=== Vorian_ is now known as Vorian [01:11] knix: When it (Eclipse) takes less than 19 hours to build on my amd64 box with 1 Gb of ram [01:12] hahaha [01:12] Makes for fun patching :) [01:12] RAOF_: get 4GB [01:12] well, these flac fixes are full of dismay. [01:12] you won't regret it [01:12] hello crimsun [01:12] hello [01:12] ajmitch: Maybe I should, at that. [01:16] crimsun: How so? [01:17] TheMuso: just looking at CVE-2007-4619 and the work necessary to backport the malloc() checks. [01:17] Multiple integer overflows in Free Lossless Audio Codec (FLAC) libFLAC before 1.2.1, as used in Winamp before 5.5 and other products, allow user-assisted remote attackers to execute arbitrary code via a malformed FLAC file that triggers improper memory allocation, resulting in a heap-based buffer overflow. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-4619) [01:17] sounds evil enough [01:17] Indeed. === _czessi is now known as Czessi === chuck_ is now known as zul === RAOF_ is now known as RAOF [03:18] quiet night [03:18] release time, so noone has anything to do [03:19] gutsy-proposed is calling you ... [03:19] Well, not ajmitch, since he doesn't upload anymore ;-) [03:20] correct, which is why I should disable my core-dev membership [03:20] thanks for reminding me to do so [03:21] Bah. [03:21] Gimme a few minutes and I can whip up a Dapper source backport for you to upload. [03:23] heh its been a few months since i've uploaded a main package, i'm sure that wil change in the next few weeks [03:23] speaking of sponsorship [03:23] pwnguin: need something ? [03:24] well yes [03:24] eventually [03:24] :) [03:24] ive been massagign thinkign finger, and figured id see what i could do to get it fixed up nicely in hardy [03:24] What a far cry from Breezy, heh. Uploading right up til the archive froze. Oh wait, I guess it's not that different after all... [03:24] just poke me, long as it isnt UDS week or OpenWeek i should have time [03:24] do you need extra acks when someone's offered a NACK? [03:25] nack? [03:25] crimsun: except that 'archive freeze' was about an hour before the isos were official & public [03:25] imbrandon: Negative ack. [03:25] crimsun: hehe yea breezy and dapper and feisty right untill freeze, gutsy is the only time i have "slacked" afaik [03:25] I have some package fuckage I did if someone is bored :-) [03:25] +edgy [03:26] err [03:26] pwnguin: afaik one nack negates all ack's, the package must be fixed [03:26] hmm [03:26] if the policy is still the same [03:27] thats sad [03:27] well, i havent officially recieved an NACK [03:27] so if im rather careful about it... [03:27] yea it rarely happens but when it does it needs to be fixed [03:28] the deal is [03:28] the bugs arent in thinkfinger [03:28] the knights who say NACK [03:28] haha [03:28] Surely the NACK is pointing out actual problems? So either you make an argument that it's too hard to fix and there's net benefit, or you fix it, right? [03:28] thinkfinger's pam module works fine with sudo [03:28] RAOF: right [03:29] it doesnt work so well with gnome-screensaver [03:29] btw howdy crimsun , ltns [03:29] Indeed. Howdie crimsun! [03:29] it feels strange to block a package because it doesnt integrate well with packages that DONT integrate with anything [03:30] So argue that point? [03:30] suppose i will [03:30] :) [03:30] pwnguin: not sure about that specific instance , but yea as RAOF said i would just argue that point and it should stand [03:30] but do you argue to the NACK guy or to some higher power? [03:30] s/argue/plead [03:30] "most" of the time the nack IS the higher power, but either [03:30] imho [03:31] does ubuntu regularly not include packages in debian testing? [03:32] no it should include all debian unstable , expiramental needs a specific "ask and you shall recieve" bug filed [03:32] automatic sync happens from debian unstable, which is where packages for debian testing come from [03:33] right. are there cases where ubuntu vetoes a debian decision to put a package in unstable? [03:33] a few cases [03:33] there's a blacklist of packages that don't get synced somewhere [03:34] afaik webmin is one [03:34] this would be handy for me to look at, i think [03:34] thats the only one i know of [03:34] imbrandon: how about all the kernel packages back in the day? [03:34] ahh yea [03:34] those too [03:34] webmin isn't in Debian [03:34] well [03:34] that doesnt count [03:34] so any blacklist for that would be legacy [03:34] kernels dont count [03:34] slangasek: yea , thats all i could rember though [03:34] :) [03:35] is there a period of time where people evaluate the blacklist and add things to it? [03:35] unlikely [03:36] however I'm not an archive admin so I can't say [03:36] i think its just done on a "as needed" hing === Whoopie_ is now known as Whoopie [03:36] thing* [03:36] ah. so things can become unpublished [03:36] slangasek: is the blacklist in the seeds file , or ummm [03:37] No, the blacklist is seperate. [03:37] figured so [03:37] actually, there is a blacklist in the seeds repo [03:37] and webmin is in it [03:37] what a surprise [03:37] Just in case... [03:37] AFAIK this is the blacklist in question [03:37] it's quality software [03:38] slangasek: ahh i was thinking i seen it on the list last i checked [03:38] but its been 6 months since i looked at the seeds [03:38] heh [03:41] just so i understand: packages published in unstable but not previously published in universe need to be looked at rather than automatically pulled in? [03:41] no durring autosync they are pulled in [03:41] pwnguin: Only after DebianImportFreeze as I understand it. [03:41] I believe that keybuk was giving them a cursory check before 'autosync' [03:43] ajmitch: Don't they just land in NEW. [03:44] ScottK: I'm not sure of the exact order of events with syncs [03:44] syncs MIGHT bypass NEW, i dont rember [03:44] I know that requested syncs hit NEW. Cause I had one rejected by (IIRC) Mithrandir. [03:45] I'd figure the automatic ones did too. [03:45] RAOF: Keep in mind the freeze names have changed [03:45] * StevenK kicks GTK and GnomeVFS [03:46] Oh, of course. I'll need to check those for Hardy. [03:46] StevenK: right, MagicalPixieDustFreeze is up first, no? [03:46] * RAOF kicks programs that don't do GnomeVFS (emacs, I'm looking at *you*) [03:47] RAOF: That'd rock. [03:47] ajmitch: Hmph. [03:48] Come on gvfs! Let me bask in your FUSE backend. [03:48] RAOF: easy to fix `sudo rm /usr/bin/emacs && sudo ln -s /usr/bin/nano /usr/bin/emacs` and stop using GNOME ;) [03:48] RAOF: I'll hold down imbrandon, you belt him. [03:48] lol === rob is now known as robii [03:48] get back under your bridge [03:49] Hrm. [03:49] Although sshfs does much of what I want emacs to do. [03:49] What's the GNOME API browser thing called? [03:49] devhelp? === robii is now known as rob [03:49] Yeah. [03:49] Great, ta [03:49] RAOF: oh, is gvfs going to be FUSE based? [03:49] * ajmitch really can't remember, having not used it [03:50] pwnguin: No, but on linux it'll have a FUSE linkage. [03:50] it would be neat if nautilus could set up something so that mplayer would play nice [03:50] pwnguin: It *can't* be FUSE based, because FUSE doesn't work on !linux [03:50] s/work/exist/ [03:50] thats what i thouhgt [03:50] i didnt think they hated BSD quite that muhc [03:50] pwnguin: it uses fuse [03:51] but it isn't fuse-based. [03:51] it also can use samba [03:51] awalton__: Uuuum. In what way "can use samba". [03:52] well, it's a VFS, so it can mount samba shares [03:52] Ah, right. [03:52] the code already works for that, I've been running Gvfs for a couple of weeks now and it works great. [03:52] cifs in the fstab for the win, fsk what the app supports, if its mounted it better "see" it ;) [03:53] That's not what we're talking about. gvfs will *export* it's stuff over FUSE. So apps that don't use gvfs can still work. [03:53] So what happens for non-Linuxes, like solaris? Will they still need gnome-vfs? [03:53] Yes [03:53] They just won't get the FUSE linkage. [03:53] Right. [03:54] it seems liek the BSD's are picking up FUSE too [03:54] The idea is that everything that can use gvfs (ie: all gnome apps) will. And the FUSE stuff is just to make non-gnome apps less annoying. [03:54] at least desiging fuse compatible API's [03:54] bah, i just say mount what you need [03:54] so that'll be cool when it all works out [03:55] kde trys that shit with media:// and such, vfs's need to diaf [03:56] imbrandon: But they really do make things easier for users, particularly when you just want to browse a network, and grab a file quickly. [03:56] Mounting is just too much screwing around for something so small. [03:56] * ScottK just uses sftp and doesn't worry. [03:56] Or to browse a network, *edit* a file quickly... [03:57] RAOF: That too. [03:57] theoretically, the vfs could just be a smart layer over mount.. but practically it doesn't make a lot of sense. [03:57] It's kinda annoying to scp, edit scp [03:57] I'm the same in that anything I use regularly, I mount, other than that, vfs FTW. [03:57] gvfs may even end up cross-desktop, too, from what I remember. [03:57] i definately dont want my laptop mounting cifs on boot [03:58] untill EVERY app supports the vfs without interaction, e.g. a lower layer , they will be crap imho [03:58] well thats what fuse is for [03:58] its a vfs in userspace api [03:58] with kernel support [03:58] VUSE! [03:58] well gvfs/gio was designed to go into glib, to supercede the posix-based stuff, so that any one who just wants to open a file will use it. [03:58] VUSEWKS [03:59] (of course the posix stuff will still be there, so there will always be apps that won't use it) [04:00] if i cant copy from the address bar of my app and paste it to a cli and use it, it wont fly imho [04:00] and vice versa [04:00] which again, is what fuse provides [04:00] sweet, it it works that way great, i just have never seen it in action === Vicfred is now known as Vicfred[Gutsy] [04:01] you've never used sshfs? [04:01] imbrandon: definitely try it out [04:01] it's great. [04:01] fuse is awesome [04:01] pwnguin: only via fstab, i only use a FS if i can mount it in fstab [04:01] imbrandon: FUSE is the best thing ever :) [04:01] imbrandon: FUSE can be mounted in fstab [04:01] jdong: i know [04:01] sshfs keeps me alive :) [04:01] the strike against FUSE is that it requires kernel support [04:02] pwnguin: agreed.... [04:02] but what it does is allow user space apps to present a mounted FS [04:02] via the fuse kernel module [04:03] how is that an issue though, load fuse once, use any fuse based FS right ? [04:03] whenever a write comes in to that fs, it's passed on to the user space program that registered itself to handle it [04:03] its an issue when you run GNOME on BSD [04:03] or Solaris [04:03] linux aint the only kernel in town ;) [04:03] or win32 or osx ? [04:03] heh [04:04] solaris is far more common, I'd say [04:04] well, osx has a fuse driver [04:04] i run kde on darwin quite often ;) [04:04] well, it doesn't sound like it would be too hard to write modules for other OSes and then just use the same userspace code. [04:04] imbrandon: you're not normal [04:04] but I haven't looked under the hood of FUSE so I wouldn't know. [04:04] ajmitch: hehe [04:05] ajmitch: its actualy faster with the darwin kernel on my G3 vs a linux one [04:05] with KDE [04:05] dunno why [04:05] OSX and FreeBSD both seem to have a FUSE driver [04:06] but yeah, FUSE aims to provide a uniform API for userland FUSE FS'es [04:07] well [04:07] darwin ought to have a fuse driver [04:07] its based on mach [04:07] 'based on' doesn't mean that it'd be easy to hack something in [04:08] it does when mach is about filesystems in userspace [04:09] really, minix and friends feel like they've got it backwards when they talk about microkernels being safer / more stable [04:09] if my filesystem crashes [04:09] i dont want to restart [04:09] i want to cry [04:10] same goes for MM [04:10] hardware drivers, on the other hand... [04:10] wow the FUSE for darwin/osx was paid for by google [04:11] http://theappleblog.com/2007/01/12/google-releases-fuse-for-osx/ [04:11] yay google. [04:11] Surely that means OS X can also get NTFS 3g. [04:11] http://googlemac.blogspot.com/2007/01/taming-mac-os-x-file-systems.html === homanj_ is now known as homanj [04:12] i didnt know that guy worked for google [04:12] heh [04:13] amit singh: [04:13] Manager of Macintosh Engineering [04:28] . [04:31] * imbrandon wonders how much bandwidth ubuntu.com is pushing right now [04:33] im pretty sure drupal errors are cheap [04:35] So "partner" is now an Ubuntu component: https://launchpad.net/ubuntu/+source/opera/+publishinghistory [04:37] hrm [04:37] sounds alot like commercial [04:37] I think that's what they renamed it. [04:37] It is [04:38] ahh so its not a ubuntu component, its a canonical one [04:38] Well not the way it shows there. [04:38] well [04:38] It shows there as an Ubuntu component. [04:38] its a component supplied by canonical which is for ubuntu. [04:38] err lifeless WHAT!?! [04:39] Package source page says "opera in Ubuntu" [04:39] no commercial was on the canonical domain not the ubuntu one [04:39] Not for. [04:41] well now its on the launchpad domain [04:41] ScottK, lifeless : must be a bug in LP in the way it shows, it dosent exist in http://archive.ubuntu.com/ubuntu/dists/gutsy/ [04:41] wait [04:42] LP has a source package to opera? [04:42] no [04:42] pwnguin: yes, but a source package can contain binary of opera to package ;) [04:42] http://launchpadlibrarian.net/9522826/opera_9.23-20070809.6gutsy1.tar.gz [04:42] what imbrandon said. [04:43] e.g. i can package any binary, i dont nessesarly have to have the source to make a src deb [04:43] well i just saw "libx11-dev build-deps" [04:43] in the changelog [04:44] ScottK: no its still on the canonical domain , http://archive.canonical.com/dists/gutsy/ [04:45] they just renamed the component and there is a bug in LP in the way it represents it [04:45] imbrandon: Why do you assume it's a bug? [04:46] because it is not part of ubuntu, and it represents its self as if it is [04:46] I agree with that, I just don't know that someone else didn't have a different idea. [04:46] actualy it dosent say what its for, those components could be for anything [04:47] It does. https://launchpad.net/ubuntu/+source/opera says, "Opera in Ubuntu". [04:51] maybe there is the need to compile some stubs or something? === jdong_ is now known as jdonginator [04:51] dosent matter, opera is packaged in the canonical repo not the ubuntu one [04:51] JanC: ^ [04:54] imbrandon, I was thinking about why the build deps on -dev packages [04:54] and I installed it from the canonical repo already earlier this night [04:59] ScottK: confirm this for me please [04:59] https://bugs.edge.launchpad.net/launchpad/+bug/153798 [05:00] Launchpad bug 153798 in launchpad "canonical partner repo packages showing as "in ubuntu"" [Undecided,New] [05:03] Done [05:05] thanks [05:05] i just poked #launchpad about it, we'll see what they say [05:18] Oh, let's give them a URL one character per line, it'd drive them mad. [05:18] Good. Someon other than me complaining. [05:18] Someon/Someone [05:19] StevenK: Good idea. [05:19] I don't understand why people are so anxious & impatient today [05:19] It's midnight where I am, so it *must* be released NOW! [05:19] * ScottK doesn't understand why they don't just install the RC and then upgrade and relax. [05:20] Fujitsu: For a moment, I thought you were serious [05:20] StevenK: because we know that sabdfl will release it personally at exactly 00:00UTC [05:20] StevenK: Sure. [05:21] ajmitch: Which was 4 and a half hours ago. [05:21] that's of no concern [05:21] StevenK: The pipes are clogged. [05:21] though there was definitely a countdown then [05:24] * TheMuso notes the time till he has to start getting ready to head down to Sydney for the Sydney release party. [05:25] * ajmitch notes that he isn't going to any release party, and hasn't been to one yet [05:27] ajmitch: Why not? [05:28] because they never tend to happen [05:30] I can just waste time at home instead [05:31] * StevenK sighs. [05:31] Now there is talk about the whole world being the 18th before it releases [05:31] it happens every time [05:31] It does. [05:31] They never get it. [05:32] the more they refresh the closer it comes. [05:32] the just _know_ it [05:35] I also find it interesting that many casual users assumes that Ubuntu/Canonical must be operating on British time. [05:35] s/interesting/amusing/, I think. [05:35] Indeed [05:36] in recent times that's been quite true [05:36] a watched iso never boils. [05:36] Hi slangasek. [05:36] * slangasek moos [05:36] slangasek: Never burns, surely? :-) [05:36] * Fujitsu sends some grass to slangasek. [05:36] slangasek: enjoying the tension of the people waiting? [05:36] StevenK: get your own metaphor [05:36] * Fujitsu accidentally hits the red button on the way. [05:36] slangasek: I enjoy borrowing other peoples [05:36] morning slangasek. so you're not hungover, then [05:37] a little early to be hungover, isn't it? [05:37] (also not morning) [05:37] Hobbsee: Not yet, you mean. [05:37] slangasek: you're with canonical - it's never too early, or late [05:37] slangasek: You are still living in the US, right? [05:37] ajmitch: Hahah. [05:37] ajmitch: that's drunkenness, not hangovers [05:37] minghua: yep [05:38] Hobbsee: Nice one. [05:38] Fujitsu: :D [05:38] true, being hungover generally requires that you actually stop drinking [05:38] ajmitch: haha. jono doesnt do that [05:38] Hobbsee: you should have been +o [05:39] * TheMuso ponders joining -releaseparty to see whats going on... [05:39] oh yeah [05:39] TheMuso: espeak won't be able to keep up! [05:39] Hobbsee: You *are* evil. [05:39] isn't it great? [05:40] TheMuso: I'd give you about 90 seconds in there... :-) [05:40] From what I heard in -devel, I think staying away from -release-party is good for my health. [05:40] wall of text crits for 8700 [05:41] Fujitsu: i know. [05:41] StevenK, Fujitsu, likely enough.. [05:41] Its easier to just watch the chaos [05:42] watching the chaos as Hobbsee stirs away [05:42] It's horrible! [05:42] Yep. [05:42] Hobbsee is doing a great job. [05:42] ajmitch: yup :) [05:42] this is the best part about release :P [05:43] yeah wait about 15 minuts till their buffers are full, then just go in hit the ubotu trigger for that and watch them all cry [05:44] Bah. Evil LaserJock. [05:45] "but it says RC on the download I want the real gutsy" [05:45] can't wait to see that one. [05:56] That #ubuntu-release-party is pure madness. [05:57] craziness! [05:57] I'm glad that craziness is in there and not #ubuntu itself [05:57] loads of fun [05:57] jcastro: yo [05:58] aj!!!! [05:58] ugh cant sleep ( again ) [05:58] heya whip [05:58] hehe [05:59] imbrandon: Read some autoconf macros. Apparently, they can induce a coma [05:59] lol [05:59] imbrandon: welcome to the insomniac's club [05:59] i could compile hurd, bddebian hints its a bedtime story [06:00] I thought you wanted sleep, not pain [06:00] jcastro: btw i ment to ask , why the nick change? just something diffrent ? [06:00] ajmitch: heh [06:00] if you're just looking to pass the time, http://ppa.launchpad.net/jldugger/ubuntu has several games i pulled from debian [06:01] imbrandon: "jorge" was taken, and like I was telling aj, a nick based on an old shitty sega game that sucked wasn't so inspiring [06:01] hehe [06:01] yea , "brandon" [06:01] was taken too [06:02] * TheMuso is off. [06:02] bye [06:02] See ya! [06:02] What, someone mentioned Hurd? :-) [06:02] as a way to cure insomnia [06:03] don't get all excited [06:03] heh [06:03] RAOF: Will I be seeing you tonight? [06:03] hehe [06:03] gnight TheMuso [06:04] TheMuso: Probably not at this point. Work and all. :/ [06:04] oh well [06:04] * TheMuso -> gone [06:04] Have fun anyway! [06:05] Gnight folks [06:08] * ScottK despairs at ever making secvpn actually work. [06:19] RAOF: come anyway, dammit. [06:20] * ajmitch would like to be there, but can't [06:21] Hobbsee: Maybe. [06:22] Hm. Someone's ubuntuforums sig is "you can enable hyperthreading on non P4 systems!!!!!1111" [06:22] Hobbsee: Are you going to be there? [06:23] RAOF, well, Sun's T1/T2 chips have some sort of hyperthreading too ;) [06:23] RAOF: yes. therefore, you have to come. [06:23] hrm i wonder if Ideastorm is a (tm) ? [06:24] JanC: :P [06:24] JanC: they have hardware threads, rather than the hyperthreading illusion [06:24] yeah, it's different, but in some ways it's similar too [06:24] though the whole chip is designed around highly parallel loads [06:25] * ajmitch got to put ubuntu on a T2000 at one point [06:25] HT, aka "our chips really, really hate pipline stalls. Let's make this happen as little as possible" [06:26] well, it seems like Intel learned to go another path :) [06:26] There's probably an 'e' in pipeline. Positioning it in the previous statement is left as an exercise for the reader. [06:26] RAOF: But that was because the pipeline was insanely long, right? [06:26] StevenK: Yes. [06:26] My CPU theory is very rusty [06:26] Because having a very high clockspeed was good for sales. [06:27] And for making your CPU dissipate >100W of heat. [06:27] mmm, toasty warm [06:27] I don't know the details, but I guess Sun makes sure taht they have enough I/O to keep data flowing through the CPU :) [06:27] And having the heatsink you sell by default only dissipate what the CPU can put out idle. [06:27] Oh wait, that was AMD [06:28] * ajmitch fondly recalls that athlon xp running at >90C [06:28] And that would be why. :-) [06:28] The P4 in my uni computer has a fan that has 3 settings: loud, worryingly loud, and 'geez I hope it's not going to die!'. [06:28] my brand-new C2Q Q6600 runs at 30 °C or something like that :) [06:28] RAOF: Loud, worrying loud, and "Geez, that fan left a big hole in the case, didn't it?" [06:29] coldest I've seen mine run at now was about 24C [06:29] and I can't even hear the fans in it [06:29] on a rather cold winters day [06:29] The first time I heard it on it's maximum settings I was seriously woried. [06:29] * StevenK hasn't tried to monitor the temperature of his AMD64 [06:29] (Yay Sydney summers.) [06:29] RAOF: (Agreed.) [06:29] yay dunedin winters [06:30] the Q6600 was running at 24°C in the BIOS too, but I guess it might be a little warmer now [06:31] That isn't accurate, since the BIOS doesn't make idle calls [06:31] you mean, it might be lower outside the BIOS? [06:31] * StevenK nods [06:31] that would be really nuts :P [06:32] Altough it's unlikely to be lower than ambient temperature :) [06:32] if they keep going that way, I'll have to isolate my next computer to prevent me from freezing [06:32] :P [06:32] I'd expect it to be a few degrees warmer than ambient [06:32] I suppose with a peltier it *could* be lower than ambient temp. [06:32] Spelling is for the week. :P [06:33] RAOF: Oh, wah [06:33] RAOF: Go back to marking, as punishment. :-P [06:33] evil [06:34] now, if they could invent better heat flow in laptops (my laptop shows 46-59 °C CPU temp most of the time while running) [06:35] StevenK: No marking this week! Test next week, then a couple of weeks, then *exams*! [06:35] and that's a mobile CPU, not a monster like this quad core thing... [06:35] hell my desktop run about 70C most of the time, well it did untill i replaced it this month [06:37] RAOF: Ah, so you won't want to scoop your eyeballs out until you're marking the exams? [06:39] Pretty much. [06:40] wb LaserJock [06:41] chaos, simply chaos [06:42] i'm just waiting for "Gutsy Released" to show up on digg [06:42] bah, I wonder if my desktop uses no more power than my laptop, while being about 5x as powerful :-/ [06:43] well, it probably uses more power, but when it would have a low-power CPU it wouldn't [06:44] and of course my desktop has no built-in LCD :) [06:47] good morning [06:50] good morning dholbach [06:50] hey luk_ === Hobbsee_ is now known as Foobar === Foobar is now known as LongPointyStick [08:16] Bug #153812 [08:16] Launchpad bug 153812 in ubuntu "Package startup-manager for repository and possibly add it to a ubuntu release" [Undecided,New] https://launchpad.net/bugs/153812 [08:17] Tag Bug #153812 as wishlist please. [08:17] Launchpad bug 153812 in ubuntu "Package startup-manager for repository and possibly add it to a ubuntu release" [Undecided,New] https://launchpad.net/bugs/153812 [08:21] good gutsy for everyone! :) [08:22] Hello, if I use reportbug will it report to Ubuntu or Debian ? [08:22] AnAnt: If it's the Ubuntu reportbug, by default it reports to Ubuntu. [08:22] AnAnt: on Ubuntu [08:23] ok, thanks ! [08:32] anyone here know anything about gnupg? [08:32] sure. it exists? [08:33] for some reason, my gnupg setup has decided to stop signing e-mail, .debs, etcetera [08:33] it says "pinentry-gtk2: Assuan processing failed: write error" :/ [08:46] nenolod: Are you on Gutsy (did you just upgrade)? [08:47] ScottK, the machine it is happening on is running 7.04 [08:47] OK. [08:47] Look at that, I just get in after a good night of sleep, and who is talking? The The-Sun-Never-Goes-Down-In-My-House ScottK! [08:48] nenolod: Look in ~/.gnupg/gpg.conf and see if it there is 'use=agent' in it. [08:48] norsetto: Yeah. I REALLY need to get to bed. [08:48] ScottK, no use=agent. [08:49] nenolod: When you are signing, I think, pinentry only gets called if gpg-agent is involved, so my suggestion would be to add that and install gpg-agent if you don't have it. [08:50] nenolod: Althought this is a Kmail specific page, the gpg part is generic: https://help.ubuntu.com/community/KMailGPGAgent#head-c6757d6675d3932eaeffb136479725842a81d9b6 [08:50] * ScottK really goes to be bed this time... No, really. [08:53] oh ffs [08:53] it ate my key [08:53] Ate your key? [08:53] yes [08:54] i created a new key, and it uses that one fine [08:54] :D [08:54] * nenolod figures out how to issue a revoke request [08:54] * persia hopes nenolod either has a revocation certificate or never sync'd the first key [08:54] persia, yeah. i'm going to make a revocation certificate now- or try to. [08:55] nenolod: You need the secret key & the passphrase to generate the revocation certificate. Perhaps you could generate one for the new key, and put it in a safe place? [08:55] persia, sounds like a good idea [08:56] nenolod: Also, if you're not sure about your backup & revocation procedures, you might want to set your key to expire in a year or so: extending expiry is fairly easy, but lost valid keys are annoying. [08:56] persia, yeah. it expires in few months anyway. [08:57] oh ffs [08:57] it won't let me generate a new key now [08:57] giving: "gpg: problem with the agent: Invalid IPC response" [08:57] so maybe my key is ok [08:58] nenolod: Are you running an agent? [08:58] yes [08:59] nenolod: Try exiting the agent, and starting the process again. [09:00] persia, no change [09:01] nenolod: Hrm. I suspect something is funny with your user (or there'd be more bugs). Perhaps try with a test user (perhaps on console) to make sure that gpg is working properly on the system, and with an agent (perhaps in a nested X, or alternate console) to get the settings right? [09:02] (remember not to upload the test user key) === ogra__ is now known as ogra [09:04] persia, sec. [09:05] it works with another user [09:05] :/ [09:05] maybe restarting will fix it. [09:06] nenolod: Try logout/login before restart. [09:09] ok. [09:09] it works on console with my user. [09:09] brb [09:12] hmm. works with pinentry-curses, but not pinentry-gtk-2. [09:13] nenolod: Are you using GNOME or xfce? [09:13] xfce. [09:13] * persia refrains from recommending seahorse-agent [09:13] would take work with icedove? [09:14] that* [09:15] nenolod: Why? I won't recommend it because it depends on lots of libraries you probably don't want to load, but I'm not sure that the choice of agent and the choice of client are related in any way. [09:21] morning all [09:26] persia, well since it works, i can probably export my private key [09:26] and then just install gutsy on here [09:26] :D [09:26] nenolod: Your original key, or the new one? [09:27] original [09:28] Excellent :) Also, for peace of mind, you might consider making a hard copy of your revocation certificate. It's not something you necessarily want to be easily available electronically, but it is something you'll want to be able to access if your computer melts. [09:30] Hi all! [09:31] * minghua just backs up the whole ~/.gnupg dir. [09:31] (and hope that's enough) [09:31] minghua: Offsite? [09:32] gpg's man page is the most confusing one I've read. [09:32] persia: Yes, on a USB disk. [09:33] minghua: That's probably good enough. It's an interesting balance: some might say keeping your secret key offsite is an invitation for compromise to those that control the offsite facility, but just keeping revocation certificates offsite means one loses the ability to self-identify in the event of data loss. [09:34] Oh. I don't think I have such high security concerns. [09:34] If someone breaks into my home and have access to my GPG key backup, I have more important things to worry about. [09:36] minghua: "my home" is offsite? I was imagining non-human influences (EM field, fire, earthquake, etc.) [09:36] persia: And shouldn't people never let their key leave them if they really care about key security anyway? [09:36] persia: Oh. So "offsite" means like somewhere across the country? [09:37] * minghua is a pretty sloppy GPG key keeper. [09:38] minghua: Perhaps. Key maintenance is always a balance. I usually define "offsite" as "not in the same physical building", although I don't typically use adjacent buildings (and I've never been involved with a catastrophe with more than ~1km direct influence) [09:40] persia: I keep one copy on a USB disk, another one on my laptop. Is that really bad maintenance? [09:40] * nenolod waits for gutsy ISO to finalise [09:41] minghua: keeping one on your laptop is regarded as poor maintenance... [09:42] slangasek: But my USB disk partition is not encrypted, I always feel that there is a higher chance that I lose my USB disk than my laptop. [09:43] minghua: a laptop is a higher-profile target for theft [09:43] I realize it's poor maintenance, I just don't know how to do better. [09:43] That's true. I hear a lot of stolen-laptop-and-gpg-key stories. [09:43] minghua: Obviously, you should have a dedicated offline keysigning machine, in a secure facility, with several revocation certificates in alternate secure facilities :) [09:44] persia: Riiiight. [09:44] persia: don't believe everything you read on Manoj's website... [09:44] I think I'll delete my key from laptop, keep the copy on my desktop at home, and go back to using my USB disk with my laptop. [09:45] It will be a bit annoying to have to keep the USB disk with you, though. [09:45] More generally, it's good practice to restrict keys to location-specific devices, perhaps with some encryption. [09:45] persia, slangasek: Thanks for the suggestions. [09:45] Hobbsee: ping [09:45] slangasek: heh. I do like the keysigning protocol though. [09:45] persia: Location-specific devices? Like what? [09:46] minghua: "my desktop at home" [09:46] minghua: More basically, my recommendations and your published plan differ by the carrying of the USB key with the secret key. [09:47] persia: So I should always connect to the desktop at home and sign stuff there? [09:47] That's probably too much trouble for me. [09:48] minghua: Um. remote-network-accessible secret keys are a whole other level of dangerous. Perhaps a known safe master key, which signs a minor key, with which you sign routine things. [09:48] Actually, I think the first story that makes me thinking about the security of my key is from slangasek. :-) [09:49] minghua: On the other hand, the web of trust has been hacked a couple times, so there's a strong possiblity you aren't you anyway. [09:49] persia: Even IP-restricted, SSH-key login only remote accessible machines? [09:50] minghua: The possibility of an exploit to get into that machine is higher than the possibility of an exploit to get into a powered-down firewall'd device, but like I said before, it's about the level of protection you want for identity. [09:52] persia: Err... What is the difference between a powered-down firewall'd device and a powered-down device with no firewall? [09:52] minghua: Wake-on-LAN [09:52] Okay. [09:53] My home desktop won't be on 24/7 anyway. I think I'll stick with the "carry a USB disk with the whole key" plan. [09:53] Not perfect, but better than I currently have, I suppose. [09:54] How about 'gpg keys on LVM-on-crypto laptop' [09:54] ? [09:55] RAOF: better than keys-on-laptop, but as slangasek said, a laptop is a greater target than a USB key. [09:56] Is the danger really that greater for un-encrypted filesystem, even if I have a strong passphrase? [09:56] If the laptop is stolen, I'm going to revoke the key regardless, I think. [09:58] minghua: Sure. The attacker only has to hack ~128 bits of passkey rather than 1024 (or whatever) bits of secret key. [09:59] (plus most passphrases are within a 50-80 bit set due to the preference for printable characters, etc.) [09:59] morning party people [09:59] persia: And that's assuming they really want my identity instead of just my laptop, and they do it before I revoke my key, right? [09:59] People tend not to choose (or remember) highly entropic passphrases, yes. === zul_ is now known as zul [10:01] minghua: Sure. 80 bits takes about 16 hours on household gaming hardware. Where's your revocation certificate (don't tell me). [10:02] (plus, they probably want your laptop anyway...) [10:02] Yeah, so I'll pass that as "not a greater danger". === thekorn_ is now known as thekorn [10:02] Some other people's identity, maybe; but not mine. [10:03] (Hmm, I hope that quote don't come back and bite me when I apply for MOTU membership extension...) [10:03] minghua: Are you in the strong set? If so, the compromise of your identity could compromise others... (not that this is especially likely either) [10:04] persia: I think not. My key is only signed by one person. A DD, though. [10:04] minghua :D [10:04] minghua, i'll sign your key if you come to my house :( [10:05] nenolod: Thanks, but I think no. Unless you live in Houston. [10:05] More signature means more responsibility, after all. [10:06] persia: I think not. My key is only signed by one person. A DD, though. <-- (did you get this line?) [10:07] minghua: I didn't. I believe you need to have 1) received two signatures, 2) made two signatures, and 3) have two independent paths to all points in the strong set (although these not need be bidirectional paths) [10:07] s/not need/need not/ [10:08] That sounds a quite strong criterion. [10:08] minghua, well. we're on the same continent at least :)) [10:08] persia: Thanks for all this talk. [10:09] (Although I really don't think I have the poorest key maintenance among MOTUs...) [10:09] * persia is sure that's true [10:10] i doubt i would have time to be a MOTU anyway. i do have time to bitch at MOTUs when they do horribly stupid things in debian/rules though :P [10:11] * warp10 is away: AFK [10:11] * pwnguin ducks [10:11] nenolod: Care to give an example of the stupid things? [10:11] heh [10:11] * minghua hopes he didn't do too many. [10:11] nenolod: There's no rush. Any contributions are welcome. If you get your keys up, and join the contributor teams, we can probably roll out your fixes faster (regardless of any future MOTU plans) [10:12] i saw one use of local system data in a debian/rules [10:12] but it got rejected from debian archive [10:12] (and rightfully so) [10:12] nenolod: Not accidentally,and not from a build-dependency? [10:12] persia, intentionally yes [10:13] persia, afaik it was dependant on environment somehow [10:13] i forget what package it was, it was few years ago [10:14] !away | warp10 [10:14] warp10: You should avoid changing your nick in a busy channel like #ubuntu - it causes unrequired scrolling which is unfair on new users. The same goes for using noisy away messages : use the command "/away " to set your client away silently - See also !Guidelines [10:14] Bah, wrong redirect, sorry. [10:14] heh [10:15] persia, recently someone tried to package code of mine passing it off as GPL3 when the license was BSD-like [10:15] (luckfully, it got rejected from both ubuntu review and debian reviews) [10:16] (not that GPL is bad, but this particular code was not GPL, and it should have said so in debian/copyright) [10:16] * minghua really hopes those packagers nenolod mentioned are not really MOTUs but only hopefuls. [10:18] i'd like to see ubuntu take a different direction in packaging a few things than debian does [10:18] minghua: Life is balanced. That's why we review each other's work, and then receive more detailed review from the archive admins. [10:18] persia: Yes, on package reviewing, I like Ubuntu's system better. [10:19] * persia doesn't see much difference other than REVU [10:19] the way audacious is packaged in debian is crap [10:19] and the packager refuses to work with upstream in any way to ensure it is packaged in a way that is useful [10:19] Actually, is there any review except NEW packaging for DDs uploading new packages? [10:19] persia: And I am not condemning careless packagers, it's just about my faith on Debian/Ubuntu's quality. [10:19] persia: AFAIK no. [10:19] well, ubuntu basically lives with debian [10:20] pwnguin: Not really. We junk a lot, and change a lot more. Each release is a greater variance as we have time/effort to do things. [10:20] persia: When I say "Ubuntu's system", that includes REVU. [10:20] minghua: Right. mentors just seems difficult to use somehow [10:20] persia: when you junk something, does it stay junked? [10:21] pwnguin: e.g. freecraft [10:21] i'd say the 100% ubuntu packages are packaged more sanely than those imported from debian [10:21] persia: Another difference is Ubuntu's package usually pass though more pairs of eyes. [10:21] nenolod: We (try to) require 0 lintian or linda warnings (let alone errors), but there's still stuff that isn't quite sane. [10:21] persia: So obvious problem are more likely to be caught. [10:22] persia, i must wonder if i redid the packaging of audacious in a sane way, if it would make it into ubuntu [10:22] persia, the way it is packaged now is non-obvious [10:23] minghua: To some degree, although it's only 1 extra pair of eyes (MOTU packager + MOTU reviewer + archive admin) and the MOTU reviewer is optional from the last published meetings of MOTU Council (although still good practice). [10:23] nenolod: Does audacious packaging in Ubuntu satisfy upstream (I assume that includes you)? [10:23] minghua, it does not presently because it is based on the debian packaging, which puts most features that users want in the -extra package [10:23] uh oh [10:24] those features aren't mp3 support are they? [10:24] no. stuff like submitting to last.fm [10:24] nenolod: It's certainly possible. Before engaging in such an effort, I'd get in touch with some of the people who have touched the package, just to make sure your efforts would be in line with their ideas. [10:24] visualisations [10:24] nenolod: I see. I think audacious has enough user interests, so if you can point out the problems in packaging, I believe there will be MOTUs willing to work with you and fix it. [10:24] persia, I am in constant contact with le_vert, and he does not think last.fm is useful [10:25] and apparently, shoutcast support wasn't useful either [10:25] :D [10:25] nenolod: the ubuntu people that touched it [10:25] imbrandon, ah. that's vasillis and Adri2000 [10:25] nenolod: If the issue is -extras, there's at least some interest in Ubuntu in keeping it out, as a minimal audacious in main is a goal for several people. [10:25] persia, yes. my plan is to split each plugin into it's own package, like xmms is done [10:26] persia, that way users can install exactly what components they want [10:26] and gstreamer and countless others [10:26] compiz is packaged the same way [10:26] nenolod: That's a lot of new packages, but definitely interesting (especially if the new packages could be combined with existing xmms / bmp / etc. plugins at a source level). [10:26] fusion, -extras, -extras-you'reinsane [10:26] persia, xmms and bmp are dead and should be dropped [10:27] heh [10:27] hahaha you'll die if you try to get xmms out of ubuntu [10:27] hehe [10:27] thats a dead horse waiting to be beaten again [10:27] nenolod: For something of that scale, I'd suggest drafting a spec of the work to be done. This would get some feedback on the idea from a wider audience, and also likely get more hands to help. [10:27] :P [10:27] i know at least one guy who wishes to be kept [10:27] xmms [10:27] well, audacious isn't actually an xmms replacement [10:28] but audacious 1.4 is an entirely different beast and is not source compatible with xmms in any way [10:28] persia: I honestly thinking mentioning specs at this stage of discussion is rather discouraging... [10:28] Regarding dropping xmms and bmp: it's been discussed, but it's not as easy as one might think. We just recently finally got it out of main, but will probably be shipping for another year or so. Merging "plugins", and disabling xmms support as "unsupportable" and "bug ridden" might be a way to reduce the pain of a later shift. [10:28] minghua: nah specs can be carried over to the next cycle easy [10:28] minghua: Why? [10:28] persia, mostly the external thirdparty components [10:29] persia, are already packaged seperately [10:29] it's just audacious-plugins which is a mess [10:29] Ah. If the source compatibility is completely broken, forget the idea of merged plugins. [10:29] persia, imbrandon: My humble opinion is that, specs are really very Ubuntu-specific and internal stuff, and should be a concern of the MOTU, instead of the upstream developer. [10:29] well, being a motu might be interesting to me in the future ;) [10:30] more! MOTUs! :) [10:30] also [10:30] Hi dholbach [10:30] the way audacious has been packaged in debian has broken the rules several times over [10:30] hey geser [10:30] and i have had to tell le_vert how to fix it [10:30] it's a little scary [10:30] minghua: To a certain degree, I can agree that a finalised spec is of interest to Ubuntu Developers, but I don't see why anyone (e.g. nenolod) who has an idea and wants to contribute can't draft something to encourage support. [10:31] yes, a spec on fixing packaging of audacious-plugins is a good idea [10:31] persia, imbrandon: If I were an upstream developer and was discussing packaging my software, and spec is mentioned, I most likely don't understand exactly what is needed (hell, I'm an MOTU and I don't), and will think "that's quite some bureaucracy, I think I won't bother". [10:31] maybe if people see that the idea is supported, maybe, just maybe, le_vert would fix it in debian [10:31] dholbach: Do you really want to do a Q&A session at 12:00 UTC on the 19th? Isn't that also the MOTU Meeting, followed by the Mentoring meeting? [10:31] (but i doubt it.) [10:31] minghua: Ah. makes sense. [10:32] what i find scary is the people who build debs of audacious snapshots using checkinstall [10:32] :( [10:32] and then they bother us upstream about it [10:32] lol [10:32] persia: I know you are engaged in many discussions on IRC, so I would suggest you say "write things on the wiki" in the future. [10:32] persia: damnit, I meant 14:00 UTC [10:32] nenolod: Just in case minghua is correct, and my suggestion to draft a spec frightens you, I'm just suggesting the creation of a wiki page describing the effort, and what needs to be done. Given the smaller scope (considering source incompatibility), perhaps it doesn't matter. [10:32] persia: thanks, sending out mails [10:33] persia, i have written specs before (although not for ubuntu) [10:33] at first when ubuntu came out, i had doubts that many people would use it too [10:34] but even now I use it [10:34] :) [10:34] nenolod: what exactly is the problem with audacious-plugins packaging in Debian then? [10:34] slangasek, most common usecases for audacious besides MP3 playback are in audacious-plugins-extra [10:35] nenolod: See? I said people would be interested. Better create a wiki page and start writing. :-) [10:35] nenolod: which common use cases are you referring to? [10:35] i'll just write a markdown page on my homedirectory [10:35] give me a sec [10:36] nenolod: Thanks. [10:38] i wonder what popcon says about that [10:39] gutsy rc out ? [10:39] http://releases.ubuntu.com/releases/7.10/ [10:39] proppy: That doesn't look like a candidate to me [10:40] s/rc// [10:40] nenolod: so we'll get a link to this page of yours when you've got it written? [10:41] yes. [10:41] Hey MOTUs [10:41] * persia wishes that the comment about MOTU joining process had been omitted from the recent email [10:42] persia: which mail? which comment? [10:43] dholbach: My mail to ubuntu-motu@l.u.c. There's lots of comments on the joining process, but no feedback on scaling for more Ubuntu-origin. I should have drafted it differently, or skipped the last bit. [10:45] persia: what do you mean by that? getting more packages into ubuntu? [10:45] 09:51:13 (9.94 MB/s) - `ubuntu-7.10-desktop-i386.iso' saved [729608192/729608192] miam [10:45] dholbach: Rather: Are we prepared to handle lots more package in Ubuntu? If we are, how should we organise the effort? Is this a MOTU function? [10:46] in the long run we need lots of teams who handle certain groups of packages [10:47] dholbach: That's one solution. My fear there is that we might well end up with lots between the cracks. It also only encourages more new packages within a team. [10:47] persia: I suspect even if you drafted it differently, that part is still going to be largely ignored. [10:48] proppy: nice connection ;-) [10:48] hey Schnitz [10:48] hi daniel [10:48] minghua: You're probably right. Alas. Unintentional evil and all that. [10:49] flamewars will happen, no matter you ignited it or not. :-) [10:49] (not that I'm saying the getdeb thread is a flamewar, of course) [10:49] minghua: Sometimes. My matches tend to fall on wet ground (and I don't think the getdeb thread is a flamewar) [10:50] me neither, I'm really happy with how João is still trying to find a way to make collaboration happening [10:51] why exactly is ubuntu in the position of trying to host packages outside of debian [10:51] ? [10:51] what does it mean when a bug is assigned to MOTU? that the motu team is working/taking care ? [10:51] I am pessimistic about it. Though I've always been a pessimist. [10:52] I think there is huge scope for collaboration, but I think someone Ubuntu-side (MOTU or not) needs to commit to wider universe and more frequent upstream updates as part of making that work, and I'm not sure the team to do that is well identified. [10:52] pwnguin: Users want to use them on Ubuntu, what's the alternative? [10:52] filing ITPs? [10:52] Schnitz: It usually means someone thought that MOTU should fix it. It doesn't usually mean that someone is working on it. It shows up on the team task list, but that doesn't always drive member tasklists. [10:53] pwnguin: What if no sponsor is available? [10:53] persia: okay i see [10:53] pwnguin: And what if the packager doesn't use Debian? [10:54] pwnguin: There are a number of Ubuntu-origin ITPs, including pointers to Ubuntu packages. There perhaps isn't the focus to get them in Debian-side (perhaps because the packager doesn't use Debian / hasn't tested on Debian) [10:54] im just trying to look at this from the perspective of a debian developer who's been told that ubuntu needs debian [10:54] I don't think it's unrealistic to try getting into Ubuntu first. As long as the ultimate goal is getting it into Debian as well. [10:54] pwnguin: I've seen too many RFS got unanswered for months. I do agree collaboration with Debian is important, of course. [10:55] request for sync? [10:55] sponsor [10:55] yes, request for sponsoring. [10:55] pwnguin: Without Debian, Ubuntu wouldn't be. Ubuntu isn't staffed to handle 15,000 source packages: only about 1500-2000 really get touched. WIth more universe, the number of Ubuntu-maintained packages may grow, but that doesn't necessarily indicate interest in the remaining 12,000 packages not maintained by Ubuntu. [10:56] dholbach: please don't take a look at my other packages before i've sent you an email again, i think i have at least the maintainer mistake in the other ones. [10:57] aww crap [10:57] Schnitz: and the version number [10:57] Schnitz: that's why I only commented on one of them :) [10:57] i forgot to set the package question answering to default [10:57] dholbach: ok great! [10:57] priority=high on 1000 packages is not fun [10:59] http://carpathia.dereferenced.org/~nenolod/audacious/debian-pkg-issues [10:59] slangasek, persia, minghua, etc [10:59] ^ [11:01] dholbach: so i should use Ubuntu MOTU Developers as maintainer, or did i get the wiki article wrong? [11:01] Schnitz: yes [11:01] Schnitz: you can preserve your name in the XSBC-Original-Maintainer field [11:02] since there appears to be some smart audio peoples at the moment [11:02] gstreamer suppport for s3m? [11:02] nenolod: The proposal makes some sense. We'd probably want to align metapackages in such a way that users could satisfy sensible (limited) defaults with main+restricted, do a fair bit with main+restricted+universe, and do everything with main+restricted+universe+multiverse, perhaps meaning a requirement for "audacious-plugins", "audacious-plugins-extra", and "audacious-plugins-nonfree" or something similar. [11:02] persia, all plugins are free [11:02] nenolod: separate binary package for *each* plugin? oh dear God no [11:03] persia, we have even managed to get upstream codec people to relicense their reference code under LGPL so that it can be used freely [11:03] nenolod: Would none of the plugin packages wish to depend on non-free libraries in multiverse? [11:03] slangasek, gstreamer does it [11:03] nenolod: Clear notes. Though I'm not sure about the gstreamer plugin structure statement -- doesn't it use -base, -good, -bad, and -ugly? [11:03] no, gstreamer does not [11:03] xmms does! [11:03] ;p [11:04] nenolod: I thought the point was to slowly make xmms more obviously redundant :) [11:04] persia, no, the point is to clean up the packaging of audacious to make it more useful for the current community which uses it [11:04] * nenolod has no interest in replacing XMMS [11:04] nenolod: Ah. === davro is now known as davromaniak [11:05] nenolod: I think persia means partly "patent-encumbered" when he says "non-free". [11:05] uh, also cdbs? [11:05] pwnguin, common debian build system [11:05] i know it has fans [11:06] but it also has enemies [11:06] I'm glad I live in a country where there are less patents that effects software. [11:06] * persia only means universe [11:06] vs. multiverse [11:06] well [11:06] * persia must change physical locations - will read logs later [11:06] nenolod: Is it possible for audacious upstream to divide the plugins into "base", "commonly used", and "uncommon" groups? [11:06] right now, debian and ubuntu ship libsidplay which contains a copy of C64 kernal [11:07] minghua, we are considering making such a recommendation with 1.4. [11:08] nenolod: IMHO that's a saner packaging approach than "one package per plugin". [11:08] well [11:08] i need to think more about that [11:08] i dont see the massive problem with one package versus one large package [11:09] well, some people want to mutilate my work for the sake of fitting it on a livecd and stuff [11:09] ;) [11:09] pwnguin: I can't parse that. [11:09] err [11:09] one per package [11:09] i dont see the massive problem with one per package versus one large package [11:09] yeah, i agree with pwnguin [11:09] this is how many other distros do it [11:09] it works pretty well [11:09] i dont think you agree with me [11:09] because if theres no difference, you have nothing to complain about [11:10] Maybe CD space, maybe main support, I don't know either, but I can imagine some. [11:10] pwnguin, it's two packages at the moment [11:10] pwnguin: as a user of audacious-plugins, I don't want to have to hunt down plugins one by one; which means I want a metapackage; which means that if this is the common case, there's not much benefit to having the individual packages [11:10] pwnguin, with non-obvious assignments [11:11] slangasek, then a better way would be to just kill off audacious-plugins-extra [11:11] so fix the non-obvious assignments, there's no need to split it into bitty pieces [11:11] nenolod: why? [11:11] slangasek: i dont think anyone was advocating the removal of metapackages [11:11] because upstream is finding all of the people who join #audacious and go [11:11] "LOL HY WHY CANT I USE LAST.FM" to be annoying [11:11] pwnguin: er? [11:11] and we might put a contract out on le_vert [11:11] you never know [11:11] ;p [11:12] just start mailing irc logs to the bts [11:12] pwnguin, great idea [11:12] :D [11:12] slangasek: I think even if it goes the one package per plugin way, there will still be a audacious-plugins meta package. [11:12] yeah [11:12] the reason why i propose splitting them up and using a metapackage is because then people who don't want all of the plugins presently in -extra can simply go [11:13] apt-get install audacious-plugin-scrobbler [11:13] or whatever [11:13] nenolod: if you did that I might be honor-bound to take his place, and I'm somewhat resilient ;) [11:13] how big is -extras in total [11:13] ? [11:13] pwnguin, huge [11:13] pwnguin, it has like 20 plugins in it [11:13] how huge? [11:13] in megabytes [11:13] several MB, but that's not the point [11:13] slangasek: one package per plugin actually sounds not a bad idea to me, if there are no archive maintenance concerns. [11:13] doko and ajmitch : I'd like to see with you about zope / plone packaging.... [11:14] audacious' behaviour can be different depending on what features you have available [11:14] minghua: in which case you would presumably want audacious to continue depending on audacious-plugins as it does today, making the split pointless [11:14] slangasek, no it doesn't [11:14] slangasek, because it's -extras that is the problem [11:14] minghua: there are archive maintenance concerns in Debian as well, but I didn't expect those to be persuasive to most people so didn't bring them up [11:14] slangasek: what about only splitting -plugins-extras? [11:15] nenolod: oh. your post didn't distinguish between -plugins and -extras [11:15] minghua, might as well do it all [11:15] that way people who don't want to listen to WMA for instance can choose to uninstall it [11:15] ;p [11:15] nenolod: Then slangasek's argument stands -- does audacious need to depend on all plugins included in -plugins meta-package? [11:15] minghua, it does not need to depend on any of them [11:16] minghua, of course, it won't play audio without codecs, etc [11:16] but it will start up as long as you have libstdio [11:16] so in the Debian sense of a dependency, it does need to depend on them [11:16] nenolod: How large is -plugins now? [11:16] -plugins is smaller than -plugins-extra, carrying only the plugins on that list. [11:16] since "depends" means "needed to make it serve its purpose", not "needed to make it start and look at you blankly" [11:17] Why would you not want the ability to listen to WMA? [11:17] nenolod: I feel you should revise your notes and concentrate on the problem of -extras. [11:17] 702kB [11:17] minghua, no. there are problems in -plugins too. [11:17] nenolod: Maybe list the plugins one by one, and argue they should be split. [11:17] minghua, audacious-plugins carries an ALAC decoder. [11:17] RAOF: maybe because it loads the plugin into memory if present? [11:18] minghua, the legality of that decoder is questionable. If included, it should be in multiverse. [11:18] minghua, moreover, the decoder has bugs [11:18] pwnguin: Fair call. [11:18] RAOF: and you dont have any need for it? or because you think you can save liveCD space? [11:18] i donno [11:18] im still on the install 'em all side [11:18] * RAOF is too. [11:18] nenolod: What do you think of the idea of keeping a refined -plugins with essential and legal plugins, and split the rest? [11:19] minghua, i think it's still non-obvious [11:19] nenolod, if they see individual plugin names being installed, they will go "hey, i can apt-cache search for this" [11:19] er. [11:19] wtf [11:19] :D [11:20] nenolod: I think listing individual plugin names in package description is a good compromise. [11:20] I don't really like many small packages, either. [11:20] But this is not really my call. [11:20] (I don't even use audacious now, sorry.) [11:21] minghua, yes. i suspect you'd actually have to be an audacious user to understand why splitting is a good idea [11:21] that is more so than "i listen to MP3s on $10 speakers" [11:21] no, I use audacious and think it's nutty :) [11:21] i got tired of xmms being a crappy way to handle playlists [11:21] well, audacious has a playlist manager [11:22] we finished it in 1.4 [11:22] because nobody dared think maybe winamp was stupid [11:22] i kinda like rhythmbox [11:22] pwnguin, there's a lot fo stupid things in winamp :D [11:22] nenolod: You should try to convince slangasek, he has a much larger chance to make things happen than I do. :-) [11:22] but this is a derail [11:22] well [11:22] if it was a big deal to me [11:23] minghua: but I'm notoriously set in my ways... [11:23] i would just split each plugin into a seperate source package [11:23] thus forcing everyone to do as i say [11:23] or drop audacious [11:23] it's 50% chance of suicide [11:23] :P [11:24] but that's an inconvenience for everyone, and me too, so i doubt i would take the LAMIP approach [11:24] nenolod: It's possible to include multiple upstream tarballs in one source package (just to let you know ;-) [11:24] (this is how they handle packaging plugins) [11:24] (seriously. see http://lamip.sourceforge.net ) [11:25] slangasek: Do you think having a core package with essential plugins, and split the rest to separate package a sane idea? [11:25] lamip is another approach in the same spirit of audacious [11:25] it doesn't get enough attention === Fujitsu is now known as Fujitsu_ [11:26] (e.g. XMMS sucks, lets fix it) [11:26] minghua: that doesn't seem totally unreasonable, but I guess it depends how many individual plugins are going to be regarded as "non-essential" [11:27] slangasek: Right. That's what I'm trying to persuade nenolod, but he doesn't seem to be interested in that approach. [11:27] (and since I don't use audacious, I'll shut up on this issue) [11:28] minghua, the reason why is because it will be bureaucracy that settles the problem (like all debian issues) and in the end the users will be screwed and everyone will bitch [11:28] or a different solution: everyone goes "hey. I listen to format X, include it in base" [11:29] so -extras gets like 3 plugins [11:29] or whatever [11:29] Well, users always bitch, so I wouldn't really worry about that. [11:29] e.g.: what is "essential" to one person, is not "essential" to someone else [11:30] so, we would still have the same basic problem we have now, except, [11:30] nenolod: er, who's complaining about this behavior aside from you? "users will be screwed" supposes a level of dissatisfaction with the current setup that I don't see evidence for [11:30] slangasek, we get in #audacious on irc.atheme.org, many people who ask "why can't i scrobble with the debian/ubuntu/mepis packages" [11:31] so if that's the primary complaint, that's easy enough to fix by moving the one plugin [11:31] or "why is format A supported but format B is not" [11:31] also, audacious-plugins-extra is a Recommends:, so in the future will be installed by default === Seveaz is now known as Seveas [11:31] and "why do I have to install a bunch of stuff I don't want to get the one thing I want" [11:31] "cause it's not Gentoo kthxbi" [11:32] slangasek, yes, well, that's one thing Gentoo gets right [11:32] nenolod: I suggest you ask them to complain to Debian/Ubuntu on your channel. [11:32] infact, the only thing Gentoo gets right [11:32] minghua, we do. [11:32] nenolod: In return for being unsupportable-by-design :) [11:32] RAOF, yes. [11:32] RAOF, i love it when i get bug reports, ask for emerge --info, and see 4 lines of CFLAGS [11:32] nenolod: be honest, it's really the Mepis users complaining isn't it. :) [11:32] RAOF, :D [11:33] RAOF, then I tell them to get hit by a bus and close their bugs [11:33] RAOF, somebody keeps telling users to not bother our IRC channel with packaging issues. I want to give them some of our SOC money when we get it. [11:33] :D [11:34] but it's annoying [11:34] because then they ask stuff like "well, why don't you TELL downstream how to package it" [11:34] and they honestly seem to disagree on the "well, because we have lives" theory [11:34] My take is still that "users will always find a way to bitch". If you split packages, then they'll go to your channel and complain they need to hunt down packages one by one. [11:35] actually [11:35] as of late, i just tell them to use some other player if they don't like it [11:35] :P [11:35] e.g. [11:35] "have you tried Amarok lately? it seems tobe all the craze" [11:35] they must love me over at Amarok [11:35] for referring so many idiots to them [11:35] :D [11:36] Sounds a good strategy to me, actually. ;-) [11:37] * minghua wishes he can recommend competitive software to whiny users, too. [11:37] try it [11:37] ubuntu is garbage [11:37] have you tried Gentoo lately? [11:37] I understand you can use it to save money on heating bills. [11:37] :D [11:37] nenolod: Nah. Now I completely ignore them, I still think that's easier. [11:38] minghua, well usually they join in the middle of an argument between me and chainsaw or me and ccr or whatever ;) [11:38] And I don't really have a good competitive software to recommend in the first place. [11:38] who said i recommended amarok because i thought it was good? [11:38] :P [11:38] s/good/viable/ [11:39] i mean i'm sure amarok is good for a jukebox, but it doesn't have a design oriented towards high quality audio reproduction - it just wraps other engines [11:39] (actually, people keep asking for solution to make amarok embed the audacious audio engine) [11:44] nenolod: My suggestion is just that, if you as upstream recommend adding/removing some plugins in the -plugins package, it will meet less resistance. If you recommend complete package structure overhaul and split all plugins, you need a more convincing case. [11:45] i disagree. gentoo gets flamewars right in spectacular fashion [11:45] nenolod: So maybe concentrate on what can be easily done is a good idea, like removing apple loseless plugin, adding last.fm plugin, etc. [11:45] well it's just observations [11:45] nenolod: you could just publish a debian dir in the upstream archive ;) [11:45] pwnguin: Are you disagreeing against me? [11:45] what should be done is better left to key people in MOTU and DDs [11:46] pwnguin, no! that's bad [11:46] says who? [11:46] pwnguin, dondelelcaro has already cussed at me and anarcat about considering that in an indymedia package known as loreley [11:46] says most Debian and Ubuntu developers. [11:46] i like hess' take on the matter [11:47] Having debian/ in upstream VCS is probably not a big deal, but putting it in tarballs causes endless of problems. [11:49] pwnguin: eeew, please don't encourage that [11:49] slangasek, i know better anyway ;) [11:49] nenolod: but he might encourage others too :( [11:49] heh [11:49] im aware people dont like it [11:50] anyway, Joey's position is aesthetically pleasing, but not very practical to date [11:50] but mostly it seems to be because it undermines their authority on packaging [11:50] a maintainer is nearly an owner [11:51] no, the better approach is to send the debian patches you make to the debian maintainer and ask him to consider it [11:51] I don't think so. You can't remove bad files from upstream's debian/. [11:51] if he respects upstream, he will [11:52] if they disagree, you cant do much because they own it [11:52] you can try the ctte [11:52] what's your point? [11:52] im not sure [11:52] ctte? [11:52] the maintainer is responsible for the package, they should have control over the outcome without having to do dances to work around upstream inclusion of debian [11:53] nenolod: I think he means Debian's technical committee [11:53] ive seen the debian technical committee referred to as ctte for some reason [11:53] slangasek, ah. the ultimate in starting debian flameage [11:54] pwnguin: If upstream is not happy with Debian's packaging, it's better to provide their own debian package than including debian/ in upstream tarball. [11:55] which is apparently done by contacting the MOTU ;) [11:56] pwnguin: I'm not sure I understand you. [11:57] pwnguin, no [11:57] pwnguin, it's done by redirecting bitching users to launchpad. [11:57] pwnguin: debian-ctte is the mailing list name; usually referred to as tech-ctte or TC though [11:58] ah [11:58] wel, just TC would be a bit confusing i think [11:58] "debian or ubuntu?" [11:59] anyways, bedtime === huats is now known as huats_ [12:02] http://carpathia.dereferenced.org/~nenolod/audacious/debian-pkg-issues [12:02] updated a little [12:03] slangasek, well, i'm sure you'll be happy to know that atheme.org has standardised on Debian and rPath machines [12:03] slangasek, before we were using gentoo - it was a nightmare to maintain :( [12:04] congratulations on finding your way :) [12:04] well, i knew the way beforehand [12:04] but the idea was that "gentoo would help us roll out our own patches more easily via overlays" [12:04] oh man, what a joke [12:04] it wound up being so painful that nothing ever got updated === tonyy is now known as tonyyarusso [12:15] ScottK: I subscribed motu-uvf to bug 64032 since its for gutsy-proposed. Let me know if there is a need for something to be added. [12:15] Launchpad bug 64032 in rpmstrap "rpmstrap using /bin/sh when it should use /bin/bash" [Medium,New] https://launchpad.net/bugs/64032 [12:16] Hey norsetto [12:16] coNP[uni]: hey asisak, how is it going? [12:17] Thanks. It is cool. [12:17] * coNP[uni] begins to miss to run an unstable version [12:18] Not that Gutsy would have been released ... === norsetto is now known as norsetto_limbo === asac_ is now known as asac === persia changed the topic of #ubuntu-motu to: Ubuntu Masters of the Universe: https://wiki.ubuntu.com/MOTU | Want to get involved with the MOTUs? https://wiki.ubuntu.com/MOTU/Contributing | http://ubuntu.joejaxx.org/ - TOP 10 Uploaders/Packages === RoninX341 is now known as etank [12:38] persia: It's still good to keep the gutsy-updates info on topic, isn't it? [12:39] minghua: You think? If we're released, I don't see how freeze status means anything. [12:39] persia: I believe the gutsy-updates stuff will apply until Hardy opens, so the normal process can be followed. [12:40] So should be in the topic, probably. [12:40] Right. Nevermind. Undoing... [12:40] persia: not the freeze part, the "start working on Gutsy SRUs (gutsy-proposed is open for motu-uvf approved uploads" part. [12:40] Poor persia. And thanks BTW. :-) === persia changed the topic of #ubuntu-motu to: Ubuntu Masters of the Universe: https://wiki.ubuntu.com/MOTU | Gutsy Gibbon released - start working on Gutsy SRUs (gutsy-proposed is open for motu-uvf approved uploads). | Want to get involved with the MOTUs? https://wiki.ubuntu.com/MOTU/Contributing | http://ubuntu.joejaxx.org/ - TOP 10 Uploaders/Packages [12:55] !SRU [12:55] Stable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates [13:10] moin all [13:10] happy release day to y'all! :) [13:11] Hey dholbach. [13:12] hey fujitsu === huats_ is now known as huats === norsetto_limbo is now known as norsetto [13:30] scottK: I assume that being assigned to bug 64032 means I have an ack to upload? [13:30] Launchpad bug 64032 in rpmstrap "rpmstrap using /bin/sh when it should use /bin/bash" [Medium,In progress] https://launchpad.net/bugs/64032 [13:31] norsetto: Yes. [13:31] norsetto: Note my comments about version numbers though. [13:32] scottK: sorry, I must be blind, which comment? [13:33] norsetto: The one I forgot to post. [13:33] Just a sec. [13:33] * norsetto thinks scottk should head to bed very soon .... [13:33] Please use ubuntu0.1 and save ubuntu1 for Hardy. [13:33] norsetto: Have slept and am back. [13:34] am also drinking coffee. [13:34] scottk: you kidding? you call 3 hours a sleep??? [13:34] Today, yes, but not as a rule. [13:35] scottk: gee man, you should take care of you better, gotta talk with your wife one of these days [13:35] I could go back to bed and then be unable to get to sleep tonight again. I just need to suck it up for one day to get back on a ~normal schedule. [13:35] norsetto: I'm making progress. Back when I was in the Navy, this was pretty normal. Now it's at least rare. === amachu_ is now known as amachu [13:37] scottK: thx for the number btw, I totally forgot about it [13:38] norsetto: No trouble. That's one of the main reasons I proposed motu-uvf ack all the uploads right now. [13:40] To all of you: WELL DONE! [13:42] this yada thingie is snappy .... [13:43] Er, I hope that isn't the yada I'm thinking of. [13:43] norsetto: I was about to warn you, but to late I guess. [13:43] Fujitsu: I'm afraid it is ..... [13:44] scottK: its ok, I just had to get used to it [13:44] * Fujitsu sobs. [13:44] * norsetto pats Fujitsu on the shoulders [13:44] * ScottK sits back and waits for fireworks. [13:45] scottK: you mean its going to make the launchpad buildd very happy? [13:45] No, I mean some people really don't like yada. [13:46] Does anyone know if we can file hardy sync requests now (I know they can't be acted on)? [13:47] ScottK: all I know is that I marked a couple of bugs as sync yesterday already [13:47] OK. [13:47] hi [13:48] howdy proppy [13:48] nice ! [13:48] back to work place [13:48] much nicer to work here than from home [13:49] proppy: at least you get paid ;-) [13:49] nop :) [13:49] proppy: so, you managed to get to work eh? No lame excuse due to the strike ... like a real man [13:49] we got fired a long time ago [13:50] we only use the place for geeking together :) [13:50] but we're not paid for that [13:50] ! [13:51] but we still have free money each month for one year :) [13:52] then we will have to figure out how to earn some [13:52] I forget to ask what is your daily work norsetto? apart for being a ubuntu dev ? [13:54] proppy: daily work? whats that!? [13:54] nice one :) [13:55] norsetto: are you able to work on the viewvc thing with me ? [13:55] (now) [13:55] proppy: is there still something to be done? [13:56] viewvc-config patch [13:57] proppy: just a sec. I need to finish something first [13:58] I thought I can do it myself, it just feel easier if you're close [13:58] :) [14:02] proppy: as if I can help you ..... you are the python wizard [14:04] cheering me up is a great help :) === TomaszD_ is now known as TomaszD [14:12] norsetto: just have written a testcase, do you agree on the behaviour http://lp152438.aminche.com/comment/comment.py.txt ? [14:12] * norsetto reads [14:12] proppy: what makes you think that I know Python? [14:13] norsetto: my guess is that testing code is pretty langage independant [14:16] proppy: I think its \n not \n\ in self.assertEquals [14:17] yep just figured that out [14:17] :) [14:18] proppy: what happens if content = "toto\ntiti tutu\ntata\n" ? [14:20] hi [14:21] norsetto: let me add the test [14:22] norsetto: what is the expected behaviour ? toto\n#titi tutu\ntata\n" ? [14:22] it passed [14:22] check [14:22] http://lp152438.aminche.com/comment/comment.py.txt [14:22] proppy: ok [14:23] what is the expected behaviour commentLine("toto\ntiti tutu\ntata\n", "tutu") ? [14:24] comment or not comment ? [14:24] my guess is that we only need to be able to comment matching the begin of the line don't we ? [14:28] updated [14:32] do you think of any other test cases before I move this to viewvc-config ? [14:34] (hint refresh the page) [14:44] proppy: (2) if you don't address to me I might not see it ...... [14:45] norsetto: sorry [14:46] norsetto: hey [14:46] norsetto: happy release day [14:46] this place starts to fill up with french ..... [14:46] huats: hrd to you too .... do you know proppy? [14:47] norsetto: let's say that the place was already filled with french, but the french knows to be silent... sometimes [14:47] :-) [14:47] proppy: nice to know you :-) [14:47] :-O [14:47] proppy->huats; huats->proppy [14:48] proppy: be kind with this poor sould from the deep south ..... [14:48] hi huats [14:48] where do you from ? [14:48] toulouse [14:48] proppy: norsetto says you know about Python? [14:48] avec l'accent ? [14:48] exactly [14:48] ScottK: I do [14:48] Up for a project? [14:48] proppy: where do you from ? [14:49] ScottK: always [14:49] huats: former "centre" now "paris" [14:49] scottK; he is also soon out of a job ... didn't you emntion an rfp some days ago? [14:49] I did. [14:49] Let me see if I can find the announcement (it's Python/Django). [14:50] ScottK: just wait I've finished treating this viewvc bug, I try to open only one task by context [14:50] proppy: http://www1.ietf.org/mail-archive/web/ietf-announce/current/msg04179.html is it's your kind of thing [14:52] proppy: I think it would be interesting to have a script to compare what versions we have with getdeb. The requestsync script in ubuntu-dev-tools (in Python) already knows how to find Ubuntu versions. The getdeb data structure was mentioned on either motu ML or devel-discuss yesterday or the day before. [14:53] ScottK: hope it's not a solo job (about the RFP) [14:53] proppy: They are looking for one contractor. Up to you how many people you get together to do it. [14:54] Maybe you and ajmitch can collaborate (dunno). [14:54] ScottK: I'm not a good contractor when I'm alone and remote [14:55] ScottK: getdeb thing is the project you were talking about ? [14:55] OK. Just mentioned it here because I thought there might be people who might be able to go after it. [14:55] proppy: Yes. [14:55] proppy: one more test case: commented = commentLine("toto\ntiti = tutu\ntata\n", "titi =") -> self.assertEquals("toto\n#titi = tutu\ntata\n", commented) [14:56] norsetto: Ran 1 test in 0.004s [14:56] it passed [14:57] proppy: ok, most probably we are missing some 100th of cases that will fail, but never mind ..... [14:57] ScottK: I don't know anything about getdeb (yet), is it related to dget ? [14:57] norsetto: it's better than nothing :) [14:58] on which ml was it discussed ? [14:58] -devel ? [14:59] proppy: devel-discuss [14:59] proppy: Ran 1 test in 0.001s -> my computer is bigger than yours, so there [14:59] :) [15:00] norsetto: I'm running the test remote :) [15:00] proppy: no lame excuse will hold [15:00] on a 1Ghz box, so it's not hard [15:00] norsetto: but I bet you can't beat my bandwitch haha [15:01] 09:51:13 (9.94 MB/s) - `ubuntu-7.10-desktop-i386.iso' saved [729608192/729608192] [15:01] don't forget kids the faster your download, the bigger your penis is :) [15:01] <_ruben> i could beat that if i'd knew a mirror that'd be bast enough :) [15:01] -r [15:01] <_ruben> s/bast/fast [15:02] ScottK: registred to the list [15:02] ScottK: looking archive [15:03] ScottK: this one https://lists.ubuntu.com/archives/ubuntu-devel-discuss/2007-October/001974.html ? [15:04] proppy: Yes. [15:04] It was also discussed some on the motu ML (IIRC). Somewhere in there they described how to get their packages/versions. [15:04] ubuntu-7.10-desktop-i386.iso [15:04] 458358784 62% 109.36MB/s 0:00:02 [15:05] proppy: it should fail if the word is not the first on the row (ie. " tete") [15:06] ScottK: they supply the binary deb ? how do they figure dependency [15:06] proppy: Don't get me started.: === x-spec-t is now known as Spec [15:06] norsetto: you mean that test ? [15:06] commented = commentLine("toto\ntiti tutu\ntata\n", "tutu") self.assertEquals("toto\ntiti tutu\ntata\n", commented) [15:06] * norsetto calms scottk down, shows him a nice cool glass of beer [15:07] norsetto: It's a little early for that here. [15:07] don't tell me your bandwith stuff i'm updating using update-manager and the german mirrors are under heavy load....50kb/s ;-( [15:07] proppy: whatever [15:07] Future Plans --------------- Move to APT based software distribution [15:07] scottk: its not here, but my fridge look as if the US navy passed over it [15:07] Heh. [15:09] ScottK: so you want a script that compare getdeb and ubuntu version ? [15:09] s/ubuntu/universe/ [15:09] ? [15:10] proppy: Yes. [15:10] ScottK: So we can deprecate GetDeb ASAP? [15:10] Ideally, what I'd like to be able to do is get a list of stuff that they are shipping, that we have in our development release (e.g. hardy) that we could backport to the current release. [15:11] Fujitsu: You catch on quick. [15:11] Their web front end is shiny and potentially useful. I just want to get them out of the packaging business. [15:11] Yes. [15:12] clever :) [15:12] * proppy hugs ScottK [15:12] It would be nice to have a repo open so we can backport new UVs post-UVF. [15:12] And then where they have newer stuff than in our development release, that'd be a work list for here. [15:12] Fujitsu: ubuntu-experimental. [15:13] Grumpy! [15:13] Someone (not me anytime soon) should also look at the new Automatix release and see what we can do for Hardy to put a few more nails in their coffin. [15:14] scottK: I was actually surprised they don't even distribute their source packages [15:15] norsetto: They do now, AFAIK. [15:15] norsetto: You must be easily suprised. [15:15] Fujitsu: I checked yesterday and couldn't find it [15:15] norsetto: Oh, for Automatix? [15:15] Fujitsu: yes [15:15] Ah, thought you were talking about GetDeb. [15:16] heh grumpy is an urban legend ;) [15:16] ScottK: does the patch in bug #129050 qualify for gutsy-proposed? [15:16] Launchpad bug 129050 in gtk2hs "ghc library packages all have bad dependencies" [Medium,Triaged] https://launchpad.net/bugs/129050 [15:17] geser: I'd say so. Let me mark in the bug. Make sure you use a version number that won't conflict with the eventual Hardy upload. [15:19] geser: Done. [15:20] ScottK: we can work on it, after the viewvc bug is fixed if you want :) [15:21] proppy: Sure. No rush. === Fujitsu_ is now known as Fujitsu [15:25] norsetto: For Automatix, there's nothing compiled, so you can disassemble their .deb to get their source. [15:25] scottk: with dpkg-deb you mean? [15:25] Personally I use ark, but whatever you like. [15:26] (04:05:57 PM) norsetto: proppy: it should fail if the word is not the first on the row (ie. " tete") [15:27] norsetto: do you confirm it's already adressed with the testing code ? or do I get you wrong ? [15:27] proppy: yes, I tested it out of curiosity and it indeed failed [15:28] by failed you mean the test failed -> it did comment the line , or the test success -> it didn't comment the line [15:29] proppy: neither : the test fails -> it didn't comment the line [15:30] show me the test [15:30] proppy: commented = commentLine("toto\n titi = tutu\ntata\n", "titi =") -> self.assertEquals("toto\n# titi = tutu\ntata\n", commented) [15:31] riiiight [15:32] proppy: is this a regex '^'+var ? [15:34] yep [15:34] \s* [15:34] do the trick [15:34] but you should change the test to commented = commentLine("toto\n titi = tutu\ntata\n", "titi =") -> self.assertEquals("toto\n#titi = tutu\ntata\n", commented) [15:34] If you don't want to care about grouping syntax [15:35] norsetto: http://lp152438.aminche.com/comment/comment.py.txt [15:35] updated [15:35] proppy: ok [15:36] it trash the space between # and keyword [15:37] proppy: yes, np [15:40] let's patch viewvc-config ? [15:41] -t for comment ? [15:42] norsetto: does this look a cool test to you http://lp152438.aminche.com/comment/test-viewvc-config.sh.txt [15:42] ? [15:44] proppy: why not with a clone of viewvc-config ? [15:45] proppy: sorry, viewvc.conf.dist ? [15:45] norsetto: I wanted to try this one http://lp152438.aminche.com/comment/test.conf.txt [15:45] norsetto: I just figure out of a new testcase that failed :) [15:47] commented = commentLine("toto\ntititi = tutu\ntata\n", "titi") -> self.assertEquals("toto\ntititi = tutu\ntata\n", commented) [15:47] proppy: btw, -k (--key) is there already [15:47] yep [15:47] that's why I used it [15:47] to know which key to comment [15:49] norsetto: does it looks reseanable to comment only line with = in it ? [15:50] proppy: yes, but, what if the user did define a template dit different from default? Shouldn't we check for it too? [15:50] proppy: s/dit/dir [15:51] proppy: but I guess we can do that in viewvc.config [15:51] yep [15:54] ScottK: do I get a confirmation that rpmstrap is in gutsy-update, so that I can send the email, or should I send the email in any case? [15:55] norsetto: It'll get marked in the bug when it's published, so as long as you are getting the bugmail, you'll get notified. [15:55] scottk: excellent [15:55] norsetto: I've updated the test and the implementation [15:58] norsetto: updated http://lp152438.aminche.com/comment/test-viewvc-config.sh.txt === cprov is now known as cprov-lunch [16:01] proppy: isn't a cat << EOF | tee test.conf needed? [16:05] norsetto: cat > test.conf < proppy: that should do too [16:10] norsetto: -t option implemented in viewvc-config [16:10] norsetto: http://hg.lp152438.aminche.com/rev/cecc6c3f0e4b [16:10] bash -x test-viewvc-config.sh || echo FAILED && echo SUCCEED print SUCCEED [16:12] proppy: Have you checked what happens if -a is specified? [16:13] if a is specified getting is one [16:13] so it should sys.exist first [16:13] but let's make a test for this [16:14] norsetto: test updated http://lp152438.aminche.com/comment/test-viewvc-config.sh.txt [16:16] proppy: for the second test case this is not true: grep "^#commentme" test.conf [16:16] doutche ? [16:16] why? [16:17] proppy: no, its ok, I didn't consider that you save to test.conf again [16:17] I added a test case for -k [16:17] to check that my code didn't alter the existing behaviour [16:18] (I should have put that test in place prior the modification) [16:18] proppy: don't forget to patch this too: # Usage: viewvc-config -k keyname -v value -c config_file ;-) [16:19] https://edge.launchpad.net/~ubuntu-dev/+polls [16:20] polls open in 8 hours for new MOTU Council candidates [16:20] please vote! [16:21] Hmm. Any MOTU can vote, right? [16:22] coNP[uni]: yes. [16:23] Hey pochu, happy release day! :) [16:23] pochu: are you a MOTU, btw? [16:24] coNP[uni]: nope :( [16:24] *yet* [16:24] norsetto: like this http://hg.lp152438.aminche.com/rev/563165dee4ac ? [16:25] Just say if you need some cheering :) [16:27] proppy: yep [16:28] what is there to vote? two candidates for two posts!? [16:29] ahah [16:30] coNP[uni]: thanks, I'll let you know :) Maybe in the Hardy cycle... [16:37] norsetto: patch updated http://lp152438.aminche.com/comment/viewvc-config.patch.txt [16:38] Heya gang [16:39] hola bddebian [16:39] Heya norsetto [16:41] what happens to stuff in ubuntu-proposed? Does it eventually get put in the main repos? [16:42] MarcC: it is eventually put to -updates [16:42] Whether they belong to main / restricted / universe / multiverse is another thing. [16:43] cool, so that means, if a change to Kopete to make it not crash is pushed to proposed, it will eventually make it into updates? [16:43] ajmitch: i got a note you expired from ubuntu-dev, was that planned or can i change it for you? [16:44] norsetto: so next ? [16:44] norsetto: update viewvc.config to call viewcv-config to comment template line and add template_dir option ? [16:44] what does SRU stand for? [16:45] Stable Release Updates [16:45] ? SRU [16:45] !SRU [16:45] Stable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates [16:45] thanks :) [17:12] ScottK: Hi Scott, long time has passed, how are you? [17:12] Hi there. [17:12] Not bad. Welcome back. [17:12] I was wondering if we'd see you again. [17:12] norsetto_: eolo999 knows about Python too and is also Italian. [17:13] I'm so rusted i don't know if i can help in any way... [17:13] I've posted metalinks for Gutsy gibbon at http://metamirrors.nl/node/139 Any comments are welcome. [17:13] * proppy hugs eolo999 [17:13] proppy: hi [17:13] norsetto_: ciao [17:13] norsetto_ may have some ideas for eolo999 to work on. [17:14] i will need some time to get in again (this time reading docs ;P) [17:14] No rush, glad to have you back. [17:15] scottK; eh, nobody is perfect ;-) [17:15] eolo999: ciao eolo [17:15] ScottK: what about you being a MOTU now? [17:16] eolo999: hey, he is THE motu :-) [17:16] norsetto_: ciao [17:16] when i've known him he was not yet... [17:17] norsetto_: di dove sei? [17:18] eolo999: he, good question: where I was born, where I lived most of my life, where I live now ? [17:18] all three. [17:18] eolo999: This is an English only channel. #ubuntu-it for that if it's in Italian. [17:19] ScottK: ok [17:19] Sorry, but it's the common denominator. [17:19] No problem. [17:19] don't worry [17:19] eolo999: ok ok ....was born in Macerata (if you don't know where that is you will be banned for life), lived most of my life in The Netherlands, and I'm now in Rome, and you? [17:20] i've been in macerata...nice [17:20] i'm from florence but i live in Rome, too. [17:20] eolo999: ok, so I know you did your army service in the Air Force ...... [17:21] i diserted military service, i've been there for my job (i'm a light technician for theatre). [17:22] norsetto_: are you a MOTU? [17:22] apt-cache search python dmx, No match too sad [17:23] eolo999: sure, norsetto_ is a MOTU-youngster. [17:23] proppy: would be nice pydmx! [17:23] proppy: You can package a new python module in just an hour or two. [17:23] eolo999: so they say, I still have to believe it myself though [17:23] ScottK: I don't know if there is a codebase for that :) [17:24] i [17:24] Actually we still need some time to notice the overall regression because of norsetto_ [17:24] ScottK: Thanks to distutils and cdbs :) [17:24] coNP[uni]: they even tried to push the release date because of me ..... [17:24] ScottK: I should upload this one, one day .. http://svn.gna.org/svn/pokersource/trunk/qpoker/debian/ [17:25] proppy: And don't forget Debian Python Modules/Apps Teams. [17:25] What is the bug number of the bug "norsetto is a MOTU"? [17:25] It is release-critical. [17:25] You can get your Python stuff uploaded to Debian pretty easily that way. [17:25] coNP[uni]: oh, that was fix-released few weeks ago .... [17:26] I use to know a DD who can sponsor upload, but since I'm a bit slow to update my package when nobody ask me too, I feel kinda unconfortable to ask him to upload other stuff [17:26] ScottK: what is Debian Python Modules/Apps Teams. ? === norsetto_ is now known as norsetto [17:27] proppy: See https://wiki.ubuntu.com/ContributingToDebian/PythonModulesTeam (it needs to be updated for the addition of the apps team). [17:28] ScottK: bye, i'll be here soon... [17:28] norsetto: If you want to visit me i work every night in Teatro Belli in Trastevere... [17:30] eolo999: thanks, I'll keep that in mind. I'm not really the night owl, but I might convince my better half one day (we live in Borghesiana ... so you know its not really downtown) [17:30] bye all [17:31] eolo999: ciao eolo ... see u again (I hope....) [17:36] seems like we lost proppy somewhere ..... [17:37] nop nop [17:37] proppy is still here AFAICT. [17:37] still here [17:37] do I miss something ? [17:38] (05:44:10 PM) proppy: norsetto: so next ? [17:38] (05:44:42 PM) proppy: norsetto: update viewvc.config to call viewcv-config to comment template line and add template_dir option ? [17:38] I'm currently looking at viewvc.config [17:39] trying to figure out where to add the brand new viewvc-config -t calls [17:39] proppy: ok, I thought you had seen that I was disconnected [17:40] oh ok [17:40] sorry [17:41] following some #gnash discussions in the meantime, and the fact that I barely understand what they are saying, lowered my attention #here :) [17:41] proppy: why not at the very end? Like a proper clean-up ..... [17:42] at the end of the file ? or at the end of the migration part ? [17:43] proppy: at the end of viewvc.config [17:43] sorry forget again to add norsetto: [17:43] :) [17:44] ....exec alias \s norsetto ..... [17:48] norsetto: let me put it in a separate script before [17:53] proppy: actually, the best place is not in viewvc.config [17:53] proppy: I think we should call it from the postinst script [17:54] ok :) [17:55] let's write it first [17:57] proppy: and I think the best place would be just after the: mv /etc/viewcvs/viewcvs.conf /etc/viewvc/viewvc.conf; what do you think? [17:58] yep I agree [17:59] norsetto, mentors meeting is tomorrow at 13 UTC so 15 for us. I'll arrive a bit late, I hope it wont finish in 20 minutes :) [17:59] bluekuja: 20 min? We'll be lucky if its over before dinner .... [17:59] norsetto: working on the script http://lp152438.aminche.com/comment/test-viewvc.config.sh.txt [18:00] bluekuja: btw, I subscribed motu-torrent to a needs-packaging about apt-torrent, you might be interested [18:00] great work guys. gutsy is out :D [18:01] is it? nobody ever tells me anything .... [18:01] when that happend?! [18:01] beside coNP[uni] of course, he tells me way too much ..... [18:01] omg, again, i overslept release... [18:01] uh some hours ago [18:01] ^^ [18:02] Bye norsetto et al. [18:02] See you tomorrow [18:02] Happy Release Day to all :) [18:02] take care coNP[uni] [18:02] jono: hoi. ubuntu --> great xD [18:02] norsetto: don't forget the talkative french... We surround you [18:03] * norsetto starts sweating [18:03] I know the question is stupid and old but "What about a extremy unstable, unsupported bleeding edge repo for gutsy?" <-- I really don't want too use debian or anything other than ubuntu ^^ [18:03] hellboy195: :) [18:03] norsetto: remember how you called me once... Master of little things ? [18:03] well, they are everywhere :-) [18:03] norsetto, perfect, I'll try to be there as soon as I can. Let me check my mail now then (just went home) :) [18:04] huats: little crawling things, makes a differnce [18:04] hellboy195: Wait until the hardy repos open and then ask for gutsy-backports. [18:04] !backports | hellboy195 [18:04] hellboy195: If new updated Ubuntu packages are built for an application, then they go into Ubuntu Backports. See https://help.ubuntu.com/community/UbuntuBackports - See also !packaging [18:05] I know the backports but ok you do a great work but I like "bleeding edge like sid" more. just only with ubuntu ^^ [18:05] norsetto: warning [18:06] * norsetto start swearing [18:06] norsetto: viewvc-config seems not to works for inserting new value, just updating existing one [18:06] norsetto: the SetValue regexp doesn't match anything if the var is not in the file contents [18:07] proppy and norsetto: It's sounding broken enough that you might want to consider contacting the upstream and work with them on it. [18:07] proppy: yes, but they are? [18:07] let me add a test to this old test-view-vc-config.sh [18:07] ScottK: dd seems to be the author of that [18:07] dd? [18:07] norsetto: I'm talking about adding template_dir [18:08] ScottK: debiandevelopers [18:08] ScottK: since the script is in debian/ [18:08] Ah. [18:08] proppy: but you don't need viewvc-configure for that! [18:08] oh ok :) [18:08] norsetto: how so ? [18:09] norsetto: I still need to insert this line in the appropriate section [18:09] whats wrong with an echo >> ? [18:09] norsetto: template_dir should be in [options] section [18:09] norsetto: a few line of sed could do it [18:09] norsetto: as well as commenting the template options line [18:09] ScottK: yes, thats the idea, we open a new bug in the debian bts with the patch and ask his advice [18:10] OK. [18:10] norsetto: I though you wanted to use viewvc-config for that ? [18:10] proppy: we can use whatever its easier to use [18:12] norsetto: we should figure where is the [options] section, and insert a line at the end of it [18:12] proppy: if you feel better using sed sure, go for it, whatever you feel will do the job with the minimum effort [18:12] norsetto: should be easy with a few regexp [18:12] norsetto: I remembered sed addressing is fun :) === bluekuja_ is now known as bluekuja [18:13] proppy: are you sure you are not a woman? [18:13] norsetto: but I fill kinda guilty not to add functionnality to the existing viewvc-config [18:13] norsetto: let me check [18:13] * norsetto hopes Hobbsee is not around ....... [18:14] norsetto: adding NewValue to viewvc-config sound fine to me [18:14] * norsetto wears his best armour suit, just in case [18:14] Hobbsee is a sed addict or a woman ? or both ? [18:15] proppy: don't know if she is a sed addict, but sure has a pointy stick and she is not afraid of using it [18:16] sed is kinda pointy [18:16] norsetto: updated http://lp152438.aminche.com/comment/test-viewvc-config.sh.txt [18:17] norsetto: now failing with python viewvc-config -k "newoption" -v "value" -c test.conf ; grep "^newoption = value" test.conf [18:18] proppy: thwre is no newoption in there? Can you add a key with it!? [18:19] not yet [18:19] what I'm looking for is made that work by patching viewvc-config ? [18:19] oh $deity, I unleashed a monster ..... [18:20] ? [18:22] maybe I should use a different option of that ? [18:22] -d like add ?፡) [18:22] jono: how are you doing with packaging now after the first try with pitivi? cause I'm also starting and I think it a little bit hard but makes me satisfied :) [18:23] hellboy195: I havent had time to do any more, but it was fun [18:24] hellboy195: if you are new, be sure to share your experiences so we know how to make it easier :) [18:24] jono: oh well your feedback was good. also for me and my wishes ;) [18:25] hellboy195: And write documentation. Once you are experienced it's impossible to write docs that work for new people. We NEED people who are new to it to write them. [18:25] ScottK: wise words :) [18:25] ^^ [18:26] ScottK: sure but I think I have to try 2-3 programms packaging first only to know the basics [18:26] hellboy195: Start writing now though. You may not publish until you've done a few, but start writing now. [18:27] ScottK: OK :) but can you tell me in what direction this sould go? [18:27] *should [18:27] hellboy195: What you want to end up with is the document that you wish had existed when you started. [18:28] It's not to early to be writing sync requests for Hardy (requestsync doesn't quite work yet) for packages that don't need to have an Ubuntu diff anymore. I got mine in (9), so you all can get in line now. [18:30] ScottK: K, will do it :) [18:30] Great. [18:32] ScottK: may I ask you a "flamewar" question? why isn't debian not able to package gnome correctly like in ubuntu? [18:33] hellboy195: I use Kubuntu, so my answer is I don't know and I don't care. [18:33] ScottK: well ok [18:34] I suspect the answer has to do with different objectives for the distros, but really don't know. [18:34] ScottK: thx anyway :) [18:35] has anyone tried to install gusty inside qemu? [18:35] I can't seem to start the installer === Spec is now known as x-spec-t [18:48] norsetto: got it working [18:50] norsetto: http://hg.lp152438.aminche.com/rev/11e1f5b7d428 [18:53] norsetto: it's easier to add the option at the beginning of the section [18:53] proppy: you are still at work? [18:53] proppy: I gotta go, wifey is calling for dinner === norsetto is now known as norsetto_limbo [18:54] norsetto_limbo: ok me tto :) [18:54] norsetto_limbo: see you later then [19:04] norsetto_limbo: all the regexp code is useless look at this :) http://docs.python.org/lib/RawConfigParser-objects.html [19:04] norsetto_limbo: see you soon have to go [19:04] see you :) [19:04] * proppy hugs ScottK [19:05] jono: wb [19:10] hi motu, any body knows if backporting team needs contribution? I've started packaging recently and have backported some packages for our local repo like Pidgin, Stardict & ... [19:11] !backports | mehdi2 [19:11] mehdi2: If new updated Ubuntu packages are built for an application, then they go into Ubuntu Backports. See https://help.ubuntu.com/community/UbuntuBackports - See also !packaging [19:11] mehdi2: For backports we only take packages from the Ubuntu development repository (with some modification if needed). [19:11] mehdi2: What we really need is people to test packages that have been requested already. [19:12] ScottK: I've backported some from Debian [19:13] Here we get them into a development repository first and then backport them. That step is not optional. [19:13] ScottK: maybe I can start/help wih testing too. [19:14] great. That's what we really need. If you have questions, jdong is the best person to ask. [19:14] mehdi2: What Ubuntu version(s) do you use? [19:15] ScottK: I just joined the mailing list to be informed about what the team is doing [19:15] ScottK: feisty/gutsy [19:15] OK. [19:16] mehdi2: https://bugs.launchpad.net/feisty-backports/+bugs - Anything marked New or Incomplete probably needs testing. [19:29] cya :) === norsetto_limbo is now known as norsetto [20:09] I just blew my streak. I clicked on a bug in #ubuntu-bugs that I couldn't invalid. I had to confirm it. [20:12] that's not good [20:14] * ScottK looks around. [20:15] * ajmitch looks at the clock [20:16] Ah. [20:24] hehe [20:43] according to https://wiki.ubuntu.com/HardyReleaseSchedule what's the next step for MOTU ?? wait for the toolchain ? [20:47] yes, or just get stuff ready for upload anyway === neversfelde_ is now known as neversfelde [21:05] leonel: We'll take Gutsy SRUs now and sync requests where needed (Ubuntu diff is safe to drop) can be filed. === _czessi is now known as Czessi [21:10] If some mentor is looking for a merge for a hopeful to work on, gramps should be reasonably straightforward and I have no interest in continuing to be the person that touched it last. [21:18] I hope postgresql 8.3 can make in hardy [21:19] Do you know when they plan to release it? [21:20] before chrismas [21:21] Well, feature freeze is Valentine's Day iirc [21:24] :D [21:24] http://www.wired.com/software/softwarereviews/news/2007/10/ubuntu_gutsy [21:25] keescook: From a security perspective, what do you think of a package making it's own user and adding that user to sudoers? Is that OK? [21:25] ScottK: making it's own user, sure. that's common. Adding the user to sudoers: yuck. There had better be a really powerfully good reason. :) [21:26] Since they give that user only limit sudo rights, I'm thinkin it's superior to running as root. [21:27] surely suid would be better? [21:27] keescook: The package is secvpn. Since it's a vpn package, I imagine it really does need to run as root. [21:27] jmg: I'd imagine so. [21:28] It's init scripts are hopelessly ancient and the more I dig, the less I like. [21:28] ScottK: while I'm sure it needs rootly powers to do some things, I'd examine why and where. [21:28] * ScottK feels suddenly in way over his head. [21:29] It does sound like a good audit ought to be done before it's put in working order. [21:29] * ScottK feels less bad it doesn't work so well ATM. [21:31] hey [21:31] secvpn looked like an evil package [21:31] at least its better than the iphone running everything as root :) [21:35] ajmitch: I'm tempted to write it off as a good thing it crashes a lot. [21:35] file a removal request [21:35] we'll all be better off, i'm sure [21:35] ajmitch: I will [21:36] Dear pitti, please remove secvpn source and binary. Reason: because it is evil. [21:37] iphone ?? idon't === ivoks_ is now known as ivoks [21:50] http://tinyurl.com/2gljnw <-- Ubuntu Article - includes MOTU and some of us [21:50] * nixternal pokes ajmitch [21:50] :P [21:51] * ajmitch hides [21:51] hehe [21:51] thought you said no to him? funny I seen your name plastered in it :p [21:51] Mr. Mitchell said, "Ubuntu what? forget that! Debian FTW!" [21:52] as he types from his sid desktop... [21:52] hehe [21:52] which is what I'm using right now :) [21:52] ya, I need to fix my Debian box....it got hosed with some KDE 4'age [22:01] jono: still awake? I'm surprised [22:01] * ajmitch watches the upgrade bugs roll in === lamego_ is now known as Lamego === oracle2025_ is now known as oracle2025 [22:10] how do I set a file to be ignore from the the diff generation during debuild ? [22:11] Delete it? :) [22:11] not of it was changed, and not just created, that would break rebuilding [22:12] Move it and put it back when you are done? [22:12] -i[regexp] [22:12] Edit the .diff.gz afterwards. [22:13] what a lot of work, when dpkg-source has options you can use [22:13] that is not an option, I hate manual actions [22:13] * ajmitch usually builds with -i.bzr, to exclude all the bzr info that would make the .diff.gz rather large [22:14] hum, let me try -i with debuild [22:15] or may I can override debsourceblah blah options on the debian/rules with some CDBS parameter [22:16] not that I'm aware of [22:18] DEB_DH_INSTALLCHANGELOGS_ARGS [22:18] there is :D [22:18] no, that's a debhelper command run in debian/rules [22:18] quite separate from dpkg-source [22:19] why do you need to have this file ignored? [22:19] well, that rule is the one called by CDBS for the package building, I am looking into the CDBS make files [22:19] because this file is changed during the building process, and I do not want that change on the building diff [22:20] ah, wait, changelogs [22:20] forget it, wrong reading [22:21] i forgot how slow the livecd can be :D === apachelogger_ is now known as apachelogger [22:43] hi, anyone in the mood to have a look at the first debdiff from a packaging newby? I have created this (http://pastebin.com/m775701b) debdiff which should fix #69455 - this is a regression of gutsy - a similar patch (http://launchpadlibrarian.net/5214902/oprofile_0.9.2-1ubuntu0.2%7Eproposed1.debdiff) has already been made by someone else for feisty. Did I make a mistake so that the debdiff contains absolute paths starting with /tmp/NQ2LJhqUFj [22:45] kai^sds: That's probably because you did it in a chroot. That's no trouble, whoever has to apply the patch should be able to apply it with -p2, so no worries. [22:46] kai^sds: bash is essentail, you don't need to specify it at depends [22:47] ok. I just did that because the feisty patch did [22:47] geser and soren: Congratulations on your MOTU council nominations. [22:48] hmm, the install from livecd has improved drastically in gutsy [22:49] how so nenolod ? [22:49] seems much faster [22:51] ScottK: \o/ [22:51] ScottK: Thanks, man. [22:51] (but maybe it took as long as last time I did an install from livecd, and it just didn't seem as long because i was listening to mp3s and cussing out an IRC troll) === Vorian_ is now known as vorian === chuck_ is now known as zul [23:30] heh, so my oprofile patch wasn't carried into gutsy? harumph. [23:36] kai^sds: bringing in either the dapper-updates or edgy-updates change will suffice [23:38] crimsun: sorry, I don't know what you mean. the important change has been made for feisty... [23:39] kai^sds: you should install patchutils. It creates a -p1 patch in most cases then. [23:39] kai^sds: I made the original change for the dapper and edgy source packages. It's simply, as you've noticed, a substitution of /bin/bash for /bin/sh. [23:39] except if its a native/debian package IIRC. [23:40] it is a native debian package afaik [23:40] oprofile? No, it's non-native. [23:41] are you sure? i am new to this, but is it possible that that has changed in gutsy? [23:41] yes, I'm sure. It carries a Debian package version in addition to an Ubuntu version suffix. [23:43] my information comes from https://help.ubuntu.com/6.10/ubuntu/packagingguide/C/basic-mistakes.html#basic-ex_files which says: [23:43] A native package is one that is specific to Ubuntu/Debian. It has debian/ directory containing the packaging information and any changes to the source included in the tarball (usually _.tar.gz). Non-native packages are more common. A non-native package splits the source package into a _.orig.tar.gz tarball that is identical ... [23:43] and oprofile does not contain a .orig tarball [23:44] kai^sds: it was mistakenly uploaded as a native one. [23:44] yes, even core-dev make mistakes from time to time :-) [23:44] ah, ok [23:45] unpossible [23:45] hehe [23:46] how are you, crimsun? [23:46] not bad, ajmitch, yourself? [23:46] managing [23:47] been busy with work? [23:48] yep, getting a breather for a bit. Gonna hit up a loco release party/installfest tomorrow afternoon [23:48] excellent [23:51] kai^sds: if you need any assistance, feel free to ping [23:52] well, I have attached the debdiff to the bugreport and anything else is beyond of my control anyway isn't it? ;) [23:53] bug 69455 [23:53] ([SRU] for oprofile, edgy-updates: bashism in oprofile's opcontrol script prevents user from setting any events) [23:53] ubotu: ping? [23:54] Cheers altogether btw! :o) [23:54] i am not ubotu but anyway: https://bugs.launchpad.net/ubuntu/+source/oprofile/+bug/69455 [23:55] wow. [23:56] kai^sds: but the bug is stilled "fix released". [23:56] Is this a regression? Has the patch been dropped for gutsy? [23:56] it is a regression and i can't change the status [23:57] the patch is for gutsy, yes [23:58] I've nominated it for release on gutsy and hardy. But someone else has to approve it, so it can track it's own status. [23:59] You can remove old diffs/attachments, can't you?