[00:00] I'm downloading several version of Ubuntu to run in virtual pc's to make packages for my program www.globulation2.org . I'm wondering if theres any tips/considerations I should know about before hand? I know that there are already glob2 packages in the motu, but these packages are outdated, and outdated clients can't play on our hosted online gaming [00:04] genixpro: is the version in hardy (0.8.23) outdated as well= [00:04] ? [00:06] sistpoty, We are at 0.9.2 right now, and very soon we are going to release 0.9.3, and we want 0.9.3 to be a major release to re-stimulate our community, so we want every major distribution to have it on release day, to have packages for everyone, etc.. [00:08] genixpro: well, we're very, very close to release right now, and we cannot easily change s.th. that has been released again. Hence we try to stabilize every package some time before release (which is coming to the final phase now) [00:09] sistpoty, what do you mean by stabalize? [00:09] genixpro: well, find out if it has bugs, and fix these [00:09] sistpoty, so you have coders that go into the actual programs themselves? [00:10] genixpro: no, not at that depth. everyone can test the development version and we rely on bugs getting reported [00:11] genixpro: which we usually forward to upstream [00:13] sistpoty, ok. If I wanted to make my own, external package that we could distribute on our own website, how would I go about it? [00:13] genixpro: though it looks like, that noone really took much care for glob2 yet (as we inherited it from debian, and obviously noone really took too much care: https://launchpad.net/ubuntu/+source/glob2/+bugs) [00:14] genixpro: I guess, I'd start with the latest debian release (same as the one in ubuntu) [00:14] Well, we do our own debian package and release it on our website, but our packager, who used to be the lead programmer and maintainer, is doing nothing [00:15] genixpro: that's sad to know [00:15] Quite. I'm being paid to maintain the project by some rich guy whos in love with the game, so making packages is one of the tasks. [00:16] genixpro: I'd grab your debian packaging, and copy the debian directory to the new version, then "make -f debian/rules binary" and see if a package falls out [00:17] sistpoty, we don't use make, we use scons, I recall the debian package scripts where modified for it to [00:18] genixpro: maybe... but for a debian package, the script to build it is always a makefile (debian/rules). This one may then in turn call scons to build the actual program [00:18] sistpoty, I see. [00:19] sistpoty, So your reccomending I get the latest version of debian, update and make our debian package, and release it as both Ubuntu and debian package? [00:19] sistpoty, Should I make seperate ones for testing, stable, and unstable? [00:19] genixpro: no, it was just a hint to get you started with building a debian package, but I may have a better idea [00:21] genixpro: we have a team for Debian and Ubuntu to manage game packges, which is hosted as a debian team. Maybe you could write a mail to them? I'm quite sure, they'll try to get glob2 updated in debian then, so that we can easily sync it after hardy is released [00:21] genixpro: mail is: debian-devel-games@lists.debian.org [00:23] sistpoty, ok, that sounds like a plan. I've been told by Fedora to do the same thing, file a bug report on the package glob2 [00:24] genixpro: actually the team can help you with packaging as well, in case you want to do it yourself... at least to some degree, I guess. [00:29] sistpoty, just sent them an email === danielm__ is now known as danielm === TheMuso_ is now known as TheMuso === danielm is now known as danielm_ === danielm_ is now known as danielm [01:18] is ubuntuwire down again? [01:20] LaserJock: people.ubuntuwire.com is. [01:20] LaserJock: works for me [01:20] But that has an outage message. [01:20] qa. is down for me [01:20] it says it'll be up by the 2nd ;-) [01:21] Shh. [01:21] qa shouldn't be. [01:21] Ah, qa2 might be. [01:21] it kinda helps doing QA if, you know, we have the info ;-p [01:22] Indeed. What do you need? I can grab most things from the backups and run them on stratos. [01:24] rcbugs is already there. [01:26] where can I get at the rcbugs? [01:26] qa.ubuntuwire.com [01:26] There should be a link. [01:26] but umm, it's down [01:26] Uh. [01:26] You're right. [01:28] There, better. [01:28] I left the 2 out of qa2 for one TLD :( [01:28] Oops. [01:29] ok, so we have Univeres rebuild results and rcbugs [01:30] can we get FTBFS? [01:30] Sure, I'll set that up now. [01:30] do we even have current FTBFS data? [01:31] We can get LP FTBFS data quickly, but I haven't run an archive rebuild in a while. [01:31] * slangasek perks up [01:32] * Fujitsu wonders why. [01:33] Fujitsu: so if I were to ask infinity to do a universe archive rebuild, would the results be used? AIUI, this used to be done but nobody was doing anything with the output so it wasn't warranted to continue committing the buildd resources [01:33] slangasek: We fixed a lot of python-central FTBFSes with the last archive rebuild I did. [01:33] We can do it in about 5 days assuming things go well. [01:33] (not with Canonical buildds) [01:35] slangasek: I'm interested in getting that data used, however I don't know how successful I'll be [01:36] it can yell and scream 'till I'm blue in the face but if people don't respond not a whole lot will happen [01:36] I thought you didn't have resources to do rebuilds of more than main/restricted. [01:36] s/it/I/ [01:36] Fujitsu: well, ok. In theory using the DC boxes would also mean being able to get FTBFS info on some/all of the ports too, if that makes a difference [01:37] Right, that's one thing we can't do. [01:37] Fujitsu: well, we don't have the resources to spend on doing rebuilds of universe if the output isn't going to be used, certainly; but if it is, I can talk to infinity and see what's what [01:39] slangasek: that would be most excellent, so I also remember as Fujitsu wrote, that we were denied this in the past due to lack of resources [01:39] s/so/though/ [01:39] sounds like the problem is "waste" rather than "lack" of resources [01:41] there are 3 things I'd love for us to really hit as much as we can before release: [01:41] 1) clear the u-u-s queue [01:41] 2) FTBFS [01:41] 3) rcbugs [01:41] that would be a big QA boost for us [01:41] LaserJock: manpower is certainly a precious resource, even IS manpower... :) [01:42] oh yeah [01:42] well, /me would like to see bugs tagged with a good priority... however I already considered going through ~9000 bugs and then did s.th. else that weekend [01:42] sistpoty: yeah, my feeling is that we could really be doing better with Importance and milestoning [01:43] There is of course the importance usage debate that makes importances useless. [01:43] If importance is distro-wide, then everything in universe should be low. If not, we've been abusing it. [01:43] Fujitsu: which one is that? [01:43] LaserJock: yes, but I think its out of scope for motu-release alone... any clues how to acheive that? [01:43] Fujitsu: I would say that the teams responsible determine the importance [01:43] Fujitsu: for motu-release it should be universe/mulitverse wide [01:43] Importance is for the people doing the work [01:44] however, from the standpoint of slangasek et al. I'm concerned about our milestoning messing up his [01:45] There's no way to filter milestone lists, is there? [01:45] that's why I've hesitated to use it so far [01:45] Fujitsu: not that I'm aware of [01:45] LaserJock: setting importance is not a problem at all; nominating bugs for release is also fair game [01:45] (and accepting them) [01:45] but there's no way to distinguish Universe from Main bugs in the milestones [01:46] so I'd rather not make the milestones useless for the core devs [01:46] the goal is that milestoning of bugs will only be used for bugs that should be blockers for the milestone; that needn't be restricted to main, I'm happy to see milestoning used for bugs that are blockers for the community flavors as long as there's an actual committment to treating them as blockers in the corresponding project [01:49] (the current milestone list is a total mess in this regard, meh) [01:50] LaserJock: LP FTBFS is working on qa.uw.c now. [01:50] slangasek: ok, I agree with that [01:52] I just think MOTU in general need to get better at identifing targets and making sure they get done [01:52] we've had some really good successes lately IMO with transitions [01:52] LaserJock: WTF did oyu do that for?? [01:52] LaserJock: now you're gonna jinx it! [01:53] :) === danielm_ is now known as danielm [01:53] heh [01:54] hmm [01:54] so Main has 1 FTBFS and Universe has ~50 for i386 [01:55] Multiverse has 22 [01:55] that's interesting [01:55] * sistpoty looks innocent for atlas *whistle* [01:55] haha [01:56] We really need to convince people to work on rcbugs. [01:56] yeah [01:57] I was just gonna say, the scariest one is rcbugs [01:57] Fujitsu, LaserJock: maybe a mail to -motu would help= [01:57] It is very scary. [01:57] damn, s/=/?/ (again) [01:57] I was thinking of announce a RCbugs day [01:57] Or three. [01:57] well, yeah [01:57] we only have a few days [01:58] if I could I'd stop all uploading for a week to focus on these topics [01:58] but well, that's not gonna happen [01:58] One we're in FinalFreeze (FiF?), people have no choice but to fix bugs. [01:58] Maybe people will look at it then. [01:58] true [01:59] but my hope was to get as many in before the freeze as possible [01:59] LaserJock: want to write a mail about rcbugs? [01:59] just to make the release teams life easier [01:59] sistpoty: will do [01:59] LaserJock: thanks, excellent! [01:59] I gotta run home [01:59] then I'll do it this evening [01:59] :) [01:59] bbiab === Spec is now known as x-spec-t [02:10] Whats the recent process Debian have introduced for maintaining individual packages called again? [02:11] Debian Maintainers. [02:11] Right I thought so, but wasn't sure. Thanks. [02:14] oh, TheMuso: are you ok with dealing with ubuntu-studio bugs or do you prefer a delegate? [02:14] sistpoty: I practically upload everything ubuntustudio related, so I don't mind doing it. [02:15] TheMuso: excellent, thanks! [02:15] TheMuso: maybe you'd like to follow up to my mail? [02:15] sistpoty: I will do. [02:15] thanks! === luisbg__ is now known as luisbg [05:36] man-di: ping [05:47] Hello, how much time does it take to approve creation of a mailing list ? [05:47] oh man [05:47] that varies a lot [05:47] I had one take 2 months, but I think that's kind of an extreme case [05:48] From an eternity, up to three eternities. [05:48] LaserJock: well, I think my case is getting close to 2 months indeed [05:49] I'm not sure what's so hard about running newlist, really. [05:49] hmm, things haven't changes so much I guess [05:49] I thought they had sped up the process [05:49] since my last one [05:49] Ha. Hahaha. Hahahaha. [05:50] ok, regarding the packages usplash-theme-ubuntume & ubuntume-themes, both have bugs that are fixed [05:50] for the usplash, the boot text was overlapping the logo for 800x600 resolution [05:51] for the themes package, the cursor theme would make problems if compiz was running [05:51] does fixing those bugs qualify for a freeze exception ? [05:51] I think my mailing lists only showed up after I didn't want them anymore [05:52] and even then they weren't the names I requested, but oh well [05:55] um, what about the freeze exception question ? [05:56] what do you have to have an exception for? [05:57] I wouldn't think those would be considered features [05:57] LaserJock: 1. for the usplash-theme-ubuntume package, the boot text was overlapping the logo for 800x600 resolution [05:57] 2. for ubuntume-themes package, the cursor theme would make problems if compiz was running [05:57] AnAnt: no, I mean what about that would require an exception? [05:58] LaserJock: I mean, can I upload the packages after I fixed those bugs ? [05:59] LaserJock: or they can't be uploaded due to freeze ? [05:59] I would think so but I would ask whoever is in charge of ubuntume first [05:59] LaserJock: erm, I am the maintainer of those packages [06:00] oh, ok [06:00] I would think it would be fine [06:01] ok [06:03] thanks [06:06] I don't think people are going to mind too much for ubuntume packages [06:06] they're just metapackages and common distro stuff === Syntux_ is now known as Syntux [06:08] Good morning [06:08] hi Syntux [06:08] hey LaserJock :-) [06:08] still today for me [06:08] I haven't gone to bed yet [06:08] :-) [06:08] LaserJock, different timezone or linking days ? [06:09] timezone [06:09] I slept less than three hours, when we were talking yesterday it was around 1:00 AM. [06:09] I last talked to you at about 3pm my time I think [06:09] Good morning [06:09] Good morning warp10 [06:09] warp10! [06:09] warp10: did I say congrats yet? [06:10] Hey Syntux! [06:10] bad me if I didn't [06:10] LaserJock: np, you just did! :D Thank you! :) [06:10] warp10, congrats but what for? [06:10] been busy with dissertation work [06:11] * Syntux being nosy [06:11] Syntux: for being accepted as a MOTU [06:11] Syntux: warp10 is now a Master of the Universe :-) [06:11] he can leap tall buildings in a single bound [06:11] he's faster than a speeding bullet [06:12] warp10, oh! you got your Motorola man! cool cool congrats [06:12] LaserJock: and can mess up the whole universe... and multiverse too :P [06:12] Syntux: thank you! :) [06:13] :-) [06:14] warp10: I prefer to say that you can fix all science packages in Universe and Multiverse ;-) [06:14] I'm learning too to get become a Jedi [06:15] LaserJock: Indeed.... well, mayavi2 works fine now! :) [06:15] warp10: excellent [06:16] warp10: dude, you have waaay too many LP emblems now [06:17] LaserJock: Heh... :D [06:19] and some wicked karma [06:19] my karma has always been pretty crappy [06:20] Question: Do we need an access to different machines in case we are building a package for different arch? [06:21] good question [06:21] do you mean a package that only works on another arch? [06:21] or just testing if it works [06:22] both [06:22] well, we don't really have machines available for people to just log into and build [06:22] what you can do is ask here if anybody can build a package for an arch you don't have [06:23] once it's actually uploaded to the Ubuntu repos it will be built on all archs available or the ones you specify [06:24] LaserJock, ah but you upload it in deb format, right? [06:24] no [06:24] we only upload source packages [06:25] uploading .debs is not allowed [06:26] LaserJock, A developer uploads a source package and a MOTU is supposed to build it and upload it, right? [06:26] the MOTU only uploads the source package [06:26] Developer == MOTU [06:26] well, that too [06:26] a developer *should* build the source package but not for the purpose of uploading the resultant .debs :) [06:26] exactly [06:27] we do test builds before uploading [06:27] but we don't actually upload the .debs [06:27] oh [06:27] ok, assuming we are done building and testing the .debs what next? [06:28] well, if you want your new package in Ubuntu you need to get 2 MOTUs to review and accept it [06:28] then it will be uploaded into the NEW queue [06:28] that's one and what if it's already in the repos and we are building and testing debs of the newer version [06:28] and the Ubuntu archive admins will have a look and make sure it's ok [06:28] in general, find someone with upload rights to review your work and upload it into Ubuntu. In the case of a new package it must be approved by two developers. [06:29] yeah, if it's already in the repos then the MOTU will just upload it [06:29] LaserJock, upload the source not the debs? [06:29] yep [06:29] but then how do we get the debs? [06:29] oh, [06:29] well we have a build farm [06:29] Launchpad handles that [06:29] with cows and milk? [06:29] ah cool [06:30] so we upload, it takes the source packages and builds them on all archs, then puts them into the repos [06:30] Syntux: you can see them at https://launchpad.net/+builds [06:31] so basically what we do on our machines is for testing purposes only and once we feel confident about the package we throw the source online and then same operation perfumed on the build farm [06:32] yep [06:32] this way the environment building the .debs is consistent [06:32] and can be changed if need-be [06:32] and the process in the build farm, is it automated ? or humans having access to the farms to build the debs [06:32] no, it's automated [06:33] it would be a disaster to have them all built by hand :-) [06:33] hmm interesting, but what if something went wrong up there in the farm? [06:34] ah, then we talk to the buildd admins [06:34] 1~0.98.2~0~jdong1~7.10prevu1 that is the prettiest and yet most horrifying version number I've concocted yet. [06:34] we call the builder machines buildds (build daemon) [06:34] * LaserJock hurls [06:34] * Fujitsu strangles jdong dead. [06:34] ;-) [06:35] most of that is not my fault ;-) [06:35] ~0. Whhhy? [06:35] interesting [06:35] Fujitsu: to make absolutely sure I'm less than any Ubuntu versions that slip in without me knowing [06:35] banshee upstream calls their newest releases 1-0.98.2 [06:35] Just ~ should do it, shouldn't it? [06:35] which is insane enough [06:35] ~ is less than nothing. [06:36] Fujitsu: yeah [06:36] but if a ~0 is uploaded he wants to be lower than that? [06:36] I don't think there's a high chance of ~0ubuntu1 or ~0anything being uploaded [06:36] maybe in a PPA [06:36] PPAs might [06:36] If a ~0 is ever uploaded, it deserves to be overwritten by anything else. [06:37] I think between PPAs, backports, -updates, and -security it's just plain not fun to think about versioning [06:37] Backports are easy. [06:37] PPAs don't matter. [06:37] -updates and -security are a pain. [06:37] at least people want to be less than other people rather than higher than [06:37] PPAs matter if you're maintaining the PPA [06:37] :-) [06:38] or we can adopt some hackish pinning system of versioning :D [06:38] !jdon [06:38] Sorry, I don't know anything about jdon - try searching on http://ubotu.ubuntu-nl.org/factoids.cgi [06:38] !jdong [06:38] jdong: yes, but you're FULL OF CRACK! [06:38] hehe [06:38] having good conventions might work too [06:38] LaserJock: haha the problem is getting $restofworld to follow them :D [06:39] *coughs and points at that glib2.0 merge* [06:39] we just don't care about $restofworld [06:39] LaserJock: the way we show we don't care is to give em a pin of 0 ;-) [06:39] LaserJock, shall I signup for launchpad team and mailing list requesting a mentor or it's too early? [06:39] it's never too early to be addicted to Ubuntu! [06:39] isn't that the point of any standard, to force the crack of the few on $restofworld? [06:40] Syntux: it's not too early, although I honestly am not sure how active the mentoring program is [06:40] * jdong fully expects to break his system in his greedy quest for new mono apps [06:40] Syntux: you might want to to talk to norsetto when he's around. He's in charge of the mentoring program [06:41] jdong: but think of the kittens!! [06:41] LaserJock, hmm well, actually I'm deadly interested in holding a packaging jam, the idea was posted by Daneil http://syntux.net/blog/2008/03/13/ubuntu-jordan-loco-floss-introduction-the-jordanian-way/ [06:41] ops, no here http://syntux.net/blog/2008/03/06/jordan-loco-ubuntu-desktop-course/ [06:42] heh, that's kinda funny to see [06:42] I didn't know there was a Jordan LoCo [06:42] Soon we might get a permanent lab in one of universities so a crash course about packaging would make a difference [06:42] I might have to blog about that ;-) [06:43] LaserJock, haha [06:43] LaserJock: well you need to first twitter that you are blogging about it :) [06:43] jdong: I don't twitter [06:43] haven't figured out yet why I or other people would care about the minutia of my life [06:44] well I don't do this whole social networking thing either :) [06:44] I don't know why people would want to listen to me any more than they already do :D [06:44] SN is a waste of time [06:44] except Linkedin. [06:44] I do Facebook and LinkedIn because well, people send me invites [06:44] but I don't do anything on Facebook but collect invites [06:44] I'm not sure what I'm exactly supposed to do with it [06:44] I'm boring and just do LinkedIn. [06:45] and I don't know how it's exactly social or networking [06:45] coffee time. brb [06:46] *grumble* why does BSD always feel like I'm using UNIX from the 60's? [06:46] would it kill their reputation to bind the HOME key? [06:46] how would you even know you little punk [06:46] LaserJock: that's how I envisioned login worked back then.... username, password, "%" [06:47] and the console beep is the most lovely piercing tone! [06:47] I was talking about punch cards with my advisor today [06:48] it's amazing, he's only like 45 [06:48] his first computer had 4k memory [06:48] stuff like that [06:49] we were cleaning out the lab to make way for a new laser and found an ancient memory board (almost 2x as big as an ATX motherboard) [06:49] cool [06:50] no wonder he likes Fortran [06:50] it must remind him of the punch cards [06:52] alright [06:53] time for this old geezer to get to bed [06:53] night! [06:53] Night LaserJock. [06:54] goodnight LaserJock ;) [06:56] OOOOOH shiny mono 1.9 in Debian [06:56] * jdong now REALLY breaks his computer :) [06:56] What's so wonderfully terrificly good about Mono? [06:57] Fujitsu: *shrug* I still have some fascination with the .NET/C# runtime and framework [06:57] I guess that carries over from my final few days on Windows [06:57] That's unhealthy. [06:58] it's everything I'd want Java for but feels more natural and "done right".... [06:59] WHy would you want Java? [07:00] as a higher level language than C/C++ with managed runtime type features and not as much a performance hit as python/ruby [07:01] guys, can I build a package for hardy if I'm running gutsy or even dapper? [07:01] on the same machine I mean. [07:01] Syntux: you sure can, pbuilder supports doing that [07:01] cool [07:10] ./configure --build i486-linux-gnu --with-sigaltstack=yes [07:10] hmm [07:10] the typos do not instill confidence. [07:11] oh nvm that's supposed to say sigaltstack === Sebast1an is now known as Sebastian [07:21] G'morning [07:22] hi there [07:22] is universe subject to translation strings deadlines like main, and if so, when are they? [07:22] i just installed rubrica and noticed that there are some strings in the en_US version which are still in Italian [07:22] We don't have translations, so I doubt we're going to freeze for them. [07:22] oh ok [07:23] so um what would be the best thing to do about that then? [07:23] Translate them to en_US, I presume. [07:24] but how do i get the translations in? go upstream and then request a sync? [07:24] Probably a bit late for that now. [07:25] You'll have to modify them in the source package, and submit a debdiff. [07:25] ok [07:37] hmm it looks like these strings are hard coded :-/ probably best to find a way to move them into a .po? === doko_ is now known as doko [11:45] oh, drat, i was supposed to read backscroll, and forgot === TheGuru is now known as Cybermatt === stratus_ is now known as stratus [15:38] Hello, can someone review : http://revu.tauware.de/details.py?package=usplash-theme-ubuntume , it's a bugfix [15:42] AnAnt: you pinged me without content. Please add content to pings. How can I help? [15:42] man-di: oh, few days ago I was saying that the problem of gcc/g++ not being of same version on Debian is over [15:42] man-di: so any news about openjdk going to Debian ? [15:43] AnAnt: we are working on it in the ubuntu-java team. if you want to help please join. [15:43] man-di: how to join ? [15:44] man-di: ubuntu-java is doing it for Debian ? [15:44] checkout bzr and test, do fixes, etc [15:44] AnAnt: we are doing this for both distros [15:44] cool === ups is now known as ups2 === ups2 is now known as ups === LjL-Temp is now known as LjL === neversfelde_ is now known as neversfelde === Syntux_ is now known as Syntux [17:08] h [17:11] hello guys [17:12] i wanted to ask how do i get my python application into a .deb or sumthin and into the ubuntu software repository? [17:12] im really new to Ubuntu,hope you guys wldnt mind helping your fellow lad here [17:14] d3vt4r: lookie here: https://wiki.ubuntu.com/PackagingGuide/ [17:19] !packaging| d3vt4r [17:19] d3vt4r: 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 [17:20] thanks guys [17:20] i'll read up on it [17:29] ScottK, doko: are there any instructions on how to run a Python application under gdb in the wiki or somewhere, to point users to them when asking for a backtrace? If not, I could write something at https://wiki.ubuntu.com/Backtrace after checking with bdmurray if that sounds good to you [17:29] damn its daunting [17:29] i have a simple python application [17:30] i didn't realise i wld be bombarded by documentation [17:30] d3vt4r: check http://people.debian.org/~srivasta/manoj-policy/ too. I think that's more for modules though [17:30] d3vt4r: does it have a setup.py file? [17:30] wld anyone care to help me out with this? [17:31] not now, but im working on it [17:31] wld it make a diff? [17:32] wld? [17:33] would [17:34] i didn't understand your last question [17:34] as to would it make a difference if i had a setup.py rather then a main.py? [17:34] to get my application into the ubuntu software repository [17:37] pochu? [17:38]