/srv/irclogs.ubuntu.com/2009/12/17/#ubuntu-motu.txt

=== ripps|sleep is now known as ripps
apacheloggerpersia: ahoy, now that nixternal did a +1 on Quintasan's application I suppose quorum is reached and Quintasan can be confirmed to be motu?01:02
persiaapachelogger: I'll probably do that today.  I was just kinda hoping that the others would chime in, but they failed to respond to requests.01:05
persiaapachelogger: Is there some urgency that means it should be done *now*, or would in a few hours be just as good?01:05
=== bdefreese2 is now known as bddebian
=== tuantub_ is now known as tuantub
=== mezgani is now known as p3rror
=== NCommander is now known as Guest48610
=== nigel_nb__ is now known as stackedagainst
ScottKdktrkranz: When you merge csound, please be sure to bump the boost build deps (we're aiming for 1.40)02:53
=== julezz is now known as julez
=== Guest48610 is now known as NCommander
=== NCommander is now known as Guest88322
ScottKbddebian: Does mapnik build with boost1.40 on Debian?03:46
bddebianScottK: Haven't tried, would you like me to?03:51
ScottKbddebian: Yes.  Please.03:51
bddebianpbuilding...03:55
=== nhandler_ is now known as nhandler
ScottKThanks.03:58
LucidFoxCan an orig.tar.gz include free TTF fonts?04:00
bddebianShouldn't04:01
* LucidFox nods04:01
ScottKLucidFox: As long as you use the system fonts and they are appropriately licensed, it's no problem.04:02
ScottK(i.e. don't use the code copies shipped in the tarball in your build)04:02
bddebianScottK: Getting some warnings about deprecated headers but so far it's going04:02
ScottKbddebian: I got to /usr/bin/ld: cannot find -lboost_python-mt04:09
bddebianScottK: Yep, it built04:10
bddebianHmm, strange04:11
ScottKAny suggestions how I might figure that out?04:11
=== Whoopie_ is now known as Whoopie
bddebianScottK: I assume that lib does exist?04:15
* ScottK goes to inspect the .deb04:16
bddebianIs it pulling in libboost-python-dev?04:16
ScottKYes and yes.04:19
bddebianVery odd04:19
ScottKMaybe ajmitch knows since he's officially our boost-python expert now.04:20
bddebianw00t :)04:20
ScottK(He fixed a bug in it, so that makes him the expert.04:21
ScottK)04:21
bddebianWFM :)04:21
dtchendoes anyone have access to an armel porter?04:22
dtchenI'm getting a slew of FTBFS that I can't easily test :(04:23
ScottKThere's a armel PPA that Canonical devs have access to.04:26
ScottKOther than that, I think you need ogra_ or NCommander.04:27
=== Guest88322 is now known as NCommander
fabrice_spbigon,  why do you need sponsorship on bug 427542: you're a motu, right?05:37
ubottuLaunchpad bug 427542 in libhildon "Merge libhildon 2.1.34.debian.1-6 (universe) from Debian testing (main)" [Wishlist,Confirmed] https://launchpad.net/bugs/42754205:37
=== asac_ is now known as asac
dtchenfabrice_sp: his membership in MOTU is active, at least05:40
dtchenso, yes05:40
fabrice_spdtchen, that's what I thought: as this package was in main some releasea ago, I think he got 'distracted' :-)05:41
fabrice_spthanks  to confirm ;-)05:41
siretart`Whoopie: IIRC we disabled dv support because vlc needed porting to the new API, or there was some general breakage with dv support. Have you tested that dv actually works?06:05
siretart`Whoopie: in any case, I've pointed xtophe to your patch, he takes a look and will probably apply it for the 1.0.4-1 upload06:06
dtchenmm, I can ditch the "nvidia" driver completely, because "nv" works great in Lucid06:34
bjsniderdtchen, is that a joke?06:42
dtchenabsolutely not06:48
dtchengranted, I don't use anything requiring libGL.so.1, so my particular case is limited06:49
bjsniderthe nv driver is there to provide basic mode-setting so you don't see a black screen the first time you boot linux06:49
bjsniderit's a transitional driver06:49
bjsnidernouveau and nvidia are much better06:50
dtchenI don't much care, frankly; if nv gives me x-terminal-emulators, I'm golden06:50
maco2who needs compositing?06:50
maco2dtchen: wait why dont you just use screen in a tty? it can do tiling06:51
dtchenI do use tmux/screen.06:51
bjsnidermaco, human beings walking around on this earth06:51
bjsniderbut i never mentioned compositing06:52
maco2well thats the point of using nvidia instead of nv, isnt it? because nv is 2D-only. i kinda though nouveau was still rather experimental though06:52
maco2*thought06:53
bjsnidernv has no xv or vdpau06:53
bjsniderfont rendering is probably a mess06:53
bjsniderno power management. no support for xrandr 1.206:54
bjsnidernot much of anything06:54
bjsniderof course the blob doesn't support xrandr 1.2 yet either06:54
bjsniderbut nouveau does06:54
bjsnidernouveau works great, but possibly not in lucid yet06:54
dtchenfont rendering is fine06:54
dtchenI used nv for about nine hours without realizing it06:55
bjsniderdtchen, you don't use your rig for anything interesting? it's all just coding shell scripts in gedit?06:55
dtchengedit's a bit too fancy for my lazy self06:56
=== Quintasan1 is now known as Quintasan
bjsniderwhy not just boot to a back screen then...06:56
dtchenI do.06:56
bjsniderjust you know, install the kernel by itself06:57
Quintasanhi06:57
maco2bjsnider: because sometimes he uses firefox instead of lynx06:57
dtchengeez, the point is that I can finally kill off the last non-Free driver06:57
maco2you could've just gotten intel graphics and not had to worry about it :P06:58
bjsnidernouveau is foss06:58
maco2yes, his point is he can ditch the binary nvidia driver06:58
bjsniderbut he's not using the best foss driver06:58
maco2he doesnt care which alternative to go to, it just so happens that nv is on his system and works, so he's going with it06:59
bjsniderwas my point06:59
maco2if it works, does it matter?06:59
dtchenI'm sure there are more optimal methods, but I can't be arsed; I'd rather fix this pot of audio bugs.06:59
maco2hmm though now you've got me wondering... is there a curses calendar app i could use with mutt to ditch kontact?07:00
=== Rh0nda is now known as Rhonda
dholbachgood morning07:43
persiadtchen: I've got an armel jaunty box (haven't been able to upgrade yet).  Is there something I could test with that?07:43
=== ara_ is now known as ara
RAOFbjsnider: If things go well, nouveau will be the default for lucid, so daniel will be using it instead of nv :)07:59
=== Adri2000` is now known as Adri2000
bigonfabrice_sp: right the pkg was previously in main, my bad08:56
paultago/ MOTU, I have a question about the process of importing from debian09:03
persiapaultag: ask it :)09:03
paultagdo we use the dsc and modify the control / changelog and use that, or do we take the upstream tgz and recreate the debian folder09:03
paultagpersia, sheesh, I was working on it :P09:03
persiapaultag: Sorry.  We get lots of folk who ask to ask and need prompting.09:04
paultagpersia, I know that deal ;)09:04
m4rtinhi all; I submitted a patch for sponsorship review (regression fix) over a month ago and haven't heard anything. Any ideas; core team v. busy? (it's in bash-completion)09:07
persiam4rtin: Which is the bug number?09:08
* persia can't help get it merged, but may be able to help check process compliance09:09
m4rtin#44934909:09
persiabug #44934909:10
ubottuLaunchpad bug 449349 in bash-completion "regression for completing remote files/dirs over ssh" [Medium,Confirmed] https://launchpad.net/bugs/44934909:10
persiam4rtin: So, is broken in lucid as well as karmic?09:13
m4rtinpersia: I would assume so... it was broken *in* Karmic, so I wrote a fix which someone then suggested I try to get into lucid09:13
m4rtinIt was owing to a bash version upgrade and the bash_completion script being incompatible09:15
persiam4rtin: So, basically you just want someone to upload the debdiff in comment #45 ?09:15
persia(and then maybe you'll look at karmic, but that's process stuff for later)09:15
m4rtinpersia: #25 I believe... as for Karmic, I was lead to believe that this would not be ported back as an update...? Is that incorrect? I've been using the fix on my Karmic systems with no problems...09:17
persiam4rtin: #25, riht.  Sorry.  Dunno how I imagined the extra 20 comments.09:18
persiaIt looks to me like you've done everything correctly, and are just stuck waiting for a sponsor.09:18
m4rtinpersia: yeah, I thought so - I'll just sit tight09:19
m4rtin(my first patch, so I didn't know what to expect)09:20
persiam4rtin: Yeah.  Sometimes it's slow.  Sorry about that.09:20
m4rtinit also fixes another very irritating issue in the same package that has existed since 8.10 - ssh completion puts far too many \ characters in to escape spaces and thereby makes it useless for completing more than 1 directory - it's all in the changelog though :)09:21
persiam4rtin: On the bright side, it looks like it got confirmed as a regression by the update verification team, which means that it's probably getting a bit more visibility than it would otherwise09:22
persia(but that was a couple weeks ago)09:22
m4rtinpersia: yeah, I saw that; guess they were busy with featuredefinitionfreeze recently, but hey, at some point they'll see it I'm sure09:23
persiaWe can certainly hope.09:24
=== etali1 is now known as etali
=== ogra_ is now known as ogra
Whoopiesiretart: yes, the DV plugin works under karmic, I don't have a lucid installation.10:23
Whoopiesiretart: I added vlc 1.0.3 to my PPA and enabled DV support for testing.10:23
siretart`Whoopie: I'd say then feel free to upload the change to lucid :-)10:28
siretart`Whoopie: or even better, let's discuss this on #debian-multimedia@oftc.net and get this change into debian proper10:28
Whoopiesiretart`: I can't upload anything.10:42
maxbwiki/SponsorshipProcess doesn't say much about what needs to appear in a merge-sponsorship-request bug. Should I be describing the changes, or should I assume that the sponsor will prefer to read the diff and debian/changelog anyway, so a brief "please merge this" is ok?10:43
persiaI like to get a quick description of the patch.10:46
persiaSomething like "Here's something that addresses ${issue} for ${release}10:46
persiaJust to have some idea what I'm looking at.10:46
persiaYes, the sponsor will be reading the diff and changelog, but a little advance hint never hurts :)10:46
persiamaxb: Actually, in the specific case of merges, some people like more.10:47
persiaFor instance, some people like to have the imported Debian/changelog entries summarised in the bug (the ones coming from Debian).10:47
persiaThis is less important before Debian Import Freeze, and it would be nice to merge everything.10:48
persiaBut after that it becomes more important, as it lets the sponsor make a more informed decision as to whether it's worth digging into the details10:48
=== mac_v is now known as _
=== _ is now known as _7hills
=== _7hills is now known as mac_v
siretart`Whoopie: okay, I'll remember that when uploading to and merging from debian10:57
persiamaxb: The only relevant thing I find in the wiki is https://wiki.ubuntu.com/UbuntuDevelopment/Merging?action=show&redirect=MOTU%2FHowToMerge#File%20a%20merge%20bug11:00
persiaWhich doesn't seem to ask for that much info.  I personally tend to file as much info as I would for a sync, but it may not matter.11:01
maxbok, thanks for the info :-)11:04
=== c0p3rn1c is now known as copernic
RoAkSoAxhey guys is merges.ubuntu.com been updated regularly?12:21
slytherinRoAkSoAx: It should be. What is the problem?12:22
RoAkSoAxslytherin, some debian package versions are still not updated in the universe list even though the debian package has been in testing for quite a few days12:23
slytherinRoAkSoAx: any example?12:23
RoAkSoAxslytherin, apgdiff (its been in testing for 4 days :) ) Or, how much does it take to update the list?12:23
slytherinRoAkSoAx: As far as I knew it used to be updated daily. May be that has changed recently.12:24
RoAkSoAxslytherin, prolly its is because they want us to change to distributed development or because the new debian source format12:25
RoAkSoAxs/prolly/probably :)12:26
slytherinAFAIK, MoM has support for new source format.12:26
RoAkSoAxslytherin, yeah but they were doing some tests this past few days in building the packages to see if they can change or not since I think some packages were FTBFS12:27
slytherinRoAkSoAx: Was that related to MoM or Launchpad?12:28
RoAkSoAxslytherin, Launchpad12:28
persiaLaunchpad still needs a bit of work in order to support uploads of the new format, but I'm not certain this is related to MoM12:28
RoAkSoAxslytherin, do you know who maintains those lists?12:28
slytherinnope12:29
slytherinpersia: AFAIK, source format 3.0 landed in edge.launchpad.net12:29
persiaslytherin: I heard it was disabled pending some software upgrades.12:30
persia(that was from #launchpad sometime in the last 12 hours)12:30
slytherinOk. Then you have latest info. I was only tracking the bug.12:31
=== ara__ is now known as ara
Laneylaunchpad but not the buildds, afaik13:08
=== ikonia_ is now known as ikonia
=== Quintasan_ is now known as Quintasan
=== propagandist1 is now known as propagandist
=== rmcbride_ is now known as rmcbride
slytherinIf a library has some part sunder this license it is free enough to be included in archives - Smack contains icons and images licensed from INCORS GmbH. You are not licensed to use these icons outside of Smack.14:16
persiaErm.14:16
persiaI'd say that belongs in multiverse, personally.14:16
Whoopiejdong: Hi, sorry for prodding but how do we continue with bug 481448?14:21
ubottuLaunchpad bug 481448 in vlc "VLC lacks build-dep on libupnp3-dev" [Undecided,New] https://launchpad.net/bugs/48144814:21
Whoopiejdong: I just need someone to upload the package.14:21
slytherinpersia: the icons will be included in jar file and not be available separately.14:23
persiaslytherin: Right, but I don't think it's DFSG free.14:23
slytherinthat is bad then. :-( I will ask on Debian lists.14:24
persiaI don't think that icons packaged that way meet the requirements for #3 and #6 at http://www.debian.org/social_contract#guidelines14:24
=== Quintasan1 is now known as Quintasan
persiaYeah, ask Debian.  I'm probably extra picky :)14:24
slytherinWhoopie: You should bug RainCT about that.14:26
slytherinpersia: And what if the library which includes the icons is not being packaged at all? I mean there are 3 jars in the package and only one contains these icons. I am not planning to ship this jar in any binary package.14:30
LaneyI'd repack to remove it then14:31
persiaslytherin: I don't think it matters: the *source* needs to be DFSG free (well UFSG, which is basically DFSG + a couple specific exceptions)14:31
persiaBecause the source is *also* being distributed.14:31
persiaBut send a note to debian-legal explaining the details and your plans, and see what they say.14:31
slytherinOk. Then repacking makes sense. That way source becomes DFSG free and binary has no problems.14:31
persiaThat also works :)14:32
slytherinby the way looking at quality of icons I am wondering why they are bothering with shipping them at all.14:32
persiaheh :)14:32
mterrypersia, I was looking at the link at the top of REVU to the diff file: http://revu.ubuntuwire.com/revu1-incoming/netbook-launcher-efl-0912171333/netbook-launcher-efl_0.2.0.0-0ubuntu1.diff14:39
RhondaOh, CoC 1.1 finally arrived. :)14:40
persiaYou're looking at an old revision: see http://revu.ubuntuwire.com/revu1-incoming/netbook-launcher-efl-0912171439/netbook-launcher-efl_0.2.0.0-0ubuntu1.diff14:40
Rhonda... but somehow the release date can't be proper, 2005-04-12. :D14:41
persiaheh, no.14:41
mterrypersia, Huh.  Yeah, I went to the main REVU page and landed on the new page.  I guess I didn't realize I could get stuck on an old revision14:41
LaneyRhonda: That's because the old CoC is considered to be compatible with this one14:41
Laney(so as not tp14:41
Laneyto expire old signatures)14:41
persiamterry: Each revision is separately available for detailed review :)14:41
RhondaLaney: Erm.  "The current version is 1.1, released 2005-04-12"14:42
persiaLaney: Except the content changed: how are we to know if we agree with the current version?14:42
* persia does, but is curious about the abstract14:42
mterrypersia, still can't upload, as eina and friends aren't updated in lucid.  Well, I could push, but it would be dependency wait.14:42
RhondaLaney: I have no clue what you refer to so that old signatures don't expire, but that sounds pretty fishy.14:42
LaneyRhonda, persia: Yeah, I'm just the messenger14:42
persiamterry: Your call, rally.14:42
mterrypersia, they did that to not invalidate old signatures, under the theory that the changes were not substantive14:43
persiaLaney: Understood.  We just kinda hoped you had a larger message :)14:43
LaneyRhonda: So if people agreed to the older version then LP considers that they agree to 1.114:43
persiaUm, the changes are substantive enough that we've an *active* developer (Rhonda) who has been waiting for them for almost two years.14:43
RhondaLaney: And that for the page has to lie about the release date of the 1.1 version?14:43
Laneyhttps://bugs.edge.launchpad.net/launchpad-registry/+bug/479870/comments/314:43
ubottuUbuntu bug 479870 in launchpad-registry "Need to support Ubuntu CoC 1.1" [Low,Fix released]14:43
LaneyRhonda: That's because LP compares the date released to the date signed to see which versions of the CoC you have agreed to14:44
persiaOh, so it's an implementation trick.14:44
Laney...so they couldn't update the date released without expiring all old sigs14:44
Laneyyes14:44
persiaWell, they *could*, but it would require coding changes, etc.14:45
Whoopieslytherin: where can I find him/her?14:45
Laneyright14:45
persia(which is more delay)14:45
Laneythey couldn't *as it stands now*14:45
persiaRight.14:45
slytherinWhoopie: him, right here, have patience.14:46
Whoopieslytherin: ok, thanks14:46
RhondaLaney: A wording change would be appropriate then. This is just confusing as it stands now. :(14:48
LaneyRhonda: I agree. Perhaps you could file an LP bug?14:48
RhondaYes, am already wondering wether launchpad-registry is the proper target? :)14:49
Rhonda... actually, was just about to ask exactly that. :)14:49
Laneya) Fix the "date released" thing. b) Fix the semantics of signing previous versions — just need to make it clear that signing old versions is still acceptable for being an Ubuntu member, etc14:49
LaneyRhonda: Not sure, I'd just file to Launchpad and let it be triaged14:50
Laneyor you could ask in #launchpad14:50
mterrypersia, my irc client must be confused, I don't see any admins of this channel.  I'm assuming you're one?  Can you mark me a reviewer in the REVU database14:50
persiaREVU admins and #ubuntu-motu admins aren't the same set, but yeah, I'm supposed to be able to do that14:51
* persia hasn't done it in a while, so it may take a few minutes14:51
persiaBY the way, the Statistics page on REVU has a list of all the admins14:52
mterrypersia, oh, the REVU wiki page is unclear on the distinction.  let me edit.  How is someone supposed to know set of REVU admins?14:53
mterrypersia, heh, whoops.  OK14:53
shriekouthttp://revu.ubuntuwire.com/p/happytimer14:53
shriekoutplease... advice....14:53
persiamterry: Should work now.14:53
persia(and as our newest reviewer, maybe you want the latest review request ? )14:53
mterrypersia, I'll start with netbook-launcher-efl14:54
persiashriekout: I don't have time for a proper review of happytimer right now, but I'll give you a capsule one here14:55
persiaPriority "extra" is for packages that break things or conflict with core: you want Priority "optional"14:56
shriekout:)14:56
shriekoutyes...14:56
shriekoutthanks :)14:56
persiashriekout: Please add a debian/watch file or get-orig-source rule14:57
shriekouthumm...14:57
persiaErr, ignore that.  The diff is sorted differently than I expected :(14:57
shriekout:)14:58
persiaI don't see anything else quick-like.14:58
persiaBut I'm doing a few other things: maybe someone else can do a full review, or I'll catch it later.14:58
shriekoutyes..14:58
shriekoutpersia, thanks :)14:59
persiaNice work!14:59
RhondaLaney, persia: https://bugs.edge.launchpad.net/launchpad/+bug/49778514:59
ubottuUbuntu bug 497785 in launchpad "codeofconduct page falsely claims 1.1 was released 2005-04-12" [Undecided,New]14:59
shriekouthave a nice day :)14:59
persiashriekout: Well, l'm an hour east of you, but thanks :)14:59
shriekout:)15:00
persiaRhonda: Thanks for the pointer.15:00
shriekouti'm a korean... I’m not so good at English... sorry...15:00
shriekoutthanks persia for your advice :)15:02
* Rhonda thanks persia and Laney :)15:02
persiaRhonda: I'm not sure I deserve that, but you're always welcome :)15:03
RhondaYou do deserve more. :)15:03
jpdsRhonda: dholbach was the person who landed the coc-1.1 branch.15:05
Rhondajpds: I'm aware of that. :)  But this isn't about the CoC itself, rather about wording on the codeofconduct page in launchpad.15:06
ulysses__Hello, I am working on bug 49774215:13
ubottuLaunchpad bug 497742 in ubuntu "[needs packaging]Please replace STasks with Smooth Tasks" [Undecided,In progress] https://launchpad.net/bugs/49774215:13
ulysses__I want a littlhe help: what should I write in the control file for Build depends?15:13
persiaulysses__: That's usually best detemined by trial and error.15:15
persiaBut you might look at some other plasmoids, and start with a minimal set like that.15:15
persiaYou might also get more specific advise in #kubuntu-devel (although there are often a fair number of KDE people here).15:15
ulysses__Thanks15:15
Quintasandholbach: I sent you answers for Behind MOTU, the picture of me is just... well I'll wait for your reaction :P15:34
LaneyI forgot to respond to that when I got MOTU :(15:35
QuintasanLaney: what a shame, you are missing you opportunity for international fame, tons of money and girls ;)15:36
Quintasanyour*15:36
dholbachLaney: do it15:37
dholbachQuintasan: I'll have a look in a bit15:37
Quintasandholbach: Great :)15:37
mattidholbach: ;]15:37
bddebianHeya gang15:40
Quintasanbddebian: \o15:41
bddebianHello Quintasan15:41
=== Lutin is now known as Guest3058
paultagHey guys, I have a quick question about importing from debian -- do we use the dsc and modify the control / changelog and use that, or do we take the upstream tgz and recreate the debian folder, any links or help would be well recieved :)15:43
ScottKpaultag: Hopefully we don't make any change at all.15:49
ScottKIf some change is needed, it's use the .dsc, modify changelog, control, etc.15:50
paultagScottK, I'm doing upstream, and I wanted to submit it for Ubuntu review15:50
paultagScottK, so it's OK to use the dsc and make small changes to control / changelog?15:50
ScottKYes.15:51
paultagScottK, thanks :)15:51
=== Guest3058 is now known as Lutin
machinaanyone know how to make a diff that would remove binary files?16:08
persiamachina: You can't.16:09
Laneyyou can't remove files, only reduce them to emptiness16:09
persia(limitation of patch)16:09
persiamachina: You could delete stuff in debian/rules clean: (which is what some people do)16:09
persiaBut if they are non-free binaries, you need to repack or complain to upstream.16:09
machinathat's kinda upsetting...16:10
machinathey aren't non-free, it just looks like the author, run ./configure or make in his new upstream16:11
machinashould I just not worry about the object files & binary file then?16:12
machinaI'll send a message to upstream though16:18
=== Quintasan1 is now known as Quintasan
EmanemHi all, I have an issue with Ubuntu deb creation17:44
Emanemis there anyone able to help?17:44
persia!ask17:45
ubottuPlease don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-)17:45
Emanemok... I'm assembling by hand a deb file17:47
persiaAh.  You probably don't want to start from there :)17:47
persiaDo you have source?17:47
EmanemI've created the debian-binary and the archive with control (control.tar.gz) and data (data.tar.gz)17:47
Emanemno17:47
EmanemI'll release the source as tar.bz2 plus just 2 packages for ubuntu 32 and 6417:48
Emanemand I just want to created a deb file with 1 only executable17:48
azeemEmanem: assembling by hand is discouraged17:48
Emanembtw I've coded and compiled the binary17:48
Emanemwhy?17:48
Emanemthen what can I use?17:49
azeemit's not easily reproducable17:49
Emanemthis software is .2 and wanted to release the source plus 32 and 64 deb just for testing purposes (eg for people that don't want to recomile it)17:49
Emanemso which tools can I use?17:50
persiaEmanem: Create a source package, and then get it built somewhere to generate your .debs.17:51
Emanemwhy do I have to do so?17:51
Emanemmy deb consist of 2 dependencies and 1 only executable17:52
Emanemplus I'm using a simple Makefile to build my project, don't want to include the ./configure etc etc17:52
geserEmanem: how can I be sure that the binary you build matches the source?17:53
wrapsterhi all.. when i dist upgrade is done this is the error I get... http://pastie.org/74746617:54
wrapsterits not ubuntu ... but related to packaging so asking here...17:54
wrapstercould anyone please help me understand what this means and about any ideas as to how i can resolve it?17:54
Emanempersia , geser : I don't think honestly, I'll provide the sources with GPLv3, so people will be able to rebuild themselves. Packages are only for people which don't want to rebuild it17:55
Emanemthat's why I want to include an executable17:55
persiaEmanem: See, you create a source package, and you have it built, and users can get the binaries.17:55
EmanemI'm getting this error btw: http://ubuntuforums.org/showthread.php?t=135647117:56
persiaWe just don't have any good tools for assembling binary packages directly.17:56
Emanemsorry but doesn't the deb format rely on ar ?17:56
EmanemI don't understand this then... how can people like Skype provide proper packages?17:57
geserEmanem: and an user can also use the source package if he needs to modify your program (e.g. apply a patch, build with newer libs, etc.) to build a new deb17:57
Emanemgeser :  the program will be released under gplv3, so they'll be free to do whatever they want :)17:58
Emanemthe point is, there must be a proper guide/howto to explain the format/which tools/programs are used to create proper packages17:58
geserEmanem: sure, but without a source package it harder to them to build an updated deb (or do you ship a script to hand-build a deb?)17:58
Emanemgeser: latter could be... but even not, the program is simple17:59
persiaOr better stated: it's harder for *you* to build a binary package.17:59
Emanemso, what am I doing wrong? the error I'm having when I try to install the deb is : dpkg-deb: file `/home/ema/TEST/packages/TEST-0.2_amd64.deb' is not a debian binary archive (try dpkg-split?)18:00
geserEmanem: Skype probably use a source package too, they just skip the compilation during the package building and install the pre-built binaries18:00
geserEmanem: how did you assemble the deb? (the last step)18:01
persiaI thought Skype used a source package and just didn't share the source.18:01
Emanemgeser: "ar -r mypackage-2.0_amd64.deb control.tar.gz data.tar.gz debian-binary"18:02
geserpersia: I don't know, I just assumed18:03
persiaEmanem: We're not really prepared to help you do it that way.  None of us do it that way.18:03
EmanemSorry guys, but can you please point me to the script that is assembling the last step? I fear that I have to manually modify the "archived" (ar -r) file18:04
geserEmanem: although debs are 'ar' archives, dpkg doesn't like all (don't know the details)18:05
Emanemand how are these created? which tool?18:05
persiaWe use debuild or dpkg-buildpackage on source packages to generate binary package.18:06
persiaSo we really don't know the manual process.18:06
persia(or we use sbuild or pdebuild or upload to some archive)18:07
Emanemsorry but there must be some sort of tools used to produce proper deb if ar can't be used...18:07
geserin the end dpkg-deb creates the deb, but as persia said, without a reason nobody looks at the low-level steps to build a packages18:08
persiaEmanem: You really will find your life easier if you just create a few files in debian/ in your source and run debuild.18:09
persiaIt's not a clean package, but it's easier than trying to create binary packages manually.18:09
Emanempersia: what is debian/ ? maybe I'm missing that bit18:10
persia!packaging18:10
ubottuThe 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 !backports18:10
persiaEmanem: Take a look at our packaging guide.18:10
persiaBasically, to create a package, unpack your source tarball, create a debian/ directory, add copyright, changelog, control, compat, and rules, and build.18:11
persia(it can get more complex, but that's the high-level overview)18:11
Emanemubottu: the guide is broken18:11
Emanempersia: do you know where can I find the online sources of dpkg-buildpackage so I can look them myself?18:14
Emanem(I'm tempted to release the 2 binaries without any deb package...and btw this is why devs have hard time doing these things... how comes that one has to use 3/4 scripts to basically create an archive file (deb)?)18:16
persiaapt-get source dpkg-dev18:16
geserEmanem: http://git.debian.org/?p=dpkg/dpkg.git18:16
Emanemthanks geser18:16
geserEmanem: http://git.debian.org/?p=dpkg/dpkg.git;a=blob;f=scripts/dpkg-buildpackage.pl;h=cf39187aa0726ef68b034012828f0920df80aeaa;hb=HEAD18:16
EmanemThe error I get is thrown here: http://git.debian.org/?p=dpkg/dpkg.git;a=blob;f=dpkg-deb/extract.c;h=cd8332282f6261903af6d62d4fad3667a1931474;hb=HEAD#l14718:19
EmanemLOL found out the error18:21
Emanemdpkg expects the first file of the archive to be debian-binary... omg...18:22
Emanemso even order matters...18:22
Emanemwell now I'm getting another error... I guess I'll source-debug dpkg and eventually find out all the issues and fix them...18:24
Emanemthanks guys for the dpkg sources18:24
=== yofel_ is now known as yofel
persiaEmanem: You really might have an easier time with a source package :)18:25
Emanempersia: I'm more a software dev/computer scientist and actually when something doesn't properly work I really enjoy dig, understand and (possibly) fix18:26
Emanemthanks for suggestions anyway :)18:26
Emanempersia, geser: YESSS! I did it! :-)18:30
blackxoredI need an sru for azureus in karmic18:31
ScottK!SRU18:31
ubottuStable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates18:31
blackxoreds/I need an sru/I need to do an sru18:32
blackxoredthe azureus issue18:33
blackxoredis confirmed18:33
blackxoredcan't run 4.2.0.8 in karmic anymore18:33
blackxoredso I got 4.3.0.0 in lucid18:33
blackxoredhow to proceed?18:33
blackxoredthis bug18:33
blackxoredBug #48850718:34
ubottuLaunchpad bug 488507 in azureus "version upgrade to 4.3 required - HD network no longer working in current version" [Undecided,Fix released] https://launchpad.net/bugs/48850718:34
Emanempersia, geser: One last question: are the following libraries [libstdc++.so.6 , libm.so.6, libgcc_s.so.1, libpthread.so.0 ] installed by default on every Ubuntu system, am I correct? I guess nothing would run without those...18:34
ScottKblackxored: Generally SRU should be a patch targetted to solve a specific problem.  You should probably discuss it with someone in ubuntu-sru.  IIRC jdong is both on ubuntu-sru and familiar with azureus.18:37
blackxoredScottK, thanks18:37
geserEmanem: you can assume that. they are so low in the dependency chain, that nothing is left if you try to remove them18:37
Emanemgeser: I definitely would imagine that after 10+ years c/c++, but you know, better double check than have bad surprises...18:38
geseryou might need to check if you need special symbols that only appear since a specific version18:40
awelool, james_w: do either of you guys know what the deal is with merge-o-matic ( merges.ubuntu.com )?  is it supposed to be up-to-date?18:42
james_wI think it keeps falling over18:42
aweI did a merge two weeks ago for gst-plugins-bad, and it hasn't updated...18:43
geserlast I heard MoM still doesn't like format 3.0 packages (even with the backported dpkg), don't know if it got fixed18:45
tombeeAre there any opportunities to get involved with security programming in Ubuntu? :)20:04
keestombee: sure thing.  see https://wiki.ubuntu.com/SecurityTeam20:06
tombeeWow great, thanks kees20:06
keesnp.  we hang out in #ubuntu-hardened  (redirected from #ubuntu-security)  see the bottom of that page for the GettingInvolved link20:06
tombeeThe thing is I'm not that experienced in software security, but it's something I'd like to gain experience in and educate myself in :)20:07
tombeeSo I'm not sure how I would go about 'getting started'20:08
tombeekees: the roadmap looks good though :)20:11
RhondaI wonder… I merged multiple accounts into my rhonda account. With that came various mail addresses. If I remove those addresses, will the merge be undone, or would that only affect future stuff that would get imported with those addresses?21:48
geserI guess you need to ask in #launchpad to get an answer (or even file a question)21:51
persiaRhonda: I'd recommend just hiding the addresses, unless there's some special reason to forcibly remove them.21:55
RhondaYeah, they are hidden anyway.21:55
Rhondageser: Oh, right. Always forget about that special-purpose channel. Sorry. %-/21:56
* wgrant is concerned that it's "special-purpose", and not for a different product, as is reality.21:59
=== jldugger is now known as pwnguin

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!