[12:13] <Kamion> mjg59: it'd be much appreciated if you could do something like my suggested usplash.postinst change above. I'm just idly browsing through my bug mail and I have quite a lot that's down to the edgy live filesystems not having built for a while.
[12:22] <mjg59> Kamion: Ok
[12:22] <mjg59> mdz: Kamion: A quick play suggests savings of around 50MB from not starting Gnome
[12:23] <mjg59> But
[12:23] <mjg59> On i810 systems, X is allocating a /huge/ amount of RAM on startup in order to be able to do 3D and stuff
[12:23] <mjg59> We could drop memory consumption by 100MB or so by asking it not to do that when just running the installer
[12:24] <mdz> mjg59: er, it allocates *100MB*?
[12:24] <mjg59> That's 100 including the 50 or so above)
[12:24] <mdz> oh
[12:24] <mjg59> mdz: No, probably more like 40
[12:24] <mdz> still, 50MB?
[12:26] <mjg59> mdz: Running X in 16 bits with no 3D, with ubiquity running, memory used is 93MB
[12:26] <Kamion> you need to run ubiquity up to partman or so to get to the memory use peak, really
[12:26] <mjg59> That's including the memory allocated for video
[12:27] <mjg59> Kamion: Yeah
[12:27] <Kamion> or possibly even gparted
[12:27] <Kamion> C++ being the fat piece of crap it is
[12:27] <mdz> mjg59: but once ubiquity has finished its job and you boot, those memory requirements go way up again
[12:27] <mdz> swap compensates somewhat, of course
[12:27] <mjg59> mdz: Indeed
[12:27] <mjg59> Kamion: Right, this is more a relative measurement than an absolute one
[12:28] <mjg59> But right now we have systems that will run Gnome acceptably that can't quite manage the desktop CD installer
[12:28] <mjg59> I think it's possibly worth reducing that requirement a bit
[12:33] <poningru_> real quick question http://plaza.ufl.edu/poningru/postinst
[12:33] <poningru_> doesnt that seem like bash and not sh
[12:33] <poningru_> perhaps /me should have waited till Seveas came in to ask that question
[12:34] <Kamion> poningru_: I see no bashisms there. What do you think is a bashism?
[12:35] <poningru_> the ' for variables
[12:35] <Kamion> poningru_: what line exactly?
[12:36] <poningru_> first and second
[12:36] <Kamion> distinguishing between ` and ' is rather important in shell
[12:36] <poningru_> wtf those are new lines
[12:36] <Kamion> but no, that's perfectly fine, nothing wrong with that in sh
[12:36] <HrdwrBoB> ` should be $(
[12:36] <Kamion> ` is fine
[12:36] <HrdwrBoB> for readability sake
[12:36] <Spads> $( ) are bash
[12:36] <Kamion> $( nests better but ` is fine
[12:36] <Seveas> HrdwrBoB, $( is a bashism
[12:36] <Kamion> Spads: also POSIX sh
[12:36] <Spads> ` ` is sh
[12:36] <elmo> Spads: no
[12:36] <HrdwrBoB> Seveas: er
[12:36] <Kamion> Seveas: wrong
[12:36] <elmo> Seveas: no
[12:36] <poningru_> ah
[12:36] <Spads> huh
[12:36] <Seveas> hmm
[12:36] <pitti> good night everyone
[12:36] <poningru_> thanks
[12:37] <Kamion> http://www.opengroup.org/onlinepubs/009695399/utilities/xcu_chap02.html#tag_02_06_03
[12:37] <poningru_> then I think the sh highlighting view in gedit has a bug
[12:37] <Spads> likewise vim
[12:37] <Kamion> poningru_: that is entirely unsurprising; editors aren't good metrics of what's valid
[12:38] <Spads> it flags them red when you're #!/bin/sh but not #!/bin/bash
[12:38] <Kamion> it probably means Bourne sh rather than POSIX sh
[12:38] <Kamion> which is good if you're on Solaris but nobody else cares any more
[12:38] <Spads> yeah, most likely
[12:38] <Seveas> ahhhh
[12:39] <elmo> Seveas: then you get to use the posix compatabile shell they provide
[12:39] <Kamion> poningru_: (by "nothing wrong with that", obviously it's poor quoting etc., although in practice that won't be a problem in this case)
[12:39] <elmo> it still doesn't make $( a 'bashism'
[12:39] <Seveas> elmo, actually I get to use tcsh :( 
[12:39] <Kamion> if anything it's a korn-sh-ism
[12:40] <elmo> no, I mean the one in /usr/xpg4/bin/sh or whatever it is
[12:40] <elmo> i.e. "the shell we'd like to be /bin/sh if we didn't have legacy systems of doom tieing us down"
[12:40] <Seveas> hehe
[12:42] <Seveas> dkaarsem@sremote ~ $ /bin/sh
[12:42] <Seveas> $ $(echo ls)
[12:42] <Seveas> syntax error: `$' unexpected
[12:42] <Seveas> yup, solaris sucks ;)
[12:55] <poningru_> rofl
[01:02] <imbrandon> lol @ Seveas
[01:37] <jono> night all
[02:12] <infinity> Kamion: No, /proc is definitely mounted during livefs builds...
[02:56] <MasterM> I love the concept behind ubuntu
[02:56] <MasterM> but i just prefer windows.
[02:56] <MasterM> ill use ubuntu as a live dvd
[02:57] <MarkShuttleworth> Hello ;)
[02:57] <Hobbsee> er...
[02:58] <Hobbsee> why are you under as that name?
[02:58] <MarkShuttleworth> It's Mark.
[02:58] <nixternal> heh, imposter again
[02:58] <MarkShuttleworth> I just came back from outta space.
[02:58] <MarkShuttleworth> How is everyone?
[02:58] <nixternal> outta space...must be an ebonics thing
[02:59] <MarkShuttleworth> hehe.
[02:59] <Hobbsee> MarkShuttleworth: rubbish.  dont impersonate people.  and this is a development channel, not a user channel.
[02:59] <MarkShuttleworth> I'm not. it's Mark i just came back from Mars.
[02:59] <jsgotangco> oh boy
[02:59] <nixternal> Hobbsee: he was here yesterday doing the same thing, actually had sabdfl choke ;)
[03:00] <MarkShuttleworth> lol
[03:00] <MarkShuttleworth> it's funny come on!
[03:00] <madduck> /ignore MarkShuttleworth    -- this is your one time chance. :)
[03:00] <MarkShuttleworth> I respect ubuntu...
[03:00] <madduck> MarkShuttleworth: uh, change your nick, sam.
[03:01] <jsgotangco> thank you
[03:01] <licio> :-|
[03:01] <MasterM> Can I ask a question?
[03:01] <ompaul> MAsterm  nalioth 
[03:01] <nixternal> lol
[03:01] <ompaul> MarkShuttleworth (n=sam@219-89-6-148.dialup.xtra.co.nz) has joined #ubuntu-devel
[03:02] <ompaul> or as he wanted to be called 
[03:02] <MasterM> do you all use ubuntu or do you use windows too?
[03:02] <nalioth> MasterM: have these nice folks asked you to depart?
[03:02] <nixternal> MasterM: #ubuntu-offtopic please, this is a developers channel. thank you
[03:02] <MasterM> Mok
[03:02] <MasterM> ok.
[03:02] <ompaul> I've already moved him out of #ubuntu
[03:03] <nalioth> yes, i know Mr. Sam
[03:03] <MasterM> #ubuntu-offtopic unable to join channel (address is banned)
[03:03] <MasterM> unban me
[03:03] <MasterM> please.
[03:03] <nalioth> ok, back to your regularly scheduled lives
[03:03] <LaserJock> nalioth: but this is the only fun we get ;-)
[03:04] <Fujitsu> Thankyou nalioth!
[03:04] <zul> i rather work..:)
[03:04] <nalioth> LaserJock: i can let him back in, if you like <EG>
[03:04] <LaserJock> nalioth: yikes, no thanks
[03:04] <nixternal> haha
[03:05] <nixternal> yesterday was funnier though
[03:05] <nixternal> sabdfl called him an imposter
[03:05] <imbrandon> heh
[03:05] <LaserJock> yeah, that was pretty funny
[03:05] <Fujitsu> Yep, that was good.
[03:05] <imbrandon> sab should probably reg that name anyhow just to twart peeps like that
[03:05] <zul> heh maybe sabdfl flew to nz to log on to irc
[03:05] <imbrandon> thwart
[03:06] <Fujitsu> Of course, zul! Why didn't we think of that :P
[03:06] <zul> oh goody he is on -bugs now
[03:06] <jsgotangco> ughh
    *cough choke splutter*
    imposter
[03:06] <nixternal> classic ^^
[03:07] <Fujitsu> Nice! I've never seen him do that before...
[03:07] <ajmitch> heh
[03:07] <Fujitsu> No.
[03:07] <Fujitsu> Only three people do.
[03:07] <Fujitsu> Seveas, dholbach, some other random person.
[03:07] <Fujitsu> coleSLAW, that's it.
[03:11] <Burgundavia> anybody else get a usb error -71 on the lastest desktop cd?
[03:11] <zul> dapper?
[03:11] <Burgundavia> edgy
[03:12] <Burgundavia> hmm, and LP search is being useless, again
[03:13] <Fujitsu> How?
[03:13] <zul> Burgundavia: all the more reason to open a new one with all of your info including dmesg
[03:14] <Burgundavia> zul: not a fatal error, so I ignored it
[03:14] <wasabi> So, opinions on launchd?
[03:14] <Burgundavia> wasabi: interesting, still a problematic license
[03:14] <Burgundavia> why could they not have chosen the lgpl?
[03:14] <wasabi> Is it?
[03:15] <wasabi> What's the license?
[03:15] <shackan_> apache
[03:15] <wasabi> Had assumed they would have chosen the same one they did for Bonjour.
[03:15] <Burgundavia> asl2 is not gpl-compatible or it is, depending on who you believe
[03:15] <Burgundavia> and they did
[03:15] <wasabi> Hmm. Was my impression the Bonjour one was BSD.  *goes to look it up*
[03:16] <Burgundavia> some of the bonjour stuff is, some isn't
[03:16] <ompaul> inarLLrion arpw dILWS 
[03:16] <ompaul> sorry 
[03:17] <wasabi> Ahh. Oh well. If the license is still a problem, any furthur talk is unneccassary.
[03:17] <shackan_> ompaul, get the cat off the keyboard :)
[03:17] <Burgundavia> not necessarily
[03:17] <wasabi> Any talk would center on the technical merits of replacing init heh
[03:17] <Burgundavia> the other concern I have is that launchd is controlled by apple
[03:17] <Burgundavia> given their past history, I am not certain that a fork won't emerge anyway
[03:17] <wasabi> Eh. If it was BSD, we could have just forked. ;)
[03:20] <shackan_> so they opened it but with such a license that noone can use it anyway ?
[03:20] <Kamion> wasabi: Keybuk says his implementation is already at the point where rebasing on launchd would be a backward step.
[03:20] <wasabi> Keybuk is working on something??!?!
[03:20] <Kamion> yes, has been for weeks
[03:20] <wasabi> Oh heck. Didn't get hte memo. url?
[03:20] <Kamion> http://wiki.ubuntu.com/ReplacementInit
[03:20] <wasabi> Duh. :0
[03:22] <Burgundavia> wasabi: he has also had lots of interest from other distros
[03:23] <wasabi> Okay. Cool. I'll read up and withdraw my interest in launchd. ;)
[03:35] <wasabi> Yes. I like keybuk's plan much better. It allows for a much more flexible system.
[03:35] <wasabi> I like the idea of broadcasting "depended on" as an event.
[03:36] <wasabi> It's also more in the unix style... communication through simple unix sockets, simple text based configuration.
[03:36] <wasabi> not xml.
[03:58] <bddebian> Heya
[04:50] <bluefoxicy> "among distributions that will break with the new 2.6.19 kernel are Ubuntu 6.06 LTS"  Who cares?  Seriously nobody is moving Dapper up past 2.6.15 and if they do then they can move udev up too.  The author of this article doesn't have a concept of distribution management.
[04:50] <bluefoxicy> <-- very bored
[08:12] <dholbach> good morning
[08:15] <pitti> good morning
[08:16] <dholbach> morning pitti!
[08:19] <pitti> hey dholbach!
[08:26] <mdeslaur> morning pitti
[08:27] <pitti> hi mdeslaur 
[08:29] <imbrandon> moins all
[09:03] <pitti> infinity: do you have a minute to talk about .ddebs?
[09:14] <jdub> whoa
[09:14] <jdub> major xgl/compiz blowup
[09:14] <jdub> compiz.real: Couldn't bind redirected window 0x1600020 to texture
[09:15] <dholbach> jdub: ajmitch is the new maintainer :)
[09:16] <infinity> pitti: Yup.
[09:17] <pitti> talking about crack, now that we have xorg 7.1, shouldn't we have aixgl now, too?
[09:18] <dholbach> pitti: isn't that xserver-xorg-air-core?
[09:18] <infinity> aiglx, even.
[09:18] <pitti> infinity: so, I guess copying the .ddebs to an outside place is possible, since it worked for translation tarballs, too; however, the problem will be the size of those .ddebs
[09:18] <infinity> pitti: Yeah, they'll be frickin' huge.
[09:19] <pitti> infinity: thus we need to ensure to not keep them on the buildds at all, and properly dominate them on people (or whereever)
[09:19] <pitti> infinity: can we distinguish between main/universe?
[09:19] <infinity> pitti: Yes, of course we can.
[09:19] <pitti> so maybe we should enable it for main only, at least for now
[09:19] <infinity> adconrad@rookery:~ $ df -h /      
[09:19] <infinity> Filesystem            Size  Used Avail Use% Mounted on
[09:19] <infinity> /dev/sda3             537G  279G  231G  55% /
[09:20] <infinity> I guess there's enough room, though elmo will kill us pretty quickly when we get enough of them up there.
[09:20] <Lathiat_> excuse my ignorance but whats a ddeb?
[09:20] <pitti> Lathiat_: debug symbols for the corresponding .deb
[09:20] <Lathiat_> ah right
[09:21] <pitti> infinity: my gut feeling is that we'll need the magnitude of 20 GB per architecture
[09:21] <infinity> pitti: I guess Team Soyuz is flat out right now, and asking them to just handle these in .changes files is out of the question? :)
[09:21] <pitti> infinity: well, that's the plan of course, but it won't happen anytime soon :(
[09:21] <infinity> (Probably very much the case)
[09:22] <infinity> pitti: Alright.  So, you're doing one .ddeb per binary package, right?
[09:22] <pitti> so this cowboying is mainly useful for developers to actually make sense of the crash reports we'll get
[09:22] <pitti> infinity: right
[09:22] <infinity> pitti: Alright.  If we're going to tie them to sources and such, I guess I'll have to generate a fake .dchanges file to tie source/version/ddebs, then send 'em to rookery.
[09:23] <pitti> infinity: would the following work: people.u.c. fetches ddebs from the buildds once a day, and the buildds remove them every day?
[09:23] <infinity> pitti: Domination will probably have to be as simplistic as "one per source package, per arch, highest version wins", rather than trying to actually match the archive contents.
[09:23] <pitti> infinity: domination> right, that's what I had in mind
[09:23] <MarcelDel> pitti, do you work on ubuntu?
[09:24] <pitti> infinity: and, for gradual disk space DoSing, maybe we should enable it for only i386 at first
[09:24] <Amaranth> Warning, that's the troll from earlier.
[09:24] <MarcelDel> troll?
[09:24] <infinity> pitti: Works for me, as long as you're sure the detached symbols work on all arches.
[09:24] <pitti> infinity: no, they won't
[09:24] <infinity> pitti: Would be a shame to only test on i386, and then realise that it actually breaks on, say, powerpc.
[09:24] <infinity> pitti: If it's broken on ia64, I don't care. :)
[09:24] <pitti> infinity: that's true of course
[09:25] <infinity> (And gdb is always broken on ia64 anyway)
[09:25] <infinity> pitti: I didn't mean "the one set of symbols should work everywhere", I meant "the concept should work everywhere", which would be nice to verify. :)
[09:25] <pitti> infinity: once it works, we can always enable more arches, right?
[09:25] <infinity> pitti: Well, yeah, obviously.
[09:25] <pitti> infinity: ah, right
[09:26] <pitti> infinity: ok, then I would write some scripts to fetch the ddebs from the buildds (stealing some stuff from lamont's scripts) and to clean up old versions
[09:29] <pitti> infinity: btw, we can free 41 GB on rookery by removing old translation tarballs :)
[09:29] <pitti> carlos: ^ oh, can we be reasonably sure to not need them any more? i. e. how is rosetta/edgy coming along?
[09:29] <infinity> pitti: Yeah, do we even need them being exported to rookery anymore at all?
[09:29] <pitti> infinity: I hope not :)
[09:29] <pitti> infinity: I haven't used them since before dapper's release
[09:30] <robtaylor> dholbach: when you get a moment, can you ping daf on #telepathy? he's been doing debian telepathy packaging, so i think you guys probably need to sync...
[09:30] <dholbach> ah nice
[09:31] <robtaylor> i think we have a repo for packaging now as well :)
[09:31] <dholbach> robtaylor: adding it to the join-on-startup list :)
[09:31] <robtaylor> dholbach: sweet :)
[09:31] <robtaylor> ok i need to go fix up my cycle.. laters :)
[09:31] <Treenaks> dholbach: watch out for the freenode channel limit
[09:32] <dholbach> Treenaks: yeah :)
[09:36] <infinity>     apport |        0.8 | edgy/universe | source, all
[09:36] <infinity> pitti: It's published on drescher, so if you have a stale archive mirror, yell at Znarl. :)
[09:37] <pitti> infinity: hm, in order to avoid race conditions, we need per-date subdirectories on the buildds
[09:38] <infinity> pitti: Already had that for translations anyway.
[09:45] <infinity> sabdfl: Registering markshuttleworth as an alternate nick?
[09:45] <pitti> infinity: so, if rookery fetches http://*.buildd/~buildd/ddeb/2006MMDD/index.txt and everything mentioned in index.txt every day at 0030 UTC, and the buildds remove this dir at 0200 UTC every day, woudl that work for you?
[09:45] <sabdfl> infinity: yes
[09:45] <infinity> pitti: Assuming that's YYYYMMDD-1, then yes. :)
[09:45] <sabdfl> infinity: it told me that it would expire in 60 days unless I identify to it
[09:45] <sabdfl> is there a way to automated that to hold onto a set of nicks?
[09:45] <infinity> Whine to lilo?
[09:45] <infinity> Otherwise, not sure.  I don't use altnicks.
[09:45] <infinity> Of course, you could write a quick irssi perl script to do the identification for you every 24 hours or something.
[09:45] <infinity> (Or whatever client you use, X-chat has scripting too)
[09:45] <sabdfl> good call - lilo fixed it
[09:45] <infinity> Well, I think your cloak screams "lilo needs to pay attention to me occasionally", so that works. :)
[09:45] <pitti> infinity: what would be a convenient format of index.txt for you?
[09:45] <infinity> pitti: Do we even need one?
[09:45] <pitti> infinity: does buildd's apache have autoindex?
[09:45] <infinity> pitti: Yeah.
[09:45] <pitti> ok, then we don't
[09:45] <infinity> pitti: And if it doesn't, I can turn it on. :)
[09:45] <pitti> yep, seems to work
[09:45] <infinity> pitti: I will generate source_version_arch.dchanges files from sbuild that list all the ddebs, and the source name and version.
[09:45] <infinity> pitti: Which should make domination and reaping a bit tidier.
[09:45] <pitti> infinity: that's mainly for correct epoch handling, right?
[09:45] <pitti> oh, also for NBS
[09:45] <infinity> pitti: Well, that and being able to tie all the ddebs together as a single entity.
[09:45] <pitti> ok, great
[09:45] <pitti> I hope that won't be too much work for you
[09:45] <lucas> ouch. both kdeedu and kgeography src packages generate a kgeography binary package
[09:45] <infinity> Nah, that's easy enough.  It won't be a full .changes file, just a cut-down version with only the fields required.
[09:45] <lucas> how is that allowed ?
[09:45] <infinity> lucas: It means one of them needs to stop, that's all.  Binary packages move between source packages all the time.
[09:45] <pitti> lucas: whichever has the higher version number wins
[09:46] <pitti> in this case, kdeedu wins both version-wise and component-wise
[09:54] <carlos> pitti: I got the approval to merge the code that will open edgy based on dapper
[09:54] <carlos> pitti: so it's a matter of polish it a bit today
[09:54] <pitti> carlos: cool
[09:54] <carlos> and merge it into production
[09:54] <Hobbsee> hi carlos, pitti 
[09:54] <pitti> carlos: I mean, can we stop the buildd->people hack for translation tarballs?
[09:55] <carlos> pitti: I think so. It's useful for debugging purposes but it's not a must
[09:55] <carlos> Hobbsee: hi
[09:55] <carlos> will be back in 30 minutes
[10:25] <hunger_work> pitti: I replied to your questions on #54530.
[10:25] <pitti> ah, thanks
[10:28] <pitti> hunger_work: "/dev/zero can not be mmaped" sounds a bit like a noexec fallout, although I don't see the point of mmap'ing /dev/zero in the first place ;)
[10:29] <hunger_work> pitti: Yeap, I thought so too, that is why I mentioned it.
[10:29] <hunger_work> pitti: OTOH /dev/zero's permission do not include execute, so noexec shouldn't matter.
[10:30] <pitti> hunger_work: however, mmap() allows you to specify PROT_EXEC
[10:30] <hunger_work> pitti: mmapping /dev/zero is a trick to get a big pre-zeroed area of memory.
[10:31] <pitti> bah, isn't memset() much more effective for that?
[10:31] <hunger_work> pitti: PROT_EXEC is about the pages in momery. I doubt that has anything to do with the filesystem permissions.
[10:31] <pitti> I don't know it either, it just came to my mind as a possible explanation
[10:31] <StevenK> Or calloc()?
[10:32] <hunger_work> IIRC the kernel catches mmaps to /dev/zero and does some magic... so it shouldn't matter.
[10:34] <pitti> ogra: ?
[10:34] <hunger_work> pitti: You might be right: PROT_EXEC may not conflict with the open mode of the file. Maybe someone opened /dev/zero with exec set? Dunno whether that is at all possible for a file that is in mode 666:-)
[10:34] <ogra> why doesnt it have a os.mount command ... grmbl
[10:35] <pitti> ogra: it's not portable at all, that might be the reason
[10:35] <ogra> yeah 
[10:35] <ogra> but it would be sooo elegant :)
[10:36] <pitti> well, subprocess.call(['/usr/bin/mount', device] ) isn't that much worse :)
[10:36] <ogra> sure, but i need to call subprocess :)
[10:37] <hunger_work> pitti: And that with python not even having a cpp to handle such issues;-)
[10:37] <ogra> os.mount(['/usr/bin/mount', device] ) would be cooler :)
[10:37] <pitti> ^ looks funny
[10:37] <ogra> well ...
[10:37] <ogra> *grin*
[10:37] <pitti> ogra: #define subprocess.call os.mount ;)
[10:38] <ogra> heh
[10:38] <pitti> (pre-processor in python 9.3)
[10:38] <StevenK> pitti: 9.3 ... ?
[10:38] <pitti> StevenK: well, it's not yet in 2.5, so I guesstimated :)
[10:39] <ogra> StevenK, might also be in 9.2 already  ;)
[10:39] <pitti> (well, preprocessor is not pythonic at all, they won't add it anytime soon)
[10:40] <StevenK> pitti: :-P
[10:41] <ogra> pitti, btw, whats the reason we dont use dev.cdrom.lock=0 in /etc/sysctl.conf ? is there some drawback i'm not aware of ?
[10:42] <pitti> ogra: bug 35695
[10:42] <Ubugtu> Malone bug 35695 in procps "Please change sys/dev/cdrom/lock default to 0" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/35695
[10:42] <pitti> ogra: however, we need to make sure to clean up properly (lazy unmounting)
[10:43] <pitti> ogra: and since hal/g-v-m handle the drive eject button nowadays in a much cleaner way, I didn't bother to work on that
[10:44] <ogra> heh, supermount-ng
[10:44] <ogra> the name of superlatives 
[10:44] <ogra> pitti, well, i have no g-v-m running on the thin clients :)
[10:45] <ogra> but i'll implement a unlock call in the python script that mounts/monitors the cd
[10:46] <ogra> i was just wondering why we dont have it by default ...
[10:46] <pitti> ogra: instead of changing the sysctl default, maybe you can just set it dynamically in ltsp?
[10:46] <ogra> but the bug explains it ... even though xubuntu might probaby want to enable it
[10:47] <ogra> pitti, yes, see above, thats what i'm doing :)
[10:47] <ogra> fcntl.ioctl(f, CDROM.CDROM_LOCKDOOR, 0)
[10:47] <pitti> ah, cool
[11:02] <Hobbsee> hi all.  i'm trying to merge:  magic-haskell, and get magic-haskell_1.0.3-0.1ubuntu1.dsc: Version older than that in the archive. 1.0.3-0.1ubuntu1 <= 1.0.3ubuntu1.  what should i upload the new version as, to overwrite both ubuntu and debian versions?
[11:03] <pitti> Hobbsee: 1.0.3ubuntu2 seems most logical
[11:03] <pitti> erm, wait, no
[11:03] <pitti> 1.0.3ubuntu1 is a native package version
[11:03] <Hobbsee> pitti: that's what i thought, then wait to fix it for the new upstream version, whenever that happens?
[11:03] <pitti> so this is seriously screwed
[11:03] <Hobbsee> yes.  heh
[11:04] <pitti> Hobbsee: I'd propose 1.0.3ubuntu1-1
[11:04] <Gloubiboulga> pitti, I think we've used versions like 1.0.3ubuntu2 already
[11:04] <Gloubiboulga> Toadstool, around?
[11:04] <Hobbsee> pitti: right.  which is native again, presumably
[11:05] <Toadstool> Gloubiboulga: yep, waking up :)
[11:05] <StevenK> Hrm, that's a point.
[11:05] <Hobbsee> pitti: sistpoty was the one who seemed to version it that way.  hmm.
[11:05] <Toadstool> hi everybody
[11:05] <pitti> Hobbsee: ah, ok; Debian just had an NMU
[11:05] <Hobbsee> hi Toadstool 
[11:05] <Gloubiboulga> Toadstool, I think you've already had a similar issue, right?
[11:05] <pitti> Hobbsee: so it indeed *is* supposed to be a native package
[11:05] <pitti> Hobbsee: 1.0.3ubuntu2 is fine then
[11:05] <Hobbsee> right
[11:06] <Toadstool> Gloubiboulga: misversioned NMU for Debian native packages?
[11:06] <Gloubiboulga> yes
[11:06] <pitti> well, not exactly 'mis'versioned
[11:06] <Hobbsee> pitti: okay, cool
[11:06] <StevenK> pitti: Sure it is.
[11:06] <pitti> debian NMUs are not supposed to 'steal' upstream version numbers
[11:06] <pitti> although 1.0.3.1 would have been more appopriate
[11:06] <StevenK> That's what I was going to suggest.
[11:06] <azeem> there's some dispute on how to version NMUs for native packages I think
[11:07] <StevenK> Add a .1 seems to work.
[11:07] <pitti> I still remember doing an NMU for one of iwj's packages and got seriously larted for stealing an upstream version number
[11:07] <Toadstool> well, i said misversioned 'cause it breaks the -ubuntuX scheme ;)
[11:10] <azeem> Toadstool: one could argue Ubuntu uses a broken versioning scheme for native packages :)
[11:10] <Toadstool> :)
[11:22] <robtaylor> azeem: -$DEBVER[.$DEBNMU] -ubuntu$UBUNTUVER[.$UBUNTUNMU]  ? ;)
[11:23] <dholbach> robtaylor: there are no NMUs
[11:24] <dholbach> robtaylor: in ubuntu :)
[11:25] <infinity> No, but we use NMU versioning in -security and -updates to avoid version overlap with newer releases.
[11:30] <pitti> oh, it's only usable from the <pid> process itself. how (un)useful
[11:38] <pitti> mvo: this burning black baloon scares me :)
[11:42] <mvo> pitti: what a couple of seconds and it will explode ;)
[11:43] <pitti> mvo: an exploding baloon? argh :)
[11:43] <pitti> mvo: maybe we should turn it upside down and mape it less pear-shaped
[11:43] <pitti> mvo: s/mape/make/
[11:43] <pitti> mvo: I'll play with it a bit
[11:45] <mvo> :)
[12:24] <infinity> Can anyone think of a filesystem, other than vfat, that has great read/write support in both Win32 and Linux, and doesn't suck?
[12:24] <Fujitsu> No.
[12:24] <Fujitsu> There isn't one.
[12:24] <Fujitsu> smbfs :P
[12:25] <infinity> Local filesystem. :)
[12:25] <Fujitsu> UDF? :P
[12:25] <infinity> Yeah, I thought of that.  Not sure what Windows will do with a hard drive formatted with UDF.
[12:25] <Fujitsu> I was wondering if it'd like it...
[12:26] <jdub> infinity: iso9660?
[12:26] <infinity> jdub: iso9660 is read-only.
[12:26] <jdub> infinity: devices are read-only :-)
[12:27] <infinity> If I wanted read-only, I'd be happy with NTFS, or even vfat, but I want something format an extrenal hard drive for sneakernet backups.
[12:28] <StevenK> infinity: What about ext2? You can get a extension for Windows to read/write it.
[12:28] <Fujitsu> And what's wrong with vfat?
[12:28] <azeem> "can get"
[12:28] <tseng> its kind of crap how that works
[12:28] <infinity> StevenK: Yeah, I'm reading up on those now.  They seem to have improved since I last looked.
[12:28] <tseng> besides not being OOTB
[12:28] <Fujitsu> infinity, they work great.
[12:28] <infinity> Fujitsu: vfat's just... Fragile.
[12:29] <infinity> Fujitsu: Especially for 400GB of data.
[12:29] <Fujitsu> infinity, that's true.
[12:51] <dholbach> ogra: if you do the gnome-screensaver update: please drop the schema patch and use debian/gnome-screensaver.gconf-defaults - you can use gnome-applets as an example to look at
[12:51] <dholbach> ogra: that's less pain for the next upstream schema changes
[12:52] <ogra> ok
[12:54] <dholbach> gracias
[01:04] <pitti> Kamion: are there any further issues with the dapper update that need testing?
[01:06] <Kamion> pitti: no - in fact publication is in progress
[01:06] <pitti> yay
[01:08] <pitti> Hobbsee: ping
[01:09] <pitti> ajmitch: ping
[01:10] <ajmitch> pitti: pong
[01:10] <pitti> ajmitch: I'd like to make you and Hobbsee administrators for ubuntu-universe-sponsors; are you fine with that?
[01:10] <ajmitch> sure
[01:15] <Kamion> pitti: when was the last thing you published to dapper-security?
[01:15] <pitti> Kamion: this morning, around 0700 UTC or so
[01:16] <Kamion> pitti: doh. can you give me a list of what you've published to dapper-security over the last four days or so?
[01:16] <StevenK> Kamion: Hey, aren't you on holidays?
[01:16] <Kamion> because I need to make sure to keep the state of the archive at dapper.1
[01:16] <pitti> Kamion: only libwmf from this morning
[01:16] <Kamion> StevenK: doing dapper point release
[01:16] <Kamion> (otherwise yes)
[01:16] <pitti> Kamion: it's a small update, doesn't really need to be on the CDs
[01:17] <Kamion> pitti: I know, I just need to construct an archive snapshot minus that
[01:17] <Kamion> i.e. what matches the CDs
[01:17] <pitti> oh, I see
[01:17] <pitti> Kamion: it might be helpful that this is the first libwmf security update ever
[01:17] <Kamion> because we don't have the proper soyuz stuff yet
[01:17] <Kamion> it is :)
[01:18] <Kamion> ok, I'll just hack that out
[01:18] <pitti> Kamion: i. e current archive minus libwmf {hoary,breezy,dapper}-security versions is the thing you want
[01:18] <pitti> sorry for the trouble
[01:19] <Kamion> no problem
[01:19] <Kamion> I should have taken the archive snapshot a few days ago really
[01:20] <pitti> Kamion: previous security update was gnupg from last Thursday
[01:21] <pitti> Kamion: 1.4.2.2-1ubuntu2.2 for dapper
[01:21] <pitti> Kamion: and that's on the CDs
[01:24] <Kamion> yep
[01:25] <pitti> wb carlos 
[01:25] <carlos> pitti: hi
[01:41] <Hobbsee> pitti: pong
[01:45] <Hobbsee> pitti: sounds good to me
[01:45] <pitti> Hobbsee: hereby I declare you a master of the universe sponsors :)
[01:46] <Hobbsee> pitti: yay :)
[01:46] <Hobbsee> pitti: what can i do now then?
[01:46] <ajmitch> Hobbsee: kick the MOTUs into sponsoring stuff
[01:46] <pitti> Hobbsee: well, since it's an open team there isn't much to administrate
[01:46] <Hobbsee> pitti: right
[01:46] <pitti> Hobbsee: but in the long term I'd like to hand ownership to a MOTU
[01:46] <Hobbsee> pitti: so i can just cause general havoc with it
[01:46] <ajmitch> like usual
[01:46] <pitti> since I think MOTUs should be self-supporting
[01:46] <ajmitch> pitti: agreed
[01:46] <gnomefreak> dholbach: ping
[01:47] <pitti> Hobbsee: you can kick people and write them rude comments about expelling them, and such :)
[01:47] <Hobbsee> pitti: my first response to that was "so why are you handing it over to me"
[01:47] <Hobbsee> pitti: ooh!  fun!
[01:47] <pitti> Hobbsee: I didn't hand it over yet, but if you want to, I'll do it now
[01:47] <Hobbsee> pitti: sounds good to me :)
[01:47] <gnomefreak> Hobbsee: your trustworthy and good at what you do?
[01:47] <Hobbsee> pitti: my point was more, it took me a while to realise that i fit into that MOTU category
[01:48] <gnomefreak> congrats Hobbsee ;)
[01:48] <Hobbsee> gnomefreak: me?  trustworthy?  you talking in irc, or in anything important here?
[01:48] <Hobbsee> i think that most people know that by now :P
[01:49] <pitti> Hobbsee: it's your's now
[01:49] <Hobbsee> pitti: thankyou :)
[01:49] <gnomefreak> if i compiled a tar without issues does that mean its a good chance that building it for ubuntu should be ok?
[01:50] <Hobbsee> pitti: i even *own* it.  cool!
[01:50] <pitti> Hobbsee: now it's *your* job to find a nice emblem, not mine any more :-P
[01:50] <ajmitch> Hobbsee: that's sort of what handing over means :)
[01:51] <Hobbsee> pitti: hehe!  that's what i was just thinking about
[01:51] <Hobbsee> pitti: where's the recommended place for emblems?
[01:51] <Hobbsee> infinity: hehe.  emblems are pretty, anyway
[01:52] <Hobbsee> hmm..  theres' a field marked contact address too.
[01:52] <pitti> Hobbsee: 'where'?
[01:52] <Hobbsee> hmmm?  where's as in, where is
[01:52] <pitti> Hobbsee: usually it's in the blue box with your Karma and name
[01:52] <infinity> Hobbsee: Make one!  Pixel art is fun.
[01:52] <infinity> Hobbsee: And you can't actually upload it without being a team admin, or an LP admin.
[01:52] <Hobbsee> pitti: no, i realise that.  more was a question of "where do you usually get your emblems from"
[01:52] <pitti> Hobbsee: yes, they are nice badges you can pin on your shoulder, or LP homepage for that matter :)
[01:53] <Hobbsee> pitti: hehe
[01:53] <pitti> Hobbsee: so far I stole them from /usr/share/icons, I'm not an artwork dude at all
[01:53] <Hobbsee> infinity: what, the team owner doesnt count in that?  :P
[01:53] <Hobbsee> pitti: ahhhh...
[01:53] <infinity> Hobbsee: I mangle GNOME icons and bend them to my will for fun and profit.
[01:53] <Hobbsee> infinity: hehe, right
[01:53] <pitti> infinity: she just became admin and owner of ubuntu-universe-sponsors
[01:54] <infinity> pitti: Ahh, I was looking at the wrong sponsors team (the one I'm a member of, for main) cause I'm half asleep.
[01:54] <infinity> pitti: Should I make an emblem for that one for you? :)
[01:54] <pitti> infinity: if you feel like it, of course :)
[01:55] <pitti> doing this stuff occasionally is fun, some months ago I spent about an hour doing a Hackergotchi for me
[01:55] <infinity> pitti: It's a good way to waste "awake, but not awake enough to do useful work" time.
[01:55] <pitti> Hobbsee: right, while you are at it, create a hackergotchi for you :)
[01:55] <Fujitsu> Yep.
[01:55] <pitti> infinity: heh, I know that feeling, for me it's usually the time around midnight
[01:55] <ajmitch> pitti: or we could supply one for her
[01:55] <Hobbsee> infinity: art never was my strong point.  although i do get the credit for kubuntu edgy being purple, as i suggested it :P
[01:56] <Hobbsee> pitti: heh.  i break cameras.
[01:56] <infinity> Hobbsee: Ahh, I've been doing various visual art for longer than I've been programming, so the pixel art is kinda midnless and fun.
[01:57] <infinity> mindless, too.
[01:57] <Hobbsee> oh, who's the channel creator person in here?
[01:57] <Hobbsee> ie, the top top person for this irc channel
[01:57] <infinity> If only I could also upload theme music for my teams.  It'd be all like LaunchMySpacePad.
[01:57] <pitti> lol
[01:57] <StevenK> Hobbsee: ChanServ says thom.
[01:58] <Hobbsee> StevenK: hmm.  never spoken to him before, iirc
[01:58] <TheMuso> Even though from my pov, I'm not.
[01:58] <Hobbsee> heh
[01:58] <Hobbsee> infinity: ooh, scary.  does that mean i can add a custom background to the page, also matching my myspace? *g*
[01:58] <Hobbsee> dont you think so, StevenK?
[01:58] <infinity> Hobbsee: For universe-sponsosrs/main-sponsors, I was thinking they typical "helping hand" (ie, two hands, interlocked) iconography for universe, and the "iron fist" for main.
[01:59] <StevenK> AAAIIIIIIIIIIIIIEEEEEE
[01:59] <Hobbsee> ROFL!
[01:59] <infinity> thom: You still own #-devel.
[01:59] <Hobbsee> thom: heya, can we get some more ops in this channel please?
[01:59] <thom> yay me
[01:59] <infinity> thom: Apparently.
[01:59] <StevenK> TheMuso: Mr Planetary Tramp!
[01:59] <TheMuso> StevenK: heh
[02:00] <StevenK> DOH!
[02:00] <StevenK> thom: Mr Planetary Tramp!
[02:00] <thom> StevenK: still not mastered irc? ;p
[02:00] <thom> Hobbsee: um, yes
[02:00] <StevenK> It's not IRC, it's typing. :-P
[02:00] <infinity> pitti: That could be tricky.  How about just the curled edge of old-skool tractor-feed paper?
[02:00] <thom> i'm open to suggestions
[02:00] <maswan> pitti: one printing text (or something else variable) in graphics mode I hope?
[02:00] <pitti> infinity: but the dot matrix printer makes you fall off your chair much harder
[02:00] <TheMuso> StevenK: It sort of applies. :p
[02:01] <pitti> at least the ones I heard in my life
[02:01] <thom> Fujitsu: hey, give that back!
[02:01] <infinity> pitti: I do like the satisfying zip of an old Raven 24.
[02:01] <Hobbsee> thom: right.  i'm one, i'm not sure who beyond that.
[02:02] <maswan> I wonder if I could manage to get an apple imagewriter ii up and running
[02:02] <pitti> infinity: the old Epson I once had sounded more like a circular saw
[02:02] <infinity> thom: I suppose I could be added to the op list, since I (apparently) never sleep.
[02:02] <thom> Hobbsee: launchpad id? 
[02:02] <Hobbsee> thom: should be under as hobbsee.  i think
[02:02] <thom> infinity: you were about to have it inflicted indeed
[02:02] <Hobbsee> hi sabdfl 
[02:02] <maswan> It had a rather nice and smattering sound, in addition to the whine
[02:02] <Fujitsu> Hi sabdfl!
[02:02] <tseng> infinity: I don't sleep, either
[02:02] <sabdfl> hey guys
[02:02] <pitti> welcome back sabdfl 
[02:02] <tseng> morn sabdfl 
[02:03] <gnomefreak> Hobbsee: it looked like you had 2 ;)
[02:03] <tseng> the real one, this time
[02:03] <gnomefreak> hello sabdfl 
[02:03] <Hobbsee> gnomefreak: say what?  2?
[02:03] <Fujitsu> tseng, hahah.
[02:03] <gnomefreak> Hobbsee: thats what a search pulled up 1 with no karma and 1 with 200k
[02:04] <Hobbsee> ah ha!  thanks thom 
[02:04] <tseng> oh cool
[02:04] <pitti> Hobbsee: accumulating power today? :)
[02:04] <Hobbsee> pitti: yeah.  arent i always?
[02:04] <Hobbsee> :P
[02:04] <pitti> phear!
[02:05] <Hobbsee> does it work?
[02:05] <tseng> Hobbsee: how cute
[02:05] <Fujitsu> Please spare me, for I am just a poor little MOTU-wannabe.
[02:05] <Hobbsee> HAH!
[02:05] <TheMuso> I can't escape. :)
[02:05] <thom> there's a reasonable collection of people on the list now
[02:05] <pitti> Hobbsee: can you please hide the whip? it scares me off
[02:05] <Hobbsee> no, cute is *not* the correct answer to that
[02:05] <Hobbsee> pitti: but...but...it's my whip....
[02:05] <ajmitch> interesting evening in -devel 
[02:06] <Hobbsee> ajmitch: very.  i'm in it, what do you expect
[02:06] <pitti> yeah, I like these totally off-topic nonsense conversations :)
[02:06] <Hobbsee> ajmitch: besides, if i do any more uploads, edgy-changes will probably break.
[02:06] <ajmitch> yes, things tended to go downhill fast
[02:06] <infinity> pitti: Was that a hint? :)
[02:06] <TheMuso> Hobbsee: Need a cane?
[02:06] <Hobbsee> TheMuso: i do have a walking stick at work.  it tends to work better.
[02:06] <pitti> infinity: rather an expression of my utter enjoyment :)
[02:06] <pitti> infinity: I like to watch Al Bundy, too :)
[02:07] <thom> hrm, launchpad ought to have retrospective karma
[02:07] <TheMuso> Yeah. It probably doesn't fall apppart with the slightest tap. :)
[02:07] <infinity> Hobbsee: Now that you've been added to the op list, you're never allowed to be offtopic, ever again.  Much like the zero-tolerance policy for poolice and drugs, and such.
[02:07] <Hobbsee> infinity: what????
[02:07] <infinity> thom: Karma times out.
[02:07] <Hobbsee> you mean i cant be hobbsee any more?
[02:07] <thom> infinity: really? ah
[02:07] <Hobbsee> infinity: but have to be the sane, prim and proper sarah hobbs?
[02:08] <ajmitch> Hobbsee: yes
[02:08] <infinity> thom: Yeah, recent contributions count for more than crusty old ones, and eventually, the crusty stuff pops off completely.
[02:08] <TheMuso> Hobbsee: Don't let us males boss you around. :)
[02:08] <Hobbsee> ajmitch: you know better than most that that's not possible.
[02:08] <Hobbsee> TheMuso: haha, never.  i can get you guys wrapped around my little finger, no problem :P
[02:09] <Hobbsee> infinity: but i was enjoying my 200K of karma :P
[02:09] <Riddell> what's the daemon that looks after cpu frequency called?
[02:09] <infinity> Riddell: powernowd
[02:09] <Riddell> hmm, I don't seem to have powernowd running
[02:09] <Riddell> "* CPU frequency scaling not supported"  that'll explain it
[02:09] <ajmitch> TheMuso: see how quickly she took over a team & got on the channel list tonight
[02:10] <Hobbsee> ah ha, we have an icon now :)
[02:11] <Hobbsee> but i dont get an icon on my LP page :(
[02:11] <infinity> Hobbsee: Refresh.  I see it on your page.
[02:11] <ajmitch> Hobbsee: a gold star?
[02:11] <ajmitch> how special
[02:11] <Hobbsee> yeah
[02:12] <pitti> Hobbsee: yo do
[02:12] <pitti> Hobbsee: you, even
[02:12] <Hobbsee> pitti: ah, right.  okay then 
[02:12] <infinity> Hope I don't run into copyright issues with Nintendo.
[02:12] <Fujitsu> Haa.
[02:12] <Hobbsee> hah
[02:12] <Hobbsee> actually, i saw a pretty fish that you could use, if you wanted.
[02:13] <pitti> infinity: a chauffeur cap might be a nice one for a sponsoring team
[02:13] <infinity> pitti: Bah, I was just drawing the mushroom.
[02:13] <Fujitsu> True, pitti.
[02:13] <infinity> pitti: You eat it and it makes you larger (ie: able to upload to main).. It totally makes sense. :)
[02:14] <infinity> (to me)
[02:14] <infinity> (late at night)
[02:14] <Fujitsu> Hahaha.
[02:14] <Hobbsee> infinity: hehe.  right.  i suspect you're very tired
[02:14] <Fujitsu> I should probably pop off into bed in a moment...
[02:15] <pitti> infinity: hah, I know! A syringe
[02:15] <pitti> since you can inject packages to something you usually don't have access to
[02:16] <sivang> re
[02:16] <pitti> Hobbsee: since infinity loves the mushroom so much, maybe you consider a syringe for the universe team
[02:16] <pitti> hi sivang, how are you?
[02:16] <Fujitsu> Hi sivang.
[02:16] <sivang> pitti: fine thanks, and you Martin ?
[02:16] <sivang> hey Fujitsu 
[02:17] <pitti> sivang: I'm well
[02:17] <Hobbsee> Fujitsu: hahaha!
[02:17] <Hobbsee> Fujitsu: how many are there now?  i havent counted
[02:17] <Fujitsu>  Neither have I.
[02:17] <Fujitsu> And you haven't been a MOTU for that long >_<
[02:17] <Hobbsee> pitti: hehe, yeah.  i'm not into needles much
[02:18] <Hobbsee> Fujitsu: yeah, dont start me on that :P
[02:18] <Fujitsu> Hobbsee, did you end up looking through my merges other than mol?
[02:18] <Hobbsee> Fujitsu: nope, didnt know what you'd built, etc
[02:18] <Fujitsu> Ah.
[02:19] <infinity> There.  Screw you all: https://launchpad.net/people/ubuntu-main-sponsors
[02:19] <infinity> :P
[02:19] <Hobbsee> haha nice, infinity!
[02:19] <Fujitsu> Nice, infinity!
[02:19] <pitti> oooh, what a nice mushroom!
[02:19] <Hobbsee> hi rodarvus 
[02:19] <pitti> thanks, infinity! :)
[02:19] <Fujitsu> It looks great!
[02:19] <Hobbsee> pitti: no, you cant eat it :P
[02:20] <rodarvus> hey Hobbsee
[02:20] <ajmitch> I seem to be collecting emblems on my launchpad page quite quickly
[02:20] <Hobbsee> good pitti.  and see, i didnt even have to use my whip on you
[02:20] <Hobbsee> ajmitch: it's more fun that way :)
[02:20] <Fujitsu> Hobbsee, I check-built all of them except mol, as it only builds on PowerPC, and my only PowerPC is at school.
[02:21] <Fujitsu> Wow... ajmitch, you have a lot of emblems.
[02:21] <ajmitch> Fujitsu: yes, I should leave some teams
[02:21] <Riddell> Hobbsee: do it
[02:22] <TheMuso> Although the emblems will probably be a little small.
[02:22] <infinity> Fujitsu: I have a few more.  I need to stop making emblems for the teams I join.
[02:22] <Fujitsu> I've just got the two :P
[02:22] <Fujitsu> A nice not-two-line list :P
[02:22] <Hobbsee> Riddell: i can do it?
[02:23] <infinity> Fujitsu: adconrad
[02:23] <Fujitsu> I knew it was something like aconrad or something.
[02:23] <mdke> freenode isn't using LP for authentication yet?
[02:23] <mdke> outrageous
[02:24] <Hobbsee> infinity: are you planning to do your universe merges?  presumably not
[02:24] <infinity> mdke: Please don't suggest that when sabdfl is in here.
[02:24] <ajmitch> mdke: would you trust your launchpad password going across irc?
[02:24] <mdke> ajmitch: was a joke
[02:24] <infinity> Hobbsee: I already checked them, and they were all just minor build failure fixes.  I gave the go-ahead to bddebian to do them, but you're welcome to.
[02:24] <ajmitch> mdke: I know, but there's always a risk that some people take it seriously
[02:25] <Hobbsee> infinity: right.  i've just been stealing merges at random, i think
[02:25] <mdke> ajmitch: guess so. The serious thing underneath was that it is nice when people have the same irc nick and LP nick... it's easier to find em
[02:25] <infinity> Hobbsee: The only remotely complex one was dmraid.  If no-one's touched that yet, and it scares you, ping me and I'll do it.
[02:25] <Hobbsee> ah yes.  
[02:27] <Hobbsee> gnomefreak: :) good man
[02:28] <infinity> Hobbsee: mdz, generally.
[02:28] <tseng> Hobbsee: mdz, Kamion 
[02:28] <pitti> Hobbsee: me
[02:28] <gnomefreak> man pages adn docs are generated for you when building a package right?
[02:28] <pitti> Hobbsee: oops, sorry; mdz
[02:28] <infinity> pitti: You do UVF exceptions now, or did you read that as MIR?
[02:28] <tseng> infinity: UnhandledException
[02:29] <Hobbsee> pitti: haha.  
[02:29] <desrt> pitti; it's dbus's fault.
[02:29] <tseng> desrt: seriously. I have the same bug in last-exit now and we dont even use gnomevfs (directly)
[02:29] <pitti> whoa, these always make my brain explode
[02:29] <pitti> hi Keybuk 
[02:29] <desrt> Hobbsee; bad call.
[02:29] <tseng> desrt: arse.
[02:29] <Keybuk> pitti: hey
[02:29] <Hobbsee> hi Keybuk 
[02:30] <Hobbsee> hehe
[02:30] <Fujitsu> Evening Keybuk.
[02:30] <desrt> Hobbsee; give him one of those and he won't recover until well after edgy
[02:30] <Hobbsee> desrt: haha.  that could be fun
[02:30] <desrt> fun for you.  bad for edgy.  :p
[02:30] <Hobbsee> desrt: well, seeing as i run edgy...
[02:31] <desrt> people would be like "hey!  cool new package that gives you a root shell when you telnet to the box without the need for pesky passwords!  it's secure because it runs on a high port number!"
[02:31] <desrt> and without pitti around, this sort of thing would be on the install CD
[02:31] <Hobbsee> Fujitsu: it already does.  it's hardlocked twice in two days.
[02:31] <Fujitsu> :O
[02:32] <mdke> btw, what package is it a bug in if when I login via ssh I get the "no warranty" message and uname -a repeated twice?
[02:32] <desrt> mdke; look at your /etc/motd.  is the message there twice?
[02:32] <mdke> desrt: no
[02:32] <desrt> weird.
[02:33] <desrt> does a hush file in your ~ get rid of one or both?
[02:34] <mdke> desrt: what do I need to do to test that?
[02:34] <desrt> touch .hushlogin
[02:34] <desrt> then login
[02:34] <desrt> supresses motd, etc. on login
[02:34] <mdke> and should I make that file?
[02:34] <desrt> ya.  touch will make it for you.
[02:34] <mdke> ah
[02:34] <desrt> in ~
[02:34] <dholbach> gnomefreak: pong
[02:34] <mdke> desrt: now I get 0
[02:35] <desrt> mdke; so the bug is in a hushlogin-respecting program
[02:35] <mdke> btw the "last login" details appear only once, after the first instance of /etc/motd
[02:36] <gnomefreak> i was able to compile gimmie ok. im gonna attempt to build it (to learn) but i remember you wanted to add that in edgy
[02:36] <mdke> then comes the second instance of it
[02:36] <desrt> grepping for hushlogin i get dropbear and sshd
[02:36] <tseng> gnomefreak: it doesnt have a make install
[02:36] <mdke> desrt: so, ssh-server i guess has the bug?
[02:36] <gnomefreak> tseng: i used checkinstall and it worked so im assuming make install should have
[02:37] <desrt> mdke; well... i know pam is also supposed to handle motd printing
[02:37] <desrt> mdke; so my bet is that sshd and pam are both doing it
[02:37] <dholbach> gnomefreak: what does the package contain afterwards?
[02:38] <gnomefreak> after compiling it?
[02:38] <dholbach> gnomefreak: yes, what does the checkinstall package contain?
[02:38] <desrt> mdke; you could try commenting out the 'pam_motd.so' line in /etc/pam.d/ssh
[02:38] <dholbach> gnomefreak: just run  dpkg -c  on it
[02:38] <mdke> desrt: if I disable pam in the sshd config, the message only appears once
[02:38] <desrt> mdke; i bet that will get you 1 message
[02:38] <desrt> hah
[02:39] <desrt> so i'm right.  sshd and pam both think that they ought to be printing the message
[02:39] <desrt> neither one is wrong, but ubuntu probably ought to decide which one it wants
[02:39] <tseng> pam is lower in the stack
[02:39] <desrt> btw
[02:39] <desrt> my /etc/ssh/sshd_config has
[02:39] <desrt> PrintMotd no
[02:40] <desrt> did you maybe change that?
[02:40] <gnomefreak> looking for the deb :(
[02:40] <dholbach> gnomefreak: oh, i thought checkinstall would provide your with a deb
[02:40] <dholbach> gnomefreak: i never used it
[02:40] <gnomefreak> it does
[02:40] <gnomefreak> i cant remember where it is
[02:40] <mdke> desrt: it is commented out, and says "Yes" (default?)
[02:40] <gnomefreak> /usr/bin im sure
[02:40] <dholbach> gnomefreak: the point is, as long as http://cvs.gnome.org/viewcvs/gimmie/Makefile.am?view=markup doesn't have anything but the funny message in install: we can't package ir properly
[02:41] <desrt> mdke; *shrug*
[02:41] <mdke> ok, I'll file it
[02:41] <desrt> mdke; if you really didn't modify the file yourself then this is the source of the bug
[02:41] <desrt> mdke; but my file on edgy is fine
[02:41] <mdke> ok, dapper here
[02:41] <desrt> it's fine on dapper for me too
[02:41] <desrt> did you run prerelease dapper on this box?
[02:42] <mdke> no
[02:42] <desrt> hm
[02:42] <gnomefreak> ok dholbach ill look into it and let you know
[02:42] <dholbach> gnomefreak: gracias
[02:42] <desrt> i don't think it's worth filing.  just fix it on your machine.
[02:42] <mdke> yes, you are probably right
[02:43] <desrt> i just checked a couple of my machines that have been upgraded.  one from breezy one all the way from hoary.  no prob on anything.
[02:44] <mdke> fine, I'll leave it, thanks for your help
[02:51] <Fujitsu> I like it!
[02:52] <pitti> infinity: kind of a negative reward? :)
[02:52] <infinity> pitti: Yeah, perhaps mvo commenting on some of the weirder strings he's written into update-manager over the years. :)
[02:53] <gnomefreak> dholbach: heres the output of dpkg -c file.deb http://paste.ubuntu-nl.org/20066
[02:53] <gnomefreak> Hobbsee: i got ff to work in kde finaly ;)
[02:53] <desrt> it's an interesting concept
[02:53] <Riddell> gnomefreak: apt-get install firefox?
[02:53] <dholbach> gnomefreak: it means the package is empty
[02:53] <Hobbsee> gnomefreak: yay!
[02:53] <gnomefreak> Riddell: no it was the font thing
[02:53] <dholbach> gnomefreak: which is because of the missing install target
[02:53] <desrt> imagine a launchpad group for "www.penis-enlargement.com lovers"
[02:53] <mvo> infinity: I worked hard to become a member for this team
[02:53] <desrt> and then the admin adds everyone on launchpad to it
[02:53] <gnomefreak> dholbach: oh ok ty :(
[02:54] <mvo> infinity: ... very hard!
[02:54] <desrt> groupspam
[02:54] <pitti> mvo: what do you have to screw up to become a member? :)
[02:54] <dholbach> infinity: there's jkakar's and jbailey's team as well :)
[02:55] <mvo> pitti: the name says it all :P 
[02:55] <mvo> dholbach: URL?
[02:55] <pitti> mvo: n-d makes me bite into my table
[02:55] <dholbach> urg, do i have to search it again :)
[02:55] <infinity> dholbach: Hah. :)
[02:55] <ajmitch> dholbach: https://launchpad.net/people/intergalacticbloodshed-hackers
[02:55] <sabdfl> elmo: do we ever want to allow mixed-arch uploads? an upload of a source with builds in different arch's
[02:56] <dholbach> ajmitch: mvo :)
[02:56] <ajmitch> mvo: ^^ :)
[02:56] <infinity> sabdfl: We don't want that for Ubuntu, but other hosted projects might.
[02:56] <sabdfl> infinity: good point. thanks
[02:56] <infinity> sabdfl: I do it in Debian all the time.  ALL the time.
[02:56] <mvo> dholbach: haha
[02:56] <mvo> pitti: notification-daemon?
[02:56] <infinity> sabdfl: Oh, and we do it with security right now too (source+all+allarches), but that's special-cased.
[02:57] <pitti> mvo: yes
[02:57] <mvo> pitti: what are you doing with it?
[02:57] <pitti> mvo: well, scary things
[02:57] <mvo> *ick*
[02:57] <pitti> mvo: Imagine, I display a notification!
[02:57] <mvo> pitti: DON'T DO THIS
[02:57] <pitti> argh, I *knew*
[02:57] <mvo> pitti: seriously, what is it doing wrong ?
[02:58] <pitti> mvo: it misplaces the notification bubble so that half of it is invisible (below the bottom screen edge) and then it just crashes
[02:58] <pitti> mvo: with apport installed, I started to notice how often it actually crashes :/
[02:58] <mvo> :(
[02:58] <pitti> ok, that's something for post feature freeze
[02:59] <mvo> pitti: I will try to have a look soonish, I need to update my tree to reflect the current merge anyway
[02:59] <pitti> mvo: oh, I just whined, don't feel blamed :)
[02:59] <mvo> pitti: thats all right
[03:20] <kagou> hi
[03:53] <Hobbsee> dholbach: three times?
[03:54] <dholbach> Hobbsee: jealous?
[03:54] <tseng> 3.times do { Hobbsee.Hug }
[03:54] <Hobbsee> dholbach: maybe.  i like hugs.  and i'm bugfixing :P
[03:54] <Hobbsee> heh
[03:55] <dholbach> :-)
[04:14] <Hobbsee> infinity: are we gettign a mass giveback on the buildds for edgy?  seeing as a whole lot of stuff failed to build while xlibs/xfonts/whatever it was was broken? 
[04:14] <Hobbsee> i'm particularly interested in diacanvas2
[04:15] <infinity> Hobbsee: There's one in progress right now.
[04:16] <Hobbsee> infinity: okay, cool :)  any ETA on when it's done?
[04:17] <infinity> Hobbsee: They take vaguely a day with the current number of failed builds.
[04:17] <Hobbsee> infinity: okay, cool :)  thanks
[04:35] <Chipzz> are there any docs on creating your own usplash boot picture?
[04:36] <Hobbsee> !usplash
[04:37] <Hobbsee> Chipzz: yes.  in a channel with ubotu in it
[04:37] <Chipzz> Hobbsee: no such channel?
[04:38] <Hobbsee> Chipz[00:38]  <ubotu> usplash is the start-up splash (before GNOME/KDE appears) in Ubuntu. To customize it, see https://wiki.ubuntu.com/USplashCustomizationHowto
[04:38] <Chipzz> Hobbsee: ah, the nick ubotu :)
[04:38] <Hobbsee> yes
[04:39] <Chipzz> Hobbsee: I thought #ubotu :P
[04:40] <Hobbsee> Chipzz: heh.  siilly :P
[04:41] <Hobbsee> iwj: what's it doing now?
[04:41] <Hobbsee> iwj: apart from being horrid, like not actually finishing building if you try with a qt engine, not gtk
[04:43] <iwj> The configure script is on crack.
[04:43] <thom> iwj: yes. but which bit in particular this time?
[04:43] <iwj> It's too complicated to explain :-).
[04:43] <iwj> The upshot is that if you build with pango and ft2 it doesn't link.
[04:43] <iwj> Presumably it works if you turn on xft but we don't.
[04:44] <thom> ew
[04:45] <Hobbsee> ah great
[04:47] <infinity> iwj: I vaguely recall hacking around that in thunderbird.
[04:47] <infinity> iwj: Check the hideous patches in the thunderbird source, and see if one is helpful.
[04:52] <ogra> seb128, ping
[04:53] <seb128> ogra: pong
[04:54] <ogra> seb128, i'm nearly done with ltsp local device support, my scripts link a new device to the users desktop which works just fine, i also add the link to .gtk-bookmarks, now i'd like to use device specific icons, where can i add the info for nautilus ? 
[04:54] <ogra> just create a file:XXX entry in .nautilus/metafiles ? 
[04:55] <ogra> (the script knows if it a camera, cdrom or harddisk/usbstick)
[04:55] <ogra> *its
[04:55] <seb128> what is that script supposed to do?
[04:56] <seb128> modifying .gtk-bookmarks and nautilus private data looks like hackish and wrong
[04:57] <ogra> well, thats the way it works ... in ltsp ... 
[04:57] <ogra> upstream as well as for us atm
[04:57] <seb128> doesn't mean it's right ;)
[04:57] <seb128> what are you trying to get listed?
[04:58] <ogra> i got it listed already, i just want to know how to make it not use the folder icon :)
[04:58] <seb128> what got listed already
[04:58] <ogra> ltspfs mounts the device from the thin client in /tmp/.$UID-ltspfs/<device>
[04:58] <seb128> I'm trying to suggest a better solution to what you are doing
[04:58] <ogra> my script links that to the desktop
[04:58] <seb128> but I need to know what you try to get done for that
[04:59] <ogra> and adds that link dynamically to .gtk-bookmarks
[04:59] <seb128> that looks wrong to me
[04:59] <ogra> that gives me an icon on the desktop and in the bookmarks list
[04:59] <seb128> those should be listed as drives
[04:59] <ogra> they cant
[04:59] <seb128> like we do for other mounts, devices, etc
[04:59] <seb128> why ?
[04:59] <ogra> because its a fuse mount
[05:00] <ogra> as the name says, filesystem in userspace
[05:00] <seb128> can't you just use a .desktop?
[05:00] <ogra> i cant do anything that needs root rights
[05:00] <ogra> does openoffice know about that ? 
[05:00] <seb128> listing devices doesn't require root rights
[05:00] <ogra> and doe it get added to .gtk-bookmarks ? 
[05:00] <ogra> *does
[05:00] <seb128> no, it gets added to the devices list
[05:01] <seb128> like if you mount an another partition
[05:01] <seb128> try mounting a windows partitions
[05:01] <ogra> well, thats the same from a user POV :)
[05:01] <ogra> hmm, i didnt know thats possible through just a .desktop file
[05:02] <seb128> ogra: right click, on the desktop, pick "create launcher"
[05:02] <ogra> so back to my question, will openoffice know to handle it ? thats the most important part for ltsp ... at least for edubuntu
[05:02] <seb128> hum, no, not a .desktop then
[05:03] <seb128> but to me it looks like they should be handled like a standard mount
[05:03] <seb128> if you mount a partition to /media/windows it get listed to the places
[05:03] <ogra> hmm, and if i want to create a device .desktop i cant use a path as arguent :)
[05:03] <seb128> why shouldn't be the case for those too?
[05:03] <seb128> ogra: you can
[05:03] <seb128> Type=Link
[05:03] <ogra> because i have no access to /media
[05:03] <seb128> URL=/path
[05:03] <ogra> ah, cool
[05:03] <ogra> ok
[05:04] <seb128> ogra: /media, or /whatever, same thing
[05:04] <seb128> a mount is a mount
[05:04] <seb128> the mountpoint doesn't really matter
[05:04] <ogra> it is no mount :) 
[05:04] <ogra> its a bit complicated to explain :)
[05:04] <ogra> there is a permanent mounted fuse mount ...
[05:04] <ogra> devices i plug in on the client get mounted *below* that
[05:05] <seb128> how is that different of a windows partition mounted from fstab at boot?
[05:05] <seb128> ah
[05:05] <ogra> in the session there is a script that monitors that dir on the server
[05:05] <ogra> if a new device shows up it shuld do xyz
[05:05] <ogra> so the user logged in on the client gets access to the device
[05:06] <ogra> currently i just link the new device t the desktop ...
[05:06] <ogra> i'll try the .desktop variant 
[05:06] <ogra> thans for now
[05:06] <seb128> np
[05:06] <ogra> :)
[05:07] <iwj> infinity: I see what you mean about hacky.  Thanks.
[05:08] <infinity> iwj: Yeah, it wasn't elegant, I was rushed and never cared to look for a saner way.
[05:12] <iwj> infinity: Well, I'm starting to lose patience with it, so I'll copy your `fix' anyway :-).  It has an least reasonable chance of working.
[05:13] <ogra> iwj, i found an easy way to make flash sound work for my ltsp users
[05:13] <infinity> ogra: Install Windows and IE?
[05:14] <ogra> apparently it always worked, but flash itself has a hardcoded test if /tmp/.esd/socket exists even though it doesnt use it at all :)
[05:14] <ogra> if it doesnt find it, it will just switch off soundoutput internally
[05:15] <infinity> ogra: Err, I don't have that here, and flash works.
[05:15] <ogra> (we dont have the /tmp/.esd dir at all in ubuntu, esd creates it as /tmp/.esd-$UID)
[05:15] <infinity> ogra: Or does it also check for /tmp/.esd-${UID}/socket ?
[05:15] <ogra> infinity, you have a ltsp setup ?
[05:15] <carlos> pitti: btw, just in case if someone asks you
[05:15] <infinity> ogra: Well, no, I mean locally.  This is only an issue with remote X?
[05:15] <carlos> pitti: the daily language packs were not being regenerated until yesterday (my fault)
[05:16] <ogra> its an issue with forwarded esd output :)
[05:16] <ogra> (you dont have a /tmp/.esd* on ltsp servers usually ... that exists only on the client)
[05:16] <seb128> carlos: hi. any edgy language pack planned?
[05:16] <carlos> pitti: and atm, the mirror database we use is not being updated daily (it's blocked by some admin work)
[05:16] <sbalneav> Kamion: ping
[05:17] <carlos> seb128: I got an approval for the code changes we had to do
[05:17] <ogra> sbalneav, he's on vacation this week
[05:17] <sbalneav> ah
[05:17] <seb128> carlos: that mean "yes, soon"? :)
[05:17] <carlos> seb128: I need to plan that with Stuart as soon as my changes land in our development repository (today)
[05:17] <sbalneav> Just wanted to chat about https://launchpad.net/distros/ubuntu/+source/openssh/+bug/54180
[05:17] <Ubugtu> Malone bug 54180 in openssh "[rfe]  sshd ought to support 'none' cipher" [Unknown,Confirmed]  
[05:18] <seb128> carlos: during previous cycle you told me that having language packs early should not be an issue for edgy
[05:18] <carlos> seb128: my part of the work is already done (once the merge is done) so it's a matter of schedule some tasks from our admins
[05:18] <ogra> sbalneav, how would that help if you cant use passwords
[05:18] <carlos> seb128: we try now to migrate what we have in dapper before opening edgy
[05:18] <seb128> carlos: having still no translations when GNOME is string frozen is really not nice for Ubuntu and make it hard to use for translators and has an impact 
[05:18] <sbalneav> ogra: see my explanation in #edu
[05:19] <carlos> and it took more time than planned
[05:19] <ogra> sbalneav, in a ltsp context ...
[05:19] <ogra> ok
[05:19] <carlos> seb128: we could open edgy since the first day
[05:19] <carlos> but the idea was to include all changes in dapper 
[05:19] <seb128> carlos: so why does it take 3 months ? :)
[05:20] <carlos> seb128: because Dapper translations are also included so no one needs to start from zero
[05:20] <seb128> carlos: right, but things like evolution which changed its translation domain is translated to 0% for months now, that's not really acceptable
[05:20] <seb128> carlos: who should I speak to to get that sorted for next cycle?
[05:21] <carlos> seb128: 0% ?
[05:21] <seb128> carlos: yeah, domain translation changed and we have no language-pack shipping the new .mo name
[05:21] <seb128> evolution-2.8.mo
[05:21] <seb128> we ship evolution-2.6.mo
[05:21] <carlos> seb128: Jordi or I
[05:21] <carlos> seb128: is it dapper ??
[05:21] <seb128> no, edgy
[05:21] <carlos> seb128: jordi, danilo or I
[05:22] <carlos> seb128: hmmm for edgy, martin
[05:22] <iwj> ogra: So all you need to do is make /tmp/.esd/socket exist and then it will use the esd lib and connect to the right esd socket ?
[05:22] <carlos> until we open rosetta to do translations
[05:22] <seb128> grumpf
[05:22] <seb128> rosetta should be open for translation when the new cycle start
[05:22] <carlos> seb128: edgy + 1 will have all code in place to open it at the same time we open soyuz for it
[05:22] <seb128> not months after
[05:22] <seb128> you already said that during the dapper cycle :/
[05:22] <carlos> seb128: my code is called at the same time soyuz is updated with a new distro
[05:23] <ogra> iwj, exactly
[05:23] <seb128> carlos: ok, let's see how it works, but current situation really sucks for Ubuntu
[05:23] <carlos> seb128: the difference is that as soon as we open edgy all .pot and .po files for main (or most of them) will be imported
[05:24] <ogra> iwj, i thought we could add an if statement that checks if LTSP_CLIENT is set to the top of /usr/bin/firefox anywhere 
[05:24] <carlos> seb128: I know and I try to do my best with every cycle to improve the situation. The fact that edgy has only 4 months is not helping at all
[05:24] <ogra> and just mkdir -p and touch the socket file...
[05:24] <seb128> carlos: k, I was just trying to point that's important that Ubuntu get translated
[05:24] <ogra> (with some error handling indeed)
[05:25] <carlos> seb128: don't worry, I'm completely aware of that
[05:25] <seb128> carlos: so you can bump the "open next distro for translations" early on your TODO or point to whoever should know that's an issue
[05:25] <carlos> seb128: for edgy + 1 ? or edgy?
[05:25] <seb128> edgy is too late now
[05:25] <carlos> for edgy is already on top
[05:25] <carlos> for edgy + 1
[05:26] <seb128> edgy has no translation
[05:26] <carlos> I already told you that the code that copy translations between distros is in place already to be called when soyuz is updated
[05:26] <seb128> after half of the cycle
[05:26] <carlos> so edgy + 1 will be open with translations directly
[05:26] <seb128> yeah, but weeks are flying and we still have no translation
[05:26] <seb128> GNOME 2.16.0 will soon be here and still no translation
[05:26] <seb128> and that starts being a real issue for users, so I'm point it again
[05:26] <seb128> ok, cool
[05:27] <seb128> s/point/pointing
[05:29] <ogra> iwj, http://wiki.ltsp.org/twiki/bin/view/Ltsp/Sound#Macromedia_Flash
[05:29] <ogra> seems edubuntu users following that had success ... (the libesd part isnt needed)
[05:31] <iwj> ogra: Why not just create this socket in your ltsp server package ?  Just `if doesn't exist, create it', during boot.
[05:31] <ogra> i have no ltsp specific intscrits on the server yet
[05:31] <ogra> *initscripts
[05:32] <iwj> You definitely can't do it in /usr/bin/firefox because it has to be done as root.
[05:32] <ogra> ah, damned, missed that part ... indeed
[05:32] <ogra> ok, i'll think about alternatives :)
[05:32] <ogra> thanks for the pointer
[05:33] <mcdonaldsguy> I'm trying to track down what I think is a bug with the dapper installer, but I'm totally unfamiliar with the debian-installer code... at what point does /target/etc/apt/apt.conf get created?
[05:33] <mcdonaldsguy> sorry if this isn't the appropriate place to ask...
[05:47] <mdz> mcdonaldsguy: apt-setup, I believe
[05:48] <ogra> seb128, hmm, Type=FSDevice together with URL for the target dir seems to work ... (looks a bit weird though)
[05:48] <seb128> ogra: cool ;)
[05:49] <ogra> but i dont get it as a device in the device list (indeed because it isnt a device) can i trick gnome anyhow to see it as device ?
[05:50] <seb128> ogra: I don't know offhand but I'll have a look to that later
[05:50] <ogra> thanks ! :)
[06:02] <infinity> mcdonaldsguy: There should be laws against nicks that make me crave quarter-pounders at 2am, when I can't do anything about it.
[06:05] <LaserJock> :-)
[06:05] <HrdwrBoB> infinity: no 24h mcdonalds?
[06:06] <thom> infinity: there should laws against craving mcdonalds
[06:06] <HrdwrBoB> actually if it's 2am, you're in au eastern states
[06:06] <HrdwrBoB> in which case there is 24hr mcdonalds
[06:06] <infinity> HrdwrBoB: Lots of 24hr McDonald's, none in my neighborhood, and no 24h trains.
[06:06] <rottingmcdonalds> better?
[06:08] <infinity> HrdwrBoB: If you've got a car, you're welcome to show your core-dev appreciation, by grabbing me some.
[06:08] <infinity> HrdwrBoB: (I'm in Armadale... See you in a few)
[06:08] <HrdwrBoB> heh
[06:09] <HrdwrBoB> you explain it to my wife
[06:09] <mdz> pitti: is something supposed to happen when I click on the crash notification icon?
[06:09] <infinity> Happily.
[06:10] <mdz> HrdwrBoB: pick up a gift for her on the way home
[06:10] <infinity> Perhaps a nice apple pie.
[06:10] <HrdwrBoB> melbourne actually has lots (relatively) of 24h florists
[06:10] <HrdwrBoB> I imagine for men who are coming home late
[06:11] <mvo> mdz: does anything happens when you run "apport-gtk" ?
[06:11] <mdz> pitti: also, why is it that the process will be in uninterruptible sleep when the helper is called by the kernel? can't we arrange for it to be simply stopped?
[06:11] <infinity> Would go well with all the out-til-six drinking that goes on here, yes.
[06:11] <mdz> mvo: command not found
[06:11] <mdz> don't have it installed
[06:11] <mvo> mdz: oh, sorry. make that /usr/share/apport/apport-gtk 
[06:12] <mdz> mvo: I don't have the apport-gtk package installed
[06:12] <mdz> apport doesn't depend on it
[06:12] <mvo> ok, I guess the icon should only show if apport-gtk is actually installed :)
[06:13] <mdz> mvo: agreed
[06:13] <mdz> I'll file a bug
[06:15] <mvo> mdz: thanks
[06:15] <mdz> mvo: bug 55971
[06:15] <mdz> bug 55791 that is
[06:15] <Ubugtu> Malone bug 55791 in apport "Can display non-functional notification icon" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/55791
[06:16] <mdz> mvo: another strange thing, when I hover over the reboot notification icon, it says "crashreport detected"
[06:16] <mdz> right now I have both icons, and both say "Crashreport detected".  clicking one displays the reboot dialog, clicking the other does nothing
[06:16] <mdz> and they look much too alike :-)
[06:17] <mdz> seb128_: gnome-panel-screenshot just crashed on me, and bug-buddy doesn't seem able to get a backtrace
[06:18] <mvo> mdz: do you have the latest u-n? I uploaded a new version that has different icons and fixes the tooltip problem
[06:18] <mdz> /usr/bin/bug-buddy --appname=gnome-panel-screenshot --pid=1 --package-ver=(null)
[06:18] <mdz> that doesn't look right...--pid=1?
[06:18] <seb128_> nop, it doesn't
[06:18] <mdz> mvo: perhaps I do but haven't logged out
[06:18] <seb128_> and (null) doesn't look like right neither ;)
[06:19] <mvo> mdz: right, that does explain it
[06:19] <mdz> bug-buddy also says the app is unknown, dunno if that's normal for g-p-s
[06:19] <seb128_> that means that the .desktop is broken
[06:19] <seb128_> new bug-buddy rely on the .desktop infos being correct
[06:20] <seb128_> having a look
[06:20] <seb128_> gnome-panel-screenshot doesn't crash on my box ... is that reproducible on your box?
[06:20] <mdz> mvo: ok, restarted update-notifier and the icon is correct now
[06:21] <mdz> my reboot icon disappeared though
[06:21] <seb128_> what version of gnome-utils do you have?
[06:21] <mdz> Version: 2.15.91-0ubuntu1
[06:21] <seb128_> could you try with .92? :)
[06:21] <mdz> yep, upgrading
[06:26] <Burgwork> mdz, do you have a time for Knot 2? IE:, should I push myself to finish Knot2 on the wiki tonight/next day?
[06:27] <mdz> Burgwork: I believe Kamion is rolling knot 2, and he is on holiday until next week
[06:27] <mvo> mdz: I will look into if/why the reboot icon might have disappeared
[06:27] <Burgwork> mdz, ok, thanks for the info
[07:01] <bluefoxicy> wow.  In Evolution there's an s-expression library with a 'struct _glib_sux_donkeys { ... }' in it
[07:01] <sabdfl> bluefoxicy: truth is not always an absolute defence, it seems
[07:02] <bluefoxicy> sabdfl:  it's just immature.
[07:25] <pitti> mdz: crash icon> update-notifier needs a small update, already settled with mvo
[07:26] <pitti> mdz: kernel sleep> tricky, I discussed this with BenC for a while; I think the new approach is more robust
[07:27] <mdz> pitti: looks like update-notifier is already fixed, I only needed to restart it
[07:27] <mdz> pitti: I'd be interested to read the discussion if you have a copy around
[07:29] <pitti> mdz: oh, it was in German :/
[07:29] <pitti> mdz: short story is, instead of checking /var/crash/* on its own, it needs to call /usr/share/apport/apport-checkreports
[07:29] <mdz> pitti: BenC speaks german?
[07:29] <pitti> mdz: oh, I thought the one with mvo, sorry ;)
[07:31] <BenC> lol
[07:32] <pitti> BenC: from my experiments it seemed to be impossible to do anything with the crashed process while the kernel was active on it; is there another option?
[07:33] <BenC> pitti: what else do you need to do?
[07:34] <pitti> BenC: well, I am fine with the 'dump core, let it die, and examine the core' approach; I'm refering to mdz's question why the crashed process cannot be stopped instead of being in kernel sleep
[07:34] <mvo> pitti: is the script ready? 
[07:35] <pitti> mvo: yes, since yesterday
[07:35] <mdz> BenC: he wants to ptrace it
[07:35] <BenC> the process has to be stopped in order for the kernel to process the core dump without the process dieing
[07:35] <mdz> stopped is fine; stopped processes can be ptraced
[07:35] <dmg> mdz: you could always learn german: http://fsi-language-courses.com/GermanBasic.aspx
[07:35] <pitti> mdz: besides, you currently cannot gdb a stopped process either
[07:35] <pitti> mdz: but this is a gdb limitation AFAIUI
[07:36] <mvo> pitti: perfect, I will add it then
[07:36] <BenC> mdz: How do you ptrace something that is just going to be reaped?
[07:36] <BenC> it's not going to do anything else other than die and gets its memory reclaimed
[07:37] <mdz> BenC: but before it does, we'd want to examine its memory
[07:37] <mdz> and registers
[07:37] <pitti> mdz: but we can get all of that from the core dump
[07:37] <BenC> mdz: which the core dump can provides, right?
[07:38] <dmg> as long as the core dump isn't being put on a read-only file system or somewhere that might run out of space.
[07:38] <BenC> dmg: it's being done the same way it would be done if you set ulimit=100M
[07:39] <mdz> pitti,BenC: sure, we can get the same info either way.  the core dump seems like an extra step
[07:39] <BenC> so if it would normally be allowed to core dump, then it would be the case here as well, and if not, then apport will just deal with what little info it can get (process name, signal, etc)
[07:39] <mdz> and core dumps have historically not always worked correctly, e.g. in threaded programs
[07:39] <mdz> is that fixed now?
[07:40] <mdz> it also seems that this would require that we enable core dumps by default in order to get any useful information
[07:40] <BenC> mdz: Current setup is going to be like this:
[07:40] <mdz> and if they're enabled by default, and apport isn't there to clean up, they'll create cruft
[07:40] <BenC> /proc/sys/kernel/crash-helper will contain the name of a program
[07:40] <BenC> the kernel checks if that helper exists and is excutable
[07:41] <BenC> if it is, it creates the dump regardless
[07:41] <BenC> calls the helper
[07:41] <BenC> if the core dump was created just for the helper (e.g. ulimits has core=0), then it removes the core
[07:42] <BenC> pitti: One thing I haven't thought about is serialization of the helper program
[07:43] <pitti> BenC: serialization?
[07:43] <BenC> pitti: would you prefer that the kernel just call it whenever, or should it only allow one helper to be active at a time?
[07:43] <pitti> (btw, I have to leave in two minutes for Taekwondo, sorry)
[07:43] <BenC> I can create a lock so that it wont call another helper until the previous one is done
[07:43] <pitti> BenC: apart from the multiple I/O load it will create, I don't see a problem with having several parallel instances
[07:44] <BenC> ok
[07:44] <BenC> just worried about a flooding caused by some gnome program failing repeatedly
[07:44] <pitti> it currently does not do anything that would need mutual exclusion
[07:44] <pitti> BenC: well, apport only creates one report per program and user
[07:44] <pitti> BenC: if there already is an unread report, it immediately exists
[07:44] <pitti> s/exists/exits/
[07:45] <BenC> gotcha
[07:45] <pitti> mdz: that's why I wanted to avoid setting the standard ulimit (avoid disk usage DoS)
[07:46] <mdz> pitti: ah, ok.  it sounds like you've addressed that issue already
[07:46] <pitti> mdz: OTOH apport shouldn't unconditionally remove core dumps, the admin will usually want them if he enabled ulimits manually
[07:46] <mdz> pitti: it should move them into /var/crash and expire them, or similar
[07:46] <pitti> ok, if there are any further issues, please mail me; sorry, I have to leave now
[07:46] <pitti> mdz: that's roughly what happens now
[07:46] <pitti> mdz: they are bzip2'ed and integrated into the crash report with the other data
[07:47] <pitti> mdz: and a cronjob cleans up old ones
[07:47] <mdz> pitti: good night
[07:47] <pitti> mdz: bzip2 makes a *huge* difference (160 MB firefox dump uncompressed -> 5 MB bzip2)
[07:47] <mdz> pitti: yes, I expected it would
[07:48] <sladen> pitti: how does gzip compare?
[07:48] <pitti> sladen: didn't try
[07:48] <mdz> probably almost as good
[07:48] <pitti> yes, the dump is full of zeros
[07:48] <mdz> probably full pages of zeros; I wonder why they're not done sparsely
[07:48] <pitti> ok, cu!
[07:48] <mdz> BenC: sparse core dumps?
[07:49] <sladen> pitti: in that case (uninitalised RAM) then, gzip might be preferable, it's stream-based and the processing overhead at the server end will be about 1% of the bzip2 one
[07:50] <mdz> sladen: you mean initialised RAM
[07:50] <mdz> I'm not too concerned about the overhead on the server side; they'll be batch processed
[07:50] <mdz> and if it saves the user even a small amount on the upload, that can be valuable
[07:50] <mdz> smaller upload -> less time spent reporting the problem -> more problem reports
[07:51] <dmg> the problem with sending core dumps is that they might contain sensitive information and there's no way to reliably scrub them.
[07:51] <mdz> dmg: hence we need to ask the user's permission
[07:53] <Nafallo> mjg59: ping
[07:54] <sladen> Nafallo: just ask
[07:54] <Nafallo> oki :-)
[07:55] <Nafallo> did anyone see the latest reply on bug #32963 ? maybe this is worth a new upstream version?
[07:55] <Ubugtu> Malone bug 32963 in xserver-xorg-driver-i810 "Xv movies on 810/i945 gives horrible color, Gamma" [Unknown,Needs info]  http://launchpad.net/bugs/32963
[07:55] <Nafallo> if someone make the package I'm happy to test it here :-)
[07:55] <Nafallo> (on edgy i386)
[07:59] <Nafallo> sladen: ah, you are the assigne :-)
[08:00] <bddebian> Heya
[08:00] <rodarvus> Nafallo, our edgy package of i810 is basically the latest version available
[08:00] <bit_doidao> hello! how to compile php5 with imap support? its possible to have it by default in php5-ubuntu.deb?
[08:00] <rodarvus> (only two small changes from the latest available version from edgy i810 package)
[08:01] <Nafallo> rodarvus: 1.6.1 in edgy, 1.6.3 is latest upstream, and the one that hopefully fixes the bug :-).
[08:01] <rodarvus> Nafallo, read what I just said :)
[08:01] <rodarvus> there are only two small changes from 1.6.1 to 1.6.3
[08:01] <Nafallo> yea, saw it afterwards :-).
[08:02] <sladen> Nafallo: I think after I bitched at the Intel people in paris they got the people in China to go and send something upstream---not that they bothered sending a copy back to me
[08:02] <mjg59> Nafallo: Hi
[08:02] <cr3> how can I determine which driver corresponds to a file under /dev, like /dev/ptmx for example
[08:02] <mjg59> Nafallo: Ah, already dealt with
[08:03] <Nafallo> mjg59: hello. oh? there is a new package to test somewhere? :-)
[08:03] <sladen> cr3: that's the pseudo terminal multiplexer...
[08:03] <Nafallo> sladen: how... kind of them not to :-P.
[08:03] <sladen> cr3: you probably want ---> #ubuntu
[08:03] <mjg59> Nafallo: Oh, no
[08:04] <mjg59> Nafallo: I'm not anything to do with X at the moment
[08:04] <mjg59> I'll see if I can reproduce somewhere
[08:04] <Nafallo> mjg59: hehe, might be a vice decision ;-)
[08:04] <Nafallo> oki
[08:15] <rodarvus> oh
[08:16] <rodarvus> Nafallo, I just found out one of the two "small changes" I mentioned is actually the merge of the 'mergedfb' tree of i810
[08:16] <rodarvus> so, not really a small change
[08:16] <Nafallo> :-)
[08:16] <rodarvus> Nafallo, I'd be thankful if you could test it, then :)
[08:17] <Nafallo> hmm, when I get the build-deps correct ;-)
[08:17] <rodarvus> I can build it quickly locally, if needed be
[08:17] <Nafallo> any idea where xineramaproto might reside? :-)
[08:18] <rodarvus> x11proto-xinerama-dev
[08:18] <Nafallo> yea. we can see who can make it build first! :-)
[08:18] <Nafallo> thanks :-)
[08:18] <rodarvus> Nafallo, actually I'm ok with you build-testing it
[08:18] <Nafallo> okidoki
[08:19] <rodarvus> btw, is anyone up to testing xorg-server on i386 or amd64?
[08:19] <Nafallo> sure, why not. after I test this one :-)
[08:19] <rodarvus> I got an updated package (with ~ 16 new patches on it)
[08:19] <dholbach> sladen: cr3 works for Canonical - and that question is not really a "#ubuntu support" question - I doubt they'd know over there.
[08:20] <rodarvus> (disclaimer: works fine on two different machines, here)
[08:20] <cr3> dholbach: that's alright, I'll figure it out :)
[08:21] <dholbach> cr3: I'd tell you, if I knew :-)
[08:21] <cr3> dholbach: I'm getting the source and I suspect I could grep for the major and minor numbers of the driver. I'll see once the darn thing finishees downloading :)
[08:22] <cr3> err, numbers of the device in the drivers sources
[08:22] <bbrazil> cr3: /proc/devices
[08:23] <cr3> bbrazil: thanks, that's a nice overview. I was using ls -l which provides that information as well
[08:24] <bbrazil> cr3: it does?
[08:24] <cr3> bbrazil: sure, major and minor device numbers
[08:24] <bbrazil> cr3: ah, I though you were looking for the drivers
[08:25] <Nafallo> rodarvus: that worked better, makes it throu configure okey now ;-)
[08:25] <Nafallo> finished building, brb. restart X :-)
[08:27] <treitter> debhelper isn't very useful of you don't have an autotooled package, right?
[08:28] <LaserJock> debhelper is useful for just about any situation, I'd think
[08:31] <bit_doidao> hello! i want to recompile php5 with only i additional stuff. how to het the original compile parameters for adding this one --with?
[08:34] <LaserJock> bit_doidao: apt-get source and look in <source>/debian/rules
[08:34] <LaserJock> bit_doidao: and #ubuntu-motu might be a better place for that :-)
[08:38] <Nafallo> rodarvus, sladen: I still have the bug :-/
[08:38] <rodarvus> thats sad
[08:39] <rodarvus> Nafallo, do you want to upgrade xorg-server to check if it solves your bug?
[08:39] <Nafallo> rodarvus: sure :-)
[08:39] <rodarvus> I don't think it will, but you'll be helping me test the new xorg-server :P
[08:39] <bbrazil> bit_doidao: phpinfo()
[08:39] <Nafallo> :-)
[08:39] <Riddell> Kamion: what's the state of the point releases?
[08:40] <rodarvus> Nafallo, http://people.ubuntu.com/~rodarvus/packages/xorg-server/
[08:40] <rodarvus> it has packages for i386 and amd64
[08:41] <sladen> Nafallo: were you the person that tried the latest upsteam from xorg?
[08:42] <Nafallo> sladen: yea, I built a package for i810 and rebooted :-)
[08:42] <Burgwork> Riddell, kamion is on vacation this week
[08:44] <bddebian> heh
[08:49] <dholbach> Burgwork, tseng: I'm sure Riddell is perfectly aware of that - Kamion was around today and yesterday to help with the releases and we are all happy he helped out with them.
[08:51] <Burgwork> dholbach, very likely, but it was an odd question
[08:52] <Nafallo> rodarvus: that works, will test video after I made my dinner ready :-9
[08:52] <rodarvus> Nafallo, thanks
[08:52] <rodarvus> are you running i386 or amd64?
[08:53] <Nafallo> i386. my rabbit ate the powercord for my amd64-laptop :-/
[08:53] <Nafallo> so can't test that :-P
[08:53] <bddebian> It's an evil beast with big pointy teeth...
[08:54] <LaserJock> Nafallo: yikes, is the rabbit ok :-)
[08:54] <Nafallo> hehe
[08:54] <Nafallo> yepp
[08:54] <Nafallo> http://www.nafallo.info/blog/ ;-)
[08:57] <Nafallo> bddebian: big and big... http://www.nafallo.info/~nafallo/pics/animals/storm.jpg
[08:57] <bddebian> "Run away, run away..."
[08:58] <Nafallo> ...and the bug still isn't solved.
[09:04] <mdz> Riddell: the point release is on its way to the mirrors, according to Kamion's message some hours ago
[09:05] <mdz> Riddell: will be announced late tonight or early tomorrow, UTC
[09:06] <Riddell> mdz: do you know if that includes kubuntu?  I don't see it on releases.ubuntu.com
[09:07] <mdz> Riddell: no, I don't
[09:08] <elmo> it does, it's still syncing
[09:08] <Riddell> elmo: cool, thanks
[09:18] <seb128> siretart: around?
[09:18] <seb128> siretart: why did you reassign bug #48138 to totem-xine?
[09:18] <Ubugtu> Malone bug 48138 in totem "totem-xine crashes when started with no argument" [Unknown,Fix released]  http://launchpad.net/bugs/48138
[09:18] <seb128> siretart: the commit pointed is a libxine one
[09:24] <crimsun> seb128: he may have just read the "bugzilla.gnome.org" URL and not seen the actual xine-lib portion
[09:25] <seb128> crimsun: yeah, probably, that's why I ping him ;)
[09:25] <crimsun> (source package corrected.)
[09:25] <seb128> crimsun: ok, thank you ;)
[09:26] <elmo> (releases.u.c is now up-to-date and cdimage.u.c is syncing at a sane speed again and will be up-to-date in 5-10 mins)
[10:20] <dilinger> i'm seeing some issues w/ the dapper installer (apt-setup) and apt
[10:20] <dilinger> dapper's apt-setup seems to create /etc/apt/apt.conf w/ Acquire::http::proxy set to "false"
[10:21] <dilinger> in one case (i'm not sure how a coworker managed to get it that way), apt.conf contained 'Acquire::::proxy "false";'
[10:21] <dilinger> somehow, $protocol never made it there
[10:22] <dilinger> in the rest of the cases, apt.conf is ending up w/ 'Acquire::http::proxy "false";', which does not work with out apt proxy (approx, apt-cacher)
[10:22] <dilinger> what i end up getting is this when i apt-get update:
[10:22] <dilinger> Failed to fetch http://zanzibar.ne.in.athenacr.com:9999/ubuntu/dists/dapper/main/binary-i386/Packages.gz  403 Forbidden
[10:23] <dilinger> however, i can wget that file; and if i comment out that Acquire::http::proxy line or set it to "" (instead of "false"), it works
[10:24] <dilinger> is this somehow related to gpg checking, or a genuine apt bug?
[10:47] <dilinger> ok, well: https://launchpad.net/distros/ubuntu/+source/apt/+bug/55813
[10:47] <Ubugtu> Malone bug 55813 in apt "Acquire::http::proxy breaks things with apt proxies like apt-cacher" [Untriaged,Unconfirmed]  
[10:47] <dilinger> heh, what timing
[11:05] <pitti> Hello again
[11:05] <dmg> pitti: have a good time beating people up?
[11:05] <dmg> or throwing them, I suppose.
[11:06] <pitti> mdz: re apport, having the kernel create a core dump is not really an extra step; we want the core dump anyway to be able to create good backtraces later; it's just that creating the core from apport itself would allow greater flexibility
[11:06] <mdz> pitti: ok
[11:06] <mdz> pitti: I'm eager to experiment with it
[11:08] <pitti> dmg: hey, I'm peaceful :)
[11:08] <pitti> mdz: heh, it seems archive.u.c. caught up a little and now has 0.8
[11:09] <pitti> mdz: I uploaded 0.9 some 6 hours ago, so it can't take long any more until it reaches the archive
[11:09] <pitti> mdz: with 0.9 I'm reasonably confident of not breaking too much; thus I'd like to add it to ubuntu-desktop for greater testing coverage soon
[11:09] <pitti> mdz: would you be fine with that?
[11:09] <mdz> pitti: yes
[11:10] <dholbach> in fact i didn't see any package updates in the last 6 hours
[11:11] <pitti> dholbach: hey, it took over a day for 0.8 to appear, so I won't complain yet :/
[11:12] <pitti> mdz: in 0.9 I also added a library (and some instructions how to use it) which 'fakes' the future kernel behaviour, so that you can get full gdb love
[11:12] <mdz> dholbach: really?  where is the blockage?
[11:12] <pitti> dholbach: yesterday it was due to the network link switch
[11:12] <mdz> pitti: what's the future kernel behaviour?
[11:12] <pitti> mdz: security.u.c. was moved yesterday, archive.u.c today
[11:13] <dholbach> mdz: i can't tell, it's just what i noticed
[11:13] <pitti> mdz: now it just calls apport while keeping the crashed process in limbo; BenC works on the solution we discussed some hours ago, the kernel dumps core and *then* calls apport with the core dump file as argument
[11:14] <pitti> mdz: and it will introduce another 'temporary' core size ulimit (if the normal ulimit is broken, it'll remove the core dump after apport finished)
[11:16] <BenC> edgy kernel upload tonight with ABI bump (you've been warned, because basically there's nothing you can do to stop me!!! :)
[11:16] <mdz> pitti: oh, I see
[11:16] <pitti> go, BenC, go! :)
[11:16] <BenC> pitti: cleaning up your apport stuff is my final task before upload
[11:17] <pitti> BenC: that might come in time for tomorrow's distro team meeting :)
[11:17] <BenC> I wont lie, I finished it up so my task list appeared as if I did something useful this past week :)
[11:18] <BenC> "it's all about the bullets baby"
[11:18] <pitti> ^ from which movie does that come?
[11:19] <BenC> the song, "It's all about the Benjamin's baby"
[11:19] <BenC> or the weird al version, "It's all about the Pentium's baby"
[11:20] <pitti> mdz: ok, then I'll wait with the main promotion/ubuntu-desktop addition until the kernel is in place, then we don't need the hideous library any more to get the full fun
[11:47] <pitti> Good night everyone
[11:47] <slomo> gn8 pitti 
[11:56] <dholbach> night guys
[11:57] <slomo> gn8 dholbach :) and enjoy your vacation
[11:57] <dholbach> gracias slomo! have a good time too!