[00:38] (On that whole overgod thing above: apart from anything else, unless I'm very much mistaken, the complainant has misunderstood the word "overgod". It doesn't mean "one who is over God"; it means "supreme God".) [00:39] (A bit like "high priest" or "overking" or whatever.) [00:42] I thought the complaint was using what was apparently a proper noun in a polytheistic context. [00:44] I read it a couple of times and as best I could make out his syntax the above was the core of his complaint. [00:44] though I imagine there was some of what you describe too [00:46] Hrm. On rereading, I can see your interpretation as well. === rgreening_ is now known as rgreening [05:16] :) === santiago-pgsql is now known as santiago-ve [05:30] * RAOF tries to remember the magic that makes pbuilder work properly on a tmpfs [05:55] Bah. How do I get pbuilder to not try to copy or hardlink files to the apt cache? [05:55] I guess it's a bug. Setting APTCACHE="" does not, in fact, make pbuilder fail to do this. === ripps_ is now known as ripps [06:24] so, what is the preferred way to add color support to something like aptitude anyway? i imagine there'd be problems with terminals that don't support color and different background colors [06:54] raylu: man terminfo [07:36] jmarsden: did that and a bit of googling... i'm fairly confused. [07:37] i'm not expected to actually parse these files myself, am i? i was looking at ncurses too, but i'm not sure if it's suitable for my needs since i don't intend to draw any windows [07:47] Hi from packagejam south africa :) [07:48] hi [07:48] so what's happening there in .za? :) [07:49] haha. im waiting for people to arrive :) [07:49] hopefully you get a good turnout [07:50] planning to package up some new stuff & get it in, or merge/fix up existing packages there? [07:51] ajmitch: our first one, my main goal is to get people interested and started, just over that first hump so they start contributing === nellery_ is now known as nellery [07:53] ok [07:53] good luck :) [07:53] and let us know how it goes, it'd be an interesting thing to run here [07:54] will do, i will blog about it, and i hope to get a couple others to do the same === pochu_ is now known as pochu [11:35] jameica, libwilluhnds, libwilluhnutil and libswtcalendar are now ready to be reviewed === tuantub_ is now known as tuantub === bastiao__ is now known as k0p === AndrewGe1 is now known as AndrewGee === krusaf|bnc is now known as krusaf [12:44] Is it appropriate to request package sync now, after DebianImportFreeze? [12:47] randomaction: if there's a good reason, yes [12:47] randomaction, absolutely [12:48] I think that a new upstream version counts as a good reason [12:48] if it's an stable version, it is :) [13:48] jameica, libwilluhnds, libwilluhnutil and libswtcalendar are now ready to be reviewed === asac_ is now known as asac === emgent is now known as emgent` === kklimonda is now known as kklimonda|test === Guest52255 is now known as kklimonda [16:11] Heyo, I have a question that I could use some help on. What would be the correct versioning to update a package in this situation: version 1.0-0ubuntu1 is the current version in the archive, but a more recent git snapshot of it is needed and the version internally is 1.1 but 1.1 has not been released. Would 1.0+git20090627-0ubuntu1 be the way to do it, or 1.1~git20090627, or something else even? [16:15] both are good as both are smaller than 1.1-0ubuntu1 (assuming 1.1 will be the next upstream release) [16:32] Help regarding revu: I already packed my application properly (already fully working on my ppa) but I am in doubt regarding the version numbers. Should I add "ubuntu1" to the end of my version number before uploading? What about the "Maintainer" field in the debian/control file? Should I change it to "Ubuntu MOTU Developers " ? [16:59] cpscotti: change Maintainer field to and XSBC-Original-Maintainer to your name / email [17:01] cpscotti: versioning is Appversion-debiannumberubuntuubuntunumber [17:01] cpscotti: like app-0ubuntu1 [17:02] cpscotti: 1.7-0ubuntu1 sorry [17:02] cpscotti: 0 because it is not in debian already, isnt it? [17:04] hey, I've been wondering is it possible to create an empty deb file that will upon installation own some (unowned by another package) files? [17:07] kklimonda|test: didnt understand what you mean [17:08] therm: hmm.. for example we create an installer that installs files and then dummy deb file that will own them and entry in sources.list so it can be upgraded just like any other package in system. [17:12] kklimonda|test: is it closed source or why do you need such an installer? [17:13] therm: well, this is just an idea I had for a closed source applications [17:15] therm: is it your application? [17:17] kklimonda|test: is it your application? [17:17] therm: no - it's just an idea I've had when I was supposed to study to my exams ;) [17:18] therm: it sounds like something that if could be done would be useful - personally I'd like to see commercial applications integrated with package system and all the goodness it brings. :) [17:19] brb [17:20] kklimonda|test: you could make a ppa where you package those apps, so people could add them to sources.list. Its as easy as running an installer [17:21] therm: but then I'd have to create a packages for every (leading) distribution. [17:21] therm: thanks !! you got it all right [17:21] kklimonda|test: I cant imaging people would like to have installers change their sources.list [17:22] therm: no problem [17:22] therm: why not? As long as it just adds another entry I see no problem. After all they (people who use them) run a closed source installers right now. Yes, it's a matter of trust.. as always [17:24] therm: but if you didnt want to make a package for every leading distribution, how do you plan to make this "and then dummy deb file that will own them"? [17:24] kklimonda|test: but if you didnt want to make a package for every leading distribution, how do you plan to make this "and then dummy deb file that will own them"? === lex79 is now known as lex [17:25] kklimonda|test: do you search a program wich is creating debs,rpms what ever automaticly by the installed distribution? === asac_ is now known as asac [17:27] kklimonda|test: something like that, I have not heard of. but opensuse hase something I dont remember the name, it creates deb/rpms for many distributions, that might help you [17:29] https://build.opensuse.org/ [17:30] jmarsden: thanks ;-) [17:36] therm: obviously dummy package file would have to be prepared but it's more a matter of space on a install medium. For example if program is 3-4GB and you install it from a dvd then you just can't prepare one dvd. [17:36] wtr build.opensuse.org it's as I said - preparing and building packages isn't a problem itself - distribution of them is. === emgent` is now known as emgent === lex is now known as lex79 [18:20] Can anyone review this package http://revu.ubuntuwire.com/p/harpia ? It is a "simulink-like" automatic programming tool for computer vision and image processing. === yofel_ is now known as yofel === bastiao_ is now known as k0 === k0 is now known as k0p [18:38] warp10 and nellery, I see you are from the MOTU's Science team, could you review this package ( http://revu.ubuntuwire.com/p/harpia ). I is, in many ways, similar to FlowDesigner. [18:38] *It [18:40] * warp10 takes a look at harpia [18:44] cpscotti: I gave a very fast look to the debian/. You should add a watch file, s/jaunty/karmic/in the changelog, and add a white line before the last line. I'll give a broader look and make more tests ASAP [18:49] cpscotti: ah, I see you are the upstream developer too. Nice! :) === JontheEchidna is now known as jmthomas [18:51] =] === Quintasan is now known as mzajac [18:51] lets fix this watch thing === Kmos_ is now known as Kmos [19:33] omg... /me sees the "remove this game.." thread on ubuntu-motu and wonders whether he should cry or laugh XDDDD [19:36] sorry again for another question and thanks for the help earlier! what's the correct format for multiple launchpad bug closers in the changelog that require a new line? should the ones on the new line start a new () or is it smart enough for this to work? [19:36] Fixes problems returning from a dpms off state. (LP: #390917, [19:36] #383973, #123456, #123456, #123456) [19:36] Sarvatt: that should work [19:36] thanks a bunch! [19:37] Sarvatt: that should work [19:37] (oops sorry,pressed up arrow + enter in the wrong window) [19:43] RainCT: full ack [19:44] RainCT: the game is indeed crap though xD [19:46] RainCT: We had a long discussion here yesterday on the game topic. [19:46] * sebner runs [19:53] sebner: hehe Right. After reading the mail I thought maybe I should blog a review on planet.ubuntu.com or send something to debaday, but after trying it out I don't think I'm going to do that :P === bizkut-miau is now known as bizkut [20:34] hello. trying to package the latest version of wireshark I get those errors at the end. http://pastebin.com/d7b8c3a14 I used the karmic package as starting point. [20:39] c_korn: my ignorant suggestion would be.. do you have fakeroot installed? [20:40] c_korn: "/usr/lib/libfakeroot/libfakeroot-sysv.so" exists? [20:41] cpscotti: hm, the build is running in a schroot. I need to check if fakeroot is installed there... [20:42] cpscotti: yes, this file exists [20:43] that fakeroot version is installed: Installed: 1.12.1ubuntu1 [20:43] in what way conflicts fakeroot with the build process here? [20:44] I never encountered any issues relating fakeroot before [20:45] seems it will use fakeroot so that files inside the package have root ownership... and things like that [20:45] seems it was just a blind guess [20:46] you changed LD_PRELOAD ? === _ruben_ is now known as _ruben [20:48] no, I did not change it [20:48] you said u started from the karmic's wireshark package right? [20:49] without changing anything you still could not build it? [20:50] well, I just edited the patches to apply on the new sources [20:50] this is all I did. no changes to debian/rules [20:50] but before applying the patches could you build it? [20:51] I was not able to build the new version. [20:51] and I did not try to not apply the patches === mzajac is now known as Quintasan [21:22] c_korn: I downloaded wireshark's source from https://launchpad.net/ubuntu/+source/wireshark/1.0.7-1ubuntu1 [21:23] Hi guys, I've got a package (in progress), the COPYING file is GPL2 but the headers on 95% of the source files are standard GPL, no version specified, shall I list the entire app as GPL-2 or GPL-any in the debian/COPYRIGHT file? [21:24] c_korn: and I could build it ok [21:24] c_korn: using ordinary shell [21:26] cpscotti: did you try the latest wireshark release? [21:27] the failing package is here: dget -ux http://abs.getdeb.net/pre_build/jaunty/wireshark_1.2.0-1~getdeb1_source.changes [21:27] I'll try [21:28] crux: If upstream is at all active, ask them what they intended. [21:29] ScottK, I already asked, they intended GPL-2 even though most of the headers don't specify the version [21:29] Then specify it that way in debian/copyright. [21:29] It's a complex copyright situation, loats of different files with different dates and authors, I used the new/proposed copyright format from the debian wiki so it should be clear to the repo admins [21:30] I don't want to make their job any harder [21:42] ScottK, lintian noticed that I have binary files with no corresponding man pages, am I ok to submit this package to revu without man pages [21:42] ? [21:42] Afterall, not all apps have man pages [21:42] crux: Odds of it getting uploaded that way are very low. [21:42] You can write them. [21:42] Aren't there plenty of apps in the repo without them though? [21:43] and to what standard/scope should I write the man pages too? [21:44] There are, but it's been policy for some time that all binaries should have one, so those that have missing man pages are buggy. [21:44] They don't need to be extensive. Even if they just have some basic info and a pointer to the real docs, that's a help. [21:45] c_korn: I builded that package you linked properly.. no errors [21:46] *uilt [21:47] ScottK, I see, well this app isn't exactly well documented, I don't think there is much I can say in the man pages, the app is really all about the GUI and I can't write docs for everyday usage of the app [21:53] Even if you just say how to get to what documentation there is, it's something. If the app has any command line switches when being started, they should be covered too. [21:53] So, applications without documentation are pretty much not going to make it into multi/universe ? [21:53] I just checked with upstream [21:53] there are no plans for documentation [21:53] If you've made a good faith effort given what upstream gives you to work with, then it's not necessarily a blocker. [21:54] I'll submit it and see what happens I guess [21:54] it's quite a niche area, if your using this app you'll already know how it works [21:54] it's for motor engine tuning [21:57] Quick thing: when uploading a new version to revu (with some packaging bugs fixed), what version number should be changed? ubuntu"X" ? [21:58] cpscotti: REVU is just for new packages. For upgrades just put the diff.gz in a bug on LP. [22:00] new version of a new package... in my case [22:02] Then it should be [newupstreamversion]-0ubuntu1 [22:03] its the same upstream version, I just added the missing "watch" file and fixed the distribution target on the changelog [22:03] but when I try to upload dput tells me I already uploaded it [22:05] is passing "-f" to dput the solution? [22:05] Removing the .upload file is a better one [22:05] If you're just updatingthe packaging on REVU, don't change the version number at all. [22:06] thanks! === hsitter is now known as Ken === rgreening is now known as Barbie [22:16] cpscotti: hm, ok. thanks. how did you build the package? [22:16] was fakeroot installed? (this library existing)? [22:16] yep === vorian is now known as hans [22:16] I downloaded the files from your link === Barbie is now known as Chewie === Chewie is now known as Vader === Vader is now known as hsitter [22:19] c_korn: then:$ dpkg-source -x wireshark_1.2.0~getdeb1.dsc tmpFold [22:19] $cd tmpFold [22:19] $debuild -S -sa -rfakeroot === Ken is now known as rgreening [22:20] c_korn: since you said somethin about the patches, I applied some of then, changed the changelog and did debuild again.. everythin fine [22:20] hm, I use sbuild which creates a schroot to build in. I will try to build it in a PPA. thanks for your help [22:20] what do you mean. some of them? [22:20] u wlc [22:20] that not all appear in the 00list file? === hans is now known as vorian [22:21] the ones in debian/patches/ [22:21] one of those scripts === hsitter is now known as not_hsitter_hone === rgreening is now known as self [22:22] (even though that doesn't make much of a sense..) === not_hsitter_hone is now known as rgreening === self is now known as hsitter === hsitter is now known as apachelogger [22:32] what's the difference between the files pbuilder creates in ../../debian and the ones it creates in /var/pbuilder/result? [22:32] both dir's contain a diff, a dsc and a build file [22:32] only the results dir has the actual binary deb === pochu_ is now known as pochu [23:04] there is a really great scanning application, gnome-scan. i have even heard it is considered to make it the default ubuntu scanning app. however the currently included version is very outdated. there are new versions available, both in launchpad and revu. and even so the original developer seems very supportive, no new version gets included... if one of you could have a look... https://bugs.edge.launchpad.net/ubuntu/+source/gnomescan/+bug [23:05] nailora: +gs [23:09] sebner: what does that mean [23:11] nailora: your link misses a "gs" at the end. bugs and not +bu ;) [23:14] Can anyone review this package http://revu.ubuntuwire.com/p/harpia ? It is a "simulink-like" automatic programming tool for computer vision and image processing. [23:14] nailora: http://revu.ubuntuwire.org/p/gnome-scan - I think the hold-up is because of the bersace user last comment: "I’ll migrate to new copyright format later." [23:15] nailora: there's also a debian request for package: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=395334 [23:15] Debian bug 395334 in wnpp "RFP: gnomescan -- Gnome Scan Infrastructure" [Wishlist,Open] [23:20] is there a shell command to llist lines in a file? [23:20] as in number of lines [23:23] wc [23:23] cat file | wc -l ? [23:23] unncessary use of cat alert! [23:24] wc -l [23:24] iirc, you get the filename printed if you use cat on its own [23:24] bencrisford: you can use head and tail together, example will retrieve first 10 lines and then the last 5 lines of those: head --lines 10 packages.txt | tail --lines 5 [23:24] rys: you do with wc -l too [23:24] I figured he was just after the raw number [23:24] err, wc on its own even [23:24] im after a raw number yeah [23:25] what i want, is to put it in a script [23:25] ahh [23:25] so i do a: [no.of lines command] > $VAR [23:26] this will show line no. 10 I think: head --lines 10 packages.txt | tail --lines 1 [23:27] yep :) [23:27] W: megatunix: copyright-refers-to-versionless-license-file usr/share/common-licenses/GPL [23:27] lintian ^ [23:27] where do I change this? [23:27] where does lintian look to get license info? [23:28] crux: pastebin your debian/copyright file somewhere so we can take a look at it [23:28] savvas, http://paste.ubuntu.com/205218/ [23:30] crux: if the source files really are GPL you can ignore/override this warning [23:30] sebner, yes, all the .c source files are GPL (version less) except two of them which are public domain, the LICENSE file that came with the app is GPL-2 [23:31] how is lintian coming up with this license error? it must be getting its info from somewhere [23:32] side note > I guess I could take that public domain files and list them as gpl-2 as that's what the app is [23:33] 1. Format-Specification: I think it should show the revision of the CopyrightFormat, e.g.: http://wiki.debian.org/Proposals/CopyrightFormat?action=recall&rev=455 [23:34] What shell command would I use to check if a directory exists? [23:34] test -d $DIRECTORY [23:35] kklimonda: I ran that and got no output [23:35] is that right? [23:36] yes - test basically exits with a status of expression [23:37] if i was writing a script [23:37] what would my if statement look like [23:37] savvas, How exactly do I do that? [23:37] so if you want any output (for debug) you can use test -d $DIRECTORY && echo 'found' [23:37] for "if $directory exists do this" [23:37] crux: for the gpl error, I have no idea, can you try to include it in quotes? try this: '/usr/share/common-licenses/GPL-2' [23:38] kklimonda: Would this work: test -d /home/user/directory && echo 'true' > $VAR [23:38] savvas, would that be for the files: * section? [23:38] then: if [ "$VAR" = "true" ]; [23:38] crux: you click on "Info" link at the copyrightformat wiki page: http://wiki.debian.org/Proposals/CopyrightFormat?action=info [23:39] bencrisford: if test -d /home/user directory; then var=true; else var=false; fi [23:39] crux: the one you pasted still shows the error you mentioned, right? [23:39] you can also use if [ -d /home/user/directory ];then var=true; else var=false; fi [23:39] savvas, yes [23:40] crux: replace all /usr/share/common-licenses/GPL-2 with '/usr/share/common-licenses/GPL-2' (not sure, but worth a try) [23:40] This is what I have in a test package: [23:40] On Debian systems, the full text of the GPL can be found at [23:40] '/usr/share/common-licenses/GPL-3'. [23:40] what means TBD (next revu day) [23:40] ? [23:40] cpscotti: to be decided [23:40] savvas, already tried that, thing is, there can be more than one license file so lintain shouldnt look at every section, it should just look at the section that declares the license for the whole app [23:41] maybe this is fixed in the backport of lintian [23:41] thanks nailora [23:41] also, im fairly new to bash, is != a valid operator for if statements and the like? [23:41] if not I'll just ignore the error and submit to revu, if someone knows more about this I'm sure they will send a comment my way [23:41] crux: I have installed on jaunty 2.2.10 from debian, no idea :) [23:42] haha [23:42] thanks for the suggestions, I'll just see what feedback I get :) [23:42] bencrisford: it is but shell programming is a bit messy.. [23:42] * bencrisford likes messy :D [23:42] bencrisford: [ 'a' != 'b' ] && echo 'test' [23:43] so it works with if [23:43] bencrisford: bash or sh? [23:43] bash [23:43] ok just checking :) [23:43] savvas: as far as I can see != works with both bash and dash [23:44] it does it does, just checking for future questions :P [23:44] yeah, I was just going ask the same.. [23:44] bashisms are really pain in the ass :/ [23:45] there was a checkbashisms script somewhere.. devscripts? [23:45] yes [23:46] grr, how do you redirect to a variable [23:46] command > $VAR doesnt seem to work [23:46] VAR=`command` [23:46] VAR = $(command) [23:47] er.. VAR=.. [23:47] what kklimonda said :) [23:47] $() also works [23:47] I don't know what's the difference [23:47] i guess one of them is "more official".. [23:48] no idea hehe