=== jamalta-afk is now known as jamalta [03:41] How much space does a pbuilder thingy take? [03:47] Some_Person: depends on what you do with it [03:47] the base image, i mean [03:48] This machine has limited HD space, so I need to knwo [03:48] s/knwo/know [03:49] hold on, lemme check mine [03:49] hi, anyone here can talk about licencing of a project (again) [03:49] Some_Person: mine seems to take around 100 MB [03:49] my emails to support have been ignored it seems [03:49] keithy: want to talk to an LP admin? [03:49] that'll do [03:50] nigelb: It'll be a tight squeeze, but I think it'll fit [03:50] keithy: The European or American working week is a much much better time. [03:50] keithy: yeah, monday to friday EU time will have better luck [03:50] How can I screw the pbuilder stuff from my system when I'm done? [03:51] Some_Person: donno what you mean + this is the wrong place for this discussion I think [03:51] nigelb: What is the right place? [03:51] wgrant: #ubuntu-motu would be beter? [03:51] And what I mean is, how can I get rid of this 100MB image when I'm done? [03:52] What's #ubuntu-motu? [03:52] Some_Person: I'm not sure how to remove the basefile other than to remove pbuilder. we generally dont need that feature [03:52] #ubuntu-motu is probably better, yes. [03:53] Some_Person: join #ubuntu-motu channel and ask there, a lot more experts will be around [03:53] What is "motu"? [03:53] !motu [03:53] 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 === jamalta is now known as jamalta-afk [06:33] It said my package was building now, but a minute later it says it'll start building in an hour [06:34] Some_Person: Are you sure you weren't looking at different architectures? [06:34] No, I just refreshed the page [06:35] Now it says it's building again, though on a different builder [06:35] It's a really fishy package, so I want to make sure it builds and works properly [06:39] Some_Person: Link to the build? [06:40] it's finished building now [07:11] are we allowed to make our ubuntu.com addresses default on LP yet? [07:12] micahg: https://bugs.launchpad.net/launchpad-foundations/+bug/5292 is still open, so I'd guess not [07:12] Launchpad bug 5292 in launchpad-foundations "People setting preferred contact address to @ubuntu.com" [Medium,Triaged] [07:12] rww: so, is that the same as having bugmail coming from the ubuntu.com address? [07:12] * micahg subscribes to that bug [07:14] micahg: i think so [07:14] rww: so I'm wondering how people do it now [07:14] micahg: example? [07:15] rww: PM? [07:15] micahg: The reason it works for me is very old and historical, and predates that bug by a long time. [07:15] persia: you were going to be one of my examples :) [07:15] (and happens to be different than the other reason it works for certain people) [07:16] k [07:17] Basically, there are some folk who have @ubuntu.com not related to launchpad (for historical reasons), and there are some folk who have @ubuntu.com cowboyed in during early days of integration with launchpad, and it works for them, but it doesn't work for new folk since the improved launchpad integration. [07:17] That's theoretically fixable, but hard, and making it work the way it works for either of the first two classes of people is just broken. [07:18] (hence the bug) [07:19] rww: nm [07:24] micahg: Sorry to disrupt your PM: I just don't think there's anything there which isn't public (the former case can be understood by reviewing MX records, and the latter case (mine) are a set of known exceptions and workarounds). [07:34] persia: I cancelled the PM request [07:35] hence my apology :) [07:39] persia: I asked for a PM because I wasn't sure if people were doing stuff wrong and didn't want to "out" them in public :) [07:39] I don't think any of the individuals did anything wrong. [07:40] persia: right, but I didn't know that when I started asking... [07:40] The issue with namespace collision on the mailserver is long-standing, and awkward to address. [07:40] The few of us who have hand-entered aliases from the beginning of the LP integration stuff are just acceidents of timing. [07:41] A real fix to the bug is to be able to separate target address from preferred address in LP. [07:42] And then to notify all the special cases that they need to set things properly. [07:42] And then set up a new mailserver, have it pull from the new LP stuff, and then change MX records. [07:42] That's hard, and it's questionable whether it's worth it. [07:43] (especially because there are a few hundred special cases) [07:44] persia: Are the special cases those Canonical employees with canonical.com aliases that happen to work on ubuntu.com too? [07:45] wgrant: Not all of them. There are two classes of special case. [07:45] Well, yes, I meant the big one. [07:45] wgrant: So, there's the @canonical.com @ubuntu.com thing (some of which I believe was addressed about 18 months ago, but I'm not sure). [07:46] Yeah, the majority are probably those. [07:46] But also there are folk (like me) that ended up with hardcoded aliases to work around bugs with the LP integration back in the early days. [07:46] Ah. [07:47] So in the beginning there was a big aliases table, and new members got added there. [07:47] Then we wanted to use LP more, and sometimes that worked, and sometimes it didn't. [07:47] Anyway, it's not actually much to do with LP -- IS handles it. [07:47] And now LP mostly just works, so nobody gets added to the aliases table anymore. [07:48] Well, there's a few bugs in IS, but the bug in LP is that it doesn't differentiate "preferred address" from "alias target address", making the bugs in IS hard to fix. [07:48] LP doesn't know about aliases. [07:49] Right. [07:49] It would have to grow that as an attribute to fix the IS stuff. [07:49] Without that attribute, it requires additions to the aliases table, which are currently (partially) automated based on LP preferred address (as I understand it: I'm not privy to the code) [07:50] Alternatively the IS thing could be replaced with a simple webapp authenticating against LP and checking ~ubuntumembers membership. [07:50] Indeed. That's probably better than trying to force LP to do it. [07:51] Now that authentication and membership verification against LP is easy, that is probably the way to do it. Plus it could be decoupled from the LP username, which is probably a good thing. [07:51] So user@ubuntu.com wouldn't necessarily be lpnet/people/user ? [07:52] I suppose it isn't always now do to how the mailserver works anyway. [07:52] It could be implementable like that if so desired. [07:52] s/do/due/ [07:52] That sounds like an excellent suggestion, except that I'm unsure IS is likely to develop such a webapp. [07:53] But I'm now convinced this isn't an LP bug. [07:57] * persia updates the bug based on this discussion [08:09] I've bumped into a bug: if a bug (e.g. bug #410028) has nominations for releases against multiple tasks, I don't seem to be able to approve one without leaving the other unapproved. Does anyone happen to know if this is already filed? [08:09] Launchpad bug 410028 in openobject-client-kde "Add view buttons" [Wishlist,New] https://launchpad.net/bugs/410028 [08:15] Err, that was supposed to be 401028, but I'm guessing nobody heard of this before anyway. [08:15] Filing now. [08:22] * persia should trust the find-similar-bug algorithm more: it's bug #271697 [08:22] Launchpad bug 271697 in malone "Javascript for approving / declining nominations is confusing " [Undecided,Triaged] https://launchpad.net/bugs/271697 [08:29] persia: You mean that you can't accept nominations for different source packages separately? [08:30] Yeah. [08:30] Bug #11195 [08:30] Launchpad bug 11195 in linux-source-2.6.15 "ripping from CD-ROM non-functional" [Medium,Invalid] https://launchpad.net/bugs/11195 [08:30] Er. [08:30] Bug #110195 [08:30] Launchpad bug 110195 in malone "Nomination for a release on one source package shouldn't affect any others" [High,Triaged] https://launchpad.net/bugs/110195 [08:30] I filed it about 3 years ago. [08:31] Is 271697 a dup, or just an expression of 110195 also in javascript? [08:31] They are separate. [08:31] Although fixing 110195 will probably fix 271697, since the nomination objects will be distinct. [08:32] But 271697 could be fixed independently. [11:49] anyone here who can help a newbie. I want to "upload" around 300+ debs. [11:50] Its for the flightgear project. There are over 300+ aircraft, which are just a bunch of files (images/xml) in a directory. [12:10] p_masho: Why do you need that in 300 different packages? [12:11] each aircradt needs to have its own ppa, they average around 10meg each.. [12:12] just realised I cant do this anyway.. after rtfm ing ;-( [12:12] heh. [12:12] Why can't you? [12:12] Packages like that are trivial to do, but often need a lot of consideration. [12:14] wats a pain in the ass is the updates [12:15] ats them moment its (once in a blue moon ) >> "cvs co " (yes cvs).. >> scan dir for updates >> make tarball >> upload tarball.. to ftp site >> mirrored === Ursinha is now known as Jorjao [15:08] Hi. I have a couple of different PPA packages which I would like to make available for different Ubuntu series. One package contains only data and is quite universal, but depends on less universal packages. Do I really have to make copies of this package for each series? [15:13] askhl: is the data package build from its own source package? [15:14] geser: yes it is. Also I wrote something wrong: it doesn't depend on anything - other packages depend on *it*. Sorry [15:14] depending on the packages you have (if they need a rebuild or not) it might be enough to copy them (through the web UI) to the other release [15:15] for a data package copying should be enough [15:15] The data package is 20MB, so I would rather avoid that [15:17] avoid the copying? [15:18] without a copy it won't get published for the other release, I assume that this won't use any additional space in your PPA [15:18] So copying for different releases doesn't count against the PPA size limit? [15:19] Anyway, I'll just create copies if that's what it takes, whether or not it takes some space. [15:20] I'm not sure, but as you re-use the same files and as they even stay in the same location I doubt it [15:22] Okay, that makes sense. If I need to update the data package, though, what is the procedure? Update the newest and then copy to other releases, or update each one individually? [15:22] I guess I should try to find these things in the Ubuntu or Debian packaging guides [15:28] I'll browse the documentation for further information. But I'll make copies to support different series in any case. Thanks a lot for the help, geser. [15:37] Hi! [15:38] Are there any plans for supporting TS (Qt Linguist) translation files in Launchpad? [15:38] I've googled a bit and only found recommendations to use translation-toolkit. [15:39] Also, there is an open wishlist bug on the tracker starting back from 2006, but it is quite inactive, and it is neither accepted nor rejected. [15:41] Converting to and from PO is a solution, but it is seems more like a workaround, and this way one would need to upload files manually. [15:41] Or keep the POs (which are, in fact, duplicates) in Bazaar repo. [17:23] hi all. i have a question regarding deb packaging for my PPA. is this the right channel to ask? [17:34] yes [17:36] i try to build a binary package from the sources by doing: 'sudo pbuilder build ../gavl_1.2.0pre1-1.dsc'. it stops with this error: "make: dh_testdir: Command not found" . however, 'dh_testdir' exists (debhelper is installed). how comes that fakeroot does not seem to find the dh_testdir command? [17:37] pbuilder uses a clean and bare chroot for building. if you need any special package during build (like debhelper) you need to specify it in Build-Depends [17:38] in debian/control [17:38] geser, thanks [17:50] geser, am i right in thinking, that i should adapt /debian/control ? i added 'debhelper' there ('doxygen' was alreaday there), but when executing the 'pbuilder build' command, i see that it is still only installing 'doxygen', but not 'debhelper'. sorry for the noobish questions [17:56] that is the complete output: http://pastebin.ca/1827292 [18:29] rdz: did you recreate the source package (debuild -S) after you changed debian/control? [18:30] geser, thanks.. yeah, i got it in the meantime [18:31] geser, now i am stuck at another issue: http://pastebin.ca/1827344. there seems some problem with the debian/rules file [18:33] this is the 'clean' section of debian/rules: http://pastebin.ca/1827345 [18:45] now, after i performed './autogen.sh' before doing 'debuild -S', it seems to work. does this mean i should add the './autogen.sh' command add to debian/rules? [18:46] or is the usual way to prepare a package to bring it into a state ready for doing ./configure ? === yofel_ is now known as yofel [19:38] I just uploaded my first source package to my PPA. Will I just have to wait until I see it appear? [19:42] rdz: you should get an accepted mail within 5 minutes [19:44] Say, if you 'import' a translation in Launchpad, what happens if the msgids are not the same? [19:45] Presumably it is strictly required when 'importing' as well as when 'uploading' that the po-file in question actually contains exactly the same msgids. [19:45] mwhuds ah.. i see.. it was not accepted .. thanks [20:43] "i386 - Pending publication": Am I required to publish this package? or do I simply need to wait? [20:43] If it requires me doing something, what shall i do? [20:46] ah.. i had simply to wait [20:46] :-) [21:01] Is OOPS-1527L2262 because somebody changed the package while I was typing for minutes on a comment which appears to be lost now? [21:01] https://lp-oops.canonical.com/oops.py/?oopsid=1527L2262 [21:06] blueyed: huh, that oops itself oopses [21:06] OOPS-1527L2276 also? [21:06] https://lp-oops.canonical.com/oops.py/?oopsid=1527L2276 [21:07] morning [21:07] blueyed: that one isn't replicated yet [21:12] lifeless: then it's probably no error.. ;) - btw: another case of the comments/data being eaten. I really get to the habit of copying the most important fields to the clipboard before submitting something on LP. Chromium does not refill the form elements when going back. [21:13] blueyed: file a bug ? [21:13] [on chromium] [21:14] I guess it's rather related to cache headers/ssl which prevents this. could be the case with FF, too. [21:15] blueyed: filing a bug means it might get fixed. [21:16] I would bet that there's a bug filed for both LP and Chrome already. and yes, it might get fixed. [21:22] shrug [21:22] up to you, I wouldn't bet anything on that bet [22:20] hi all. i added a package to my ppa which has version 1.2.0pre1-1. ubuntu repos have version 1.1.0-2. although i added my ppa to the sources, aptitude still wants to install ubuntu's version. how are versions compared? do i have to specify the package at a different location as well (beside debian/control)= [22:20] ? [22:38] rdz: have you run "aptitude update" after you added your PPA? [22:40] geser, i did [22:40] what's the package name? [22:41] libgavl1 [22:43] I have a 75MB package in my PPA. I need to get my debian folder out of the package but do not want to download 75MB because of my slow connection. Anything I can do? [22:44] Some_Person: if it's a non-native package and the upstream tarball didn't contain the debian dir, then downloading the diff.gz is enough [22:45] Package has a completely different rules file than the one upstream due to changes since that version [22:45] rdz: does "apt-cache policy" list your PPA? [22:45] Also, wouldn't the diff.gz only give me changes since the last version uploaded? [22:46] geser, no, it doesn't [22:46] the diff.gz contains all differences compared to the upstream tarball (like e.g. a debian directory or patches) [22:46] There have been numerous changes since upstream [22:47] upstream package is 2 years old, lots of development since then [22:47] rdz: then you probably didn't add your PPA correctly. how did you do it? [22:47] then the .diff.gz will contain all those changes (if they aren't part of the .orig.tar.gz) [22:48] geser, sudo add-apt-repository ppa:reduzierer/rdz-pd-extra+deps [22:55] geser, when adding the ppa the traditional way (adding to /etc/apt/sources.list manually, adding the key manually) it works [22:56] although it did not work before, my ppa was listed by the software-properties-gtk gui program [22:59] geser, thanks for your help [23:09] i have troubles finding the information: how can i tell the ppa to compile for many supported distros? [23:20] you need to re-upload for each release you want to support (and update the version and distribution field in the changelog) [23:21] geser, i see. why does the verion also need to be updated? [23:21] *version [23:22] because you can upload a version only once (and because all files for a package are stored together and the version is part of the filename) [23:23] hm..how do you call the different versions then (that are actually the same)? [23:23] sorry for the noobish questions... [23:24] you can just append e.g. ~karmic or ~jaunty to your version string. that's enough to make the different [23:25] geser, thanks [23:39] hi thumper - which bug has >1000 subscribers? [23:41] oh nm, i can find out [23:41] OOPS-1526ED777 [23:41] https://lp-oops.canonical.com/oops.py/?oopsid=1526ED777 [23:42] bug 1 of course [23:42] https://bugs.launchpad.net/ubuntu/+bug/1 (Timeout) [23:42] jeez, why would someone want to subscribe to that? [23:42] I was just thinking the same thing [23:50] lifeless, btw https://bugs.launchpad.net/bugs/534066 broke hydrazine bugclient fairly completely :-/ [23:50] Launchpad bug 534066 in malone "can't update bugtask importance via api" [Undecided,New] [23:59] poolie: ugh [23:59] I wonder if it's the heat bug. [23:59] But it seems unlikely that that would fail every time.