [22:21] sebner: Why did you do that? [22:21] Darren: click that link -> To get started right away on your MOTU adventure, click here! [22:21] Darxus: there is every information you need [22:21] irvingpop: Please don't recommend checkinstall. [22:22] !checkinstall | irvingpop [22:22] irvingpop: checkinstall is a wrapper to "make install", useful for installing programs you compiled. It will create a .deb package, which will be listed in the APT database and can be uninstalled like other packages. See https://help.ubuntu.com/community/CheckInstall - Read the warnings at the top and bottom of that web page, and DO NOT interrupt CheckInstall while it's running! [22:22] ScottK, there are no warnings on that page [22:23] I've been told its a big no-no [22:23] There aren't? [22:23] Grumble. [22:23] irvingpop: true ;) [22:23] irvingpop: Yeah, the "To get started right away on your MOTU adventure, click here!" link on https://wiki.ubuntu.com/MOTU looks like exactly what you're looking for. [22:23] it actually makes checkinstall look sound like a decent solution just glancing through it [22:24] checkinstall is part of the RPM packaging process as well, which is why I started there [22:24] but that's it, I won't mention it again [22:24] :) [22:24] superm1: And then when I tried to look at the page history I got an internal server error. [22:25] Yeah, checkinstall sounds quite lovely :) [22:25] ScottK, what were the big reasons that were recommending against it? looking at the history, it looks like some fo the big warnings that were related to permissions that got fixed in newer versions [22:26] superm1: I'm not sure. [22:26] It's not possible to automatically produce policy complaint packages. [22:26] ScottK: not wondering, no script/app can do that [22:26] ScottK: Yeah but it seems like a great place to start. [22:26] So even if some of the more dangerous flaws have been corrected, it's fundamentally unsuitable. [22:27] Darxus: I disagree. [22:27] I think knowing what you are doing is a great place to start. [22:27] ScottK: Document your objections on https://help.ubuntu.com/community/CheckInstall [22:31] Darxus: well, if you want to do something, do it right. https://wiki.ubuntu.com/PackagingGuide/Complete [22:32] sebner: I understand. But somebody should document disadvantages on the checkinstall page. [22:32] Darxus: true, as no one uses checkinstall here it might have been forgotten or people just don't care [22:33] although i wouldn't recommend it's usage over learning to package, i would be interested to find out what about it's packaging is outside of policy too [22:34] Two of the links (the blog from Darxus and the Packaging/Complete wiki page) seem to be enough to get me started [22:34] superm1: just choose an application to package a) the right way b) checkinstall and then compare :) [22:34] it's not easily reproducible, including overwriting changelog on slight change. [22:35] i'm not even satisified it does much more than dh_make anyway. [22:36] From what I gather, checkinstall is not a sustainable way to build packages which integrate into a build system (like Launchpad), but it is suitable for one-off "I just need this packaged quickly" jobs [22:36] * ScottK marked on the page and prepares for wikiwar. [22:36] irvingpop: something like that, yes [22:37] OK, I'm going to go make a package. sorry if I rocked the boat :) [22:38] irvingpop: hrhr, np. I do a make install if I need an app quickly though. Not used checkinstall since years [22:38] go ScottK go [22:38] I tried to be low key about it. We'll see. [22:39] sebner: I need it quickly on a dozen desktops, otherwise I wouldn't bother [22:39] ScottK: what was the name of this horrible application which installed stuff and did apt-get --force and that stuff [22:39] automatix [22:39] It's still around under a different name, AFAIK. [22:39] Plus, if anyone has tried to use a Webcam with Flash Player recently they would feel my pain [22:40] bbl [22:40] ScottK: write something like that: Checkinstall is the automatix for packaging. HAHA [22:41] * ScottK just said don't us it for pacakges intended for distribution. [22:42] bah :P [22:43] Does checkinstall even create a .deb? [22:44] If people want to blow up their own systems, that's really their business. [22:44] Darxus: yes [22:45] ScottK: hmm as long as they don't file bugs on LP then .... [22:45] Yep. [22:45] Back when automatix was popular I used to sometimes search in LP bugs for automatix and invalid any where it came up. [22:45] I used automatix :P [22:45] Darxus: not surprising :P [22:46] So did I, before I knew better. [22:46] The real beauty of the way it was broken was that it would work, but would screw future upgrades, so they almost never got the blame. [22:46] Yeah, my video broke bad during an upgrade. [22:47] Drivers. Nvidia. [22:47] But it provided useful stuff that ubuntu didn't. [22:48] If it'd been done in a sane way, it would have been fine. [22:55] Do I have to do anything else to get this debdiff into karmic? https://bugs.launchpad.net/ubuntu/+source/azureus/+bug/428514 [22:55] Launchpad bug 428514 in azureus "Azureus crashes both openjdk and sun java." [Undecided,New] [22:59] It's such a minor change, and it make azureus not crash :/ [23:03] Hey ScottK: you were so picky about licensing with pycryptopp that I've pre-emptively gone and cleaned up all licensing issues that I can think of for Tahoe-LAFS. ;-) [23:04] When I get off work today I intend to try uploading a new Tahoe-LAFS to REVU myself, which is just like the current one up there except with a bunch of grant of rights statements and the copyright file fully fleshed out and so on. [23:05] Since the Beta Freeze is coming up, and since some people already advocated for an earlier upload of Tahoe-LAFS which is functionally identical to this one, can I somehow mark it on REVU as partially advocated or something? [23:08] One person -- kirkland -- advocated Tahoe-LAFS. [23:12] in what folder needs to place the script that start and stops a server ? [23:13] and in what folder are the server library files [23:14] Davedan1: /etc/rc* [23:14] Davedan1: Server library files? Something other than /usr/lib ? [23:16] Darxus: /etc/rc* is for the init.d script? [23:17] Davedan1: https://wiki.ubuntu.com/PackagingGuide/Complete <- it should be in there. [23:18] Darxus: I've read it but it doesn't tell you where to put the program files [23:19] Davedan1: You need to put a script in /etc/init.d/ and then simlink it to the various /etc/rc* directories corresponding to the runlevels you want it started and stopped for. [23:19] Does ubuntu really not document this? [23:19] Might be worth checking debian policy. [23:20] Darxus: I know about the init.d script but sometime servers call a nother script from the init.d script [23:20] so the init.d script just has the bare minumum and it excute commands from server_name_ctl script [23:21] Okay. [23:22] so my question is were to put server_name_ctl script [23:22] Davedan1: Ah. I don't know. [23:23] Davedan1: Probably /usr/sbin/ [23:24] Davedan1: I was serious about checking debian policy. [23:24] Davedan1: http://www.debian.org/doc/debian-policy/ [23:25] http://www.debian.org/doc/debian-policy/ch-opersys.html#s-sysvinit 9.3 System run levels and init.d scripts [23:25] thanks [23:25] You're welcome. [23:26] POX, pochu, ScottK: do I have to wait untill Phatch is uploaded to Debian to file a FFE? [23:28] stani: it will have to go thru NEW in Debian (new binary packages), so pochu should probably upload it to Ubuntu once I'll do it in Debian [23:31] Davedan1: Debian policy says file system hierarchy must comply with http://www.debian.org/doc/packaging-manuals/fhs/ [23:32] Davedan1: http://www.pathname.com/fhs/pub/fhs-2.3.html#USRSBINNONESSENTIALSTANDARDSYSTEMBI [23:32] Davedan1: Yeah, I think /usr/sbin would be right. [23:32] I've read it [23:32] ok. thanks [23:45] stani: What POX says makes sense. Go ahead and file the FFe. [23:46] ScottK: thanks