[03:46] <pygi> morning Hobbsee 
[03:47] <Hobbsee> hiya pygi 
[03:47] <voltagex> I think I'm in the wrong channel, I'd like to discuss a bug report with someonw.
[03:47] <pygi> which bug?
[03:47] <voltagex> https://bugs.launchpad.net/ubuntu/+source/vlc/+bug/103025
[03:48] <ubotu> Launchpad bug 103025 in vlc "(Feisty) VLC sound quality is poor for many video files" [Medium,Invalid]  
[03:48] <pygi> no idea about vlc, sorry
[03:48] <voltagex> I don't believe it's a VLC issue
[03:48] <pygi> hey LaserJock 
[03:48] <voltagex> and from what other people in that discussion have said, it's an alsa problem
[03:49] <voltagex> but personally I think it needs investigating, because there's a lot of confusion in that thread.
[03:49] <pygi> if it's really alsa problem, then crimsun is your guy, but he's probably sleeping now
[03:49] <voltagex> which timezone?
[03:49] <voltagex> sorry, which timezone is crimsun in?
[03:49] <LaserJock> hi pygi 
[03:50] <pygi> voltagex, no idea
[03:50] <voltagex> (so I know when to look for him :) )
[03:50] <pygi> LaserJock, how is it going at this early morning? :)
[03:50] <LaserJock> pygi: it's 6:50 pm here
[03:50] <pygi> voltagex, ah, he's already following the bug
[03:50] <voltagex> 11:50 -!- crimsun is away: moving.  Offline until sometime Monday
[03:51] <pygi> LaserJock, 3:50 :)
[03:51] <pygi> 3:50AM :)
[03:51] <voltagex> 1150AM here.
[03:51] <pygi> voltagex, he's looking at the bug, don't worry
[03:51] <voltagex> I'd like to get more involved in bug reporting, because I seem to keep stumbling onto them.
[03:52] <voltagex> pygi: what would be the best way to help him?
[03:52] <pygi> voltagex, reproduce bug, attach traces, and such :)
[03:52] <minghua> voltagex: for help reporting bugs in general or alsa bugs in specific?
[03:53] <voltagex> ah...complex
[03:53] <voltagex> this bug in particular, because I have some free time now to chase it down.
[03:54] <minghua> then just put whatever you found in the bug report, since crimsun is already following the report
[03:55] <pygi> nod :)
[03:56] <voltagex> crap, just realised what I wrote is very unhelpful - I made a post before I came in here.
[03:58] <voltagex> so, reproduce this bug by playing a "crackling file", then copy some kind of verbose output to the report? Stack traces are beyond my current skill level, but if someone could point me in the right direction I'm willing to learn.
[05:18] <Hobbsee> siretart: er, are you talking about the fact that libxine-ffmpeg and plugins and whatnot depend on libxine1, or that libxine1 depends on itself?
[05:21] <Hobbsee> looks like a circular dependancy between libxine1-ffmpeg and libxine1 too, where recommends are installed by default.
[05:21] <Hobbsee> maybe i'm on crack or something
[05:21] <pygi> Hobbsee, possible :)
[07:14] <pitti> Good morning
[07:14] <Hobbsee> morning pitti!
[07:14] <pitti> it's a Hobbsee! *hug*
[07:14] <Hobbsee> pitti: it is!  *hugs back* :)
[07:15] <pitti> hi StevenK 
[07:15] <StevenK> Turns, you say?
[07:16] <Hobbsee> pitti: can you reject the first ubuntu-restricted-extras v7 please?
[07:16] <Hobbsee> StevenK: surely i'm not *always* violent...
[07:18] <minghua> Hobbsee: so, emm, resumes? :-P
[07:18] <Hobbsee> minghua: sorry?
[07:19] <Hobbsee> minghua: oh, resumes being violent?
[07:19] <minghua> Hobbsee: yes
[07:20] <StevenK> pitti: Regenerate cruft when and if you feel so inclined.
[07:20] <StevenK> Please.
[07:20] <fabbione> morning guys
[07:21] <Hobbsee> morning fabbione!
[07:21] <fabbione> hey Hobbsee 
[07:22] <ion_> Hobbsee already did that to me. http://heh.fi/tmp/cephalopod
[07:22] <nixternal> hey, I can't get upstream (and I have asked very nicely) to incorporate the license into his tarball. he does post the license on his website though (LGPL), this is for an icon package. Is there any work around what so ever, to get this in and legal?
[07:22] <Burgundavia> nixternal: which package is it?
[07:22] <ion_> nixternal: I think youll only have to recreate orig.tar.gz, adding COPYING.
[07:22] <pitti> Hobbsee: sorry, nothing in NEW or accepted; I'm afraid you have to upload a new version
[07:23] <ion_> nixternal: I could be wrong, though. :-)
[07:23] <nixternal> Burgundavia: icon package for kde
[07:25] <Hobbsee> pitti: okay, will do
[07:26] <Hobbsee> ion_: hehe, yummy
[07:26] <nixternal> I would love it if I could rerecreate the orig.tar.gz with the license file..this icon package is the yummiest out there :)
[07:27] <ion_> nixternal: Id like to see a screenshot, btw.
[07:29] <nixternal> http://www.kde-look.org/content/show.php/Crystal+Project?content=60475
[07:29] <nixternal> the screenshot on that page doesn't do it justice..the link to the homepage has more
[07:29] <ion_> Thanks
[07:30] <Hobbsee> fabbione: just what i wanted.  to be covered in ink...
[07:30] <nixternal> http://www.everaldo.com/crystal/?action=preview
[07:30] <nixternal> there you go
[07:30] <fabbione> :)
[07:31] <ion_> The Home icon reminds me a lot of the equivalent icon in the good old Amiga icon set, whatever its name was.
[07:34] <pitti> StevenK: it finally finished, updated
[07:35] <StevenK> pitti: Thanks
[07:35] <pitti> siretart: can I remind you about the ffmpeg soname transition?
[07:37] <StevenK> pitti: spandsp is waiting on a new release of asterisk-plugins-spandsp to hit unstable, I'm not really comfortable poaching an unreleased version from pkg-voip's SVN.
[07:38] <pitti> StevenK: that's fine; but a simple rebuild won't do?
[07:38] <StevenK> pitti: No, the current package can't build against Asterisk 1.4
[07:39] <ajmitch> hi pitti 
[07:40] <Hobbsee> ah well, closed a couple of bugs
[07:40] <pitti> hey ajmitch 
[07:43] <Hobbsee> pitti: where can i find info on the ffmpeg soname transition?
[07:43] <Hobbsee> as in, to find out which packages are affected
[07:45] <pitti> Hobbsee: same as where StevenK works from: http://people.ubuntu.com/~pitti/tmp/cruft/
[07:45] <Hobbsee> pitti: bah.  where's that duncecap?
[07:45] <StevenK> My blood, sweat, tears, and occasionally other bodily fluids are in those lists.
[07:51] <StevenK> pitti: Can I bug you to look at the MIRs for OpenAL and co to banish rss-glx from that list? :-)
[07:54] <pitti> StevenK: ah, yes
[08:20] <StevenK> pitti: Okay, that should be everything bar libavg for the libavformat0d transition. They should all get published and built over the next 2 hours.
[08:20] <pitti> StevenK: yay you
[08:21] <fabbione> do we already have GutstyReleaseNotes ?
[08:21] <pitti> fabbione: we have them in distributed form in the specs at least
[08:21] <fabbione> pitti: this is a bug fix.. there was no spec for it and it needs a note
[08:22] <fabbione> "whops we just figured that we were writing an almost totally wrong SUN disk label on sparc.. and you really want to update it by running fdisk manually and reboot"
[08:22] <pitti> fabbione: I don't see an existing one, but feel free to create https://wiki.ubuntu.com/GutsyGibbon/ReleaseNotes
[08:22] <fabbione> kind of thing..
[08:23] <fabbione> pitti: yeah i couldn't find one either.. just making sure the RM god was ok with that before doign
[08:23] <Hobbsee> fabbione: he'll come and hit you with lightening soon if he isnt
[08:24] <fabbione> Hobbsee: no thanks. i am not in "getting spanked" mood today
[08:24] <pitti> lol
[08:24] <Hobbsee> fabbione: i wouldnt have called lightning an implement for spanking...more "blasting"...but whatever you prefer :)
[08:25] <pitti> nah, unlike Hobbsee I don't eat/burn people :)
[08:25] <Hobbsee> pitti: you sad, deprived person.
[08:26] <Hobbsee> pitti: besides, i'm not the one who normally eats people.
[08:27] <StevenK> Yay. libavg fails to build because of avformat API changes. Not totally unexpected.
[08:31] <pitti> StevenK: ok, openal/freealut checked and promoted; yay for two more obscure libraries in main :/
[08:31] <fabbione> BRRRRRRRR
[08:31] <fabbione> pitti: i use them almost regularly
[08:31] <Hobbsee> pitti: could you be persuaded to do pinentry-qt too?
[08:31] <fabbione> pitti: x-plane needs them too :)
[08:32] <pitti> right, but this is highly universe games stuff
[08:32] <fabbione> pitti: x-plane is not in archive :)
[08:32] <pitti> Hobbsee: that sounds more delicate, I'll have another look later; breakfast first :)
[08:32] <Hobbsee> pitti: heh
[08:33] <fabbione> pitti: http://www.x-plane.com/ <- better than FlightSimulator
[08:33] <StevenK> fabbione: Better than FlightGear?
[08:33] <fabbione> StevenK: mostlikely
[08:34] <StevenK> Ahh, x-plane ENOFREE
[08:34] <pitti> fabbione: looks cool
[08:36] <fabbione> pitti: it is.. 9 dual layers DVD + updates + plugins + SDK.. highly portable etc.
[08:36] <fabbione> pitti: i can take it with me in London to show
[08:36] <fabbione> StevenK: right, but it runs everywhere and it's FAA certified for official pilot training
[08:37] <StevenK> Which automatically kicks FlightSim all over the place.
[08:37] <fabbione> yeps
[08:38] <fabbione> it has a lot more tho.. you should really give it a shot
[08:38] <fabbione> StevenK: http://people.ubuntu.com/~fabbione/pics/random/xplane.jpg
[08:38] <fabbione> (running on Linux expect on ppc where port is beta stage)
[08:38] <StevenK> I think I'd rather get a handle on flying with FlightGear before I shell out cash for x-plane. :-)
[08:39] <fabbione> StevenK: it's cheap tho.. like 60 USD or so
[08:39] <fabbione> to be a game of that size is almost no money
[08:39] <fabbione> (given the 10 dual layer DVD)
[08:41] <persia> fabbione: Can x-plane actually handle all 34 buttons on that joystick under linux?
[08:41] <fabbione> persia: yes. I have 2 joysticks attached to it
[08:41] <fabbione> it's fully configurable
[08:41] <fabbione> 2 joysticks and i still can't cover all the config options for flying
[08:42] <persia> fabbione: Thanks.  I wasn't sure it would work (given flightgear's arrangements).  Thanks.
[08:44] <fabbione> pitti: could you check if gutsy-changes died again?
[08:44] <Hobbsee> fabbione: should be alive.
[08:44] <Hobbsee> fabbione: i saw at least one from you before
[08:44] <fabbione> ok then it's in the queue to arrive here i guess
[08:45] <StevenK> fabbione: There ought to be about four from me.
[08:45] <Hobbsee> accepted parted
[08:45] <Hobbsee> fabbione: ^
[08:45] <fabbione> Hobbsee: ok.. i just didn't get them yet
[08:45] <Hobbsee> :)
[08:45] <fabbione> Hobbsee: yeah parted and util-linux
[08:51] <fabbione> StevenK: got one second for me?
[08:52] <StevenK> fabbione: Certainly.
[08:52] <fabbione> StevenK: could you try to resolve smtp.fabbione.net from where you are?
[08:53] <fabbione> or telnet to it on port 25
[08:53] <fabbione> looks like emails are not coming back to me
[08:53] <StevenK> It responds.
[08:53] <fabbione> ok thanks
[08:53] <StevenK> 220 trider-g7.fabbione.net ESMTP Postfix (Ubuntu)
[08:53] <fabbione> yeps..
[08:53] <fabbione> that's right
[08:53] <fabbione> thanks mucho
[08:54] <StevenK> No problem
[09:03] <StevenK> Okay, then. quilt is .... interesting.
[09:04] <persia> StevenK: Why?
[09:04] <pitti> fabbione: I don't have particular powers to check that
[09:05] <StevenK> persia: It's the first time I've used it.
[09:05] <pitti> fabbione: I can just talk to the soyuz guys; did you get a reject email again?
[09:05] <persia> StevenK: Ah.  Yes, it's a little different.
[09:06] <StevenK> persia: A little you say? :-)
[09:07] <persia> StevenK: Yes, in the same way that the width of the pacific ocean is a little far to swim (although I believe it has been done).
[09:09] <pitti> Hobbsee: WTH did you do?
[09:09] <lifeless> Hobbsee: !!!
[09:09] <Hobbsee> cut myself on a can by accident
[09:11] <pitti> fabbione: it seems to work for StevenK at least
[09:12] <Hobbsee> right.  fixed.  darned inconvienent, that
[09:12] <Hobbsee> heya lifeless 
[09:13] <lifeless> hiya.
[09:17] <StevenK> pitti: libavg fixed and just uploaded, too.
[09:17] <pitti> cheers
[09:18] <pitti> StevenK: publisher is still running for the promotion change
[09:19] <StevenK> Ahh
[09:19] <fabbione> pitti: it seems to be a canonical mail server issue.
[09:19] <fabbione> pitti: nothing you can do about it
[09:22] <StevenK> Hrm. My libavformat fixing should have fixed libavcodec too.
[09:23] <StevenK> pitti: Was it decided that the Beryl stuff would be killed?
[09:29] <pitti> Hobbsee: I have the entire Douglas Adams collection here :)
[09:29] <pitti> StevenK: not really officially decided, but I guess it makes sense to kill those
[09:29] <Hobbsee> pitti: then you know what the friendly green letters say on the front cover :)
[09:29] <Hobbsee> i'm disappointed - my school library never had all of them.
[09:30] <Hobbsee> i wonder if the uni library does.....
[09:30] <pitti> Hobbsee: they don't have the real essentials? :/
[09:31] <Hobbsee> pitti: seems not
[09:31] <Hobbsee> pitti: our school wasnt great on such things.
[09:32] <Hobbsee> they had the first 2 books of tomorrow when the war began - and note that the first was on the prescribed reading list for one of the subjects
[09:32] <StevenK> Heh. My wife and I both have a complete set.
[09:34] <siti> nag: https://bugs.launchpad.net/ubuntu/+source/gksu/+bug/66518 <-- debdiff here for a very annoying bug, it's been sitting here far to long :(
[09:34] <ubotu> Launchpad bug 66518 in gksu "[Edgy + Feisty]  Startup Notification broken" [Medium,Confirmed]  
[09:34] <siti> please could some dev look at it :'(
[09:35] <seb128> mvo: ^
[09:35] <mvo> siti: looking
[09:36] <siti> thanks
[09:38] <mvo> siti: we usually do only fix major crashers or security problems in already released versions of ubuntu. 
[09:38] <siti> yes, but for gutsy would be good
[09:38] <mvo> siti: right, gutsy should be fixed, but let me double check
[09:38] <siti> (the debdiff would need to be changed for gutsy)
[09:38] <siti> thanks
[09:40] <pitti> carlos: the current gutsy update langpacks are relative to 20070625?
[09:41] <seb128> pitti: would it be easy to make apport ignoring g_logv and g_log for the bug title?
[09:41] <pitti> seb128: it would be cleaner to also 'unwind' it from StacktraceTop; would that be right?
[09:41] <seb128> pitti: ie, bug #123462
[09:41] <seb128> yes
[09:41] <pitti> seb128: so the new gtk/glib now always calls g_logv() for a crash?
[09:41] <ubotu> Launchpad bug 123462 in gnome-vfs2 "gnome-vfs-daemon assert "volume->priv->drive == drive"" [Medium,Confirmed]  https://launchpad.net/bugs/123462
[09:41] <seb128> g_assert_warning () from /usr/lib/libglib-2.0.so.0
[09:41] <seb128>  gnome_vfs_volume_unset_drive_private ()
[09:42] <seb128> pitti: no, only for a gdk_x_error or an assert
[09:42] <pitti> seb128: seems that we actually should unwind until (including) g_assert_warning()?
[09:43] <carlos> pitti: they should
[09:43] <carlos> let me check to be sure...
[09:43] <pitti> seb128: so this bug would be shown as a crash in gnome_vfs_volume_unset_drive_private()
[09:43] <seb128> pitti: this bug is an assertion in gnome_vfs_volume_unset_drive_private() yes
[09:44] <seb128> we want to keep the g_assert_warning() though
[09:44] <seb128> to indicate it's an assertion
[09:44] <pitti> seb128: ah, ok; as you wish
[09:44] <pitti> seb128: right now I don't have logic to construct the bug title from a different StacktraceTop line than the first
[09:45] <pitti> seb128: that would introduce heuristics on two different places and might interact in a weird way
[09:45] <seb128> pitti: I'm fine having the title mentionning the g_assert_warning
[09:46] <pitti> seb128: ok; so I only unwind g_log() and g_logv() for now?
[09:47] <seb128> pitti: yes
[09:47] <pitti> cool, will do that
[09:47] <seb128> pitti: or do you think dropping the g_assert_warning as well would make sense?
[09:47] <pitti> seb128: it might, depending on how useful it is for dup detection
[09:48] <pitti> seb128: if it's just a standard function that happens on every assert, it should probably go
[09:48] <seb128> would it be easy to change the title when there is a g_assert_warning?
[09:48] <pitti> seb128: would be an ugly special case, so I would need to think about some generalization, but yes
[09:49] <seb128> "gnome-vfs-daemon crashed with signal 5 in g_logv()" is the current title
[09:49] <seb128> "gnome-vfs-daemon crashed on assert in gnome_vfs_volume_unset_drive_private()"
[09:49] <seb128> that would be better ;)
[09:49] <pitti> seb128: so g_assert() actually raises signal 5 now?
[09:52] <carlos> pitti: yes, it is
[09:52] <carlos> only 172 files exported yesterday
[09:52] <seb128> pitti: g_assert calls g_log and then "abort()"
[09:52] <carlos> pitti: btw, I think I will need to delay when the process starts because the mirror took a bit more time today so the exports failed..
[09:54] <seb128> pitti: afaik g_assert didn't change
[09:57] <cjwatson1> Hobbsee: FWIW, self-dependencies are harmless
[09:58] <cjwatson1> (dpkg ignores them)
[09:58] <Hobbsee> cjwatson1: ahhh, right
[09:58] <cjwatson> Hobbsee: circular dependencies are *mostly* harmless, though can lead to postinst scripts being executed in something other than the order you might hope for
[09:58] <ion_> I cant believe i didnt think of this earlier. No need to quit Firefox anymore when watching a movie (since theres always some stupid script in some tab thats going to steal the precious CPU time from mplayer). I simply added pkill -STOP firefox and pkill -CONT firefox to my panel. :-)
[09:58] <Hobbsee> cjwatson: right
[10:00] <cjwatson> circular pre-dependencies are bad; circular build-dependencies are sometimes necessary but are a pain
[10:04] <Hobbsee> ahhh
[10:09] <Hobbsee> hooray!
[10:10] <siti> is there anyone passionate about no more source packages, or wants the debian packaging to be more collbrative? could they have a look at: https://wiki.ubuntu.com/NoMoreSourcePackages/GitImpl
[10:10] <siti> it's just a very rough idea, but so far I already see huge advantages
[10:13] <Mithrandir> siti: given our big commitment to bzr, I doubt we would want to build a solution around git's model.
[10:14] <siti> Mithrandir: yeah, I've thought about that, but would debian ever move to bzr?
[10:14] <siti> also bzr is probably to slow for large packages
[10:14] <Mithrandir> siti: debian is probably never going to adopt just one VCS.
[10:14] <siti> git is gaining traction...
[10:15] <Mithrandir> and bzr is getting faster
[10:15] <siti> is bzr as easy to script for, e.g. could I do something like this in bzr: git-ls-files --ignored -x \* -o --directory | grep -v debian | xargs rm -rf
[10:15] <siti> it automatically deletes all non-referenced files (much easier than writing cleaners in the rules file)
[10:16] <Mithrandir> bzr unknowns |xargs rm -rf?
[10:16] <siti> ok
[10:16] <cjwatson> (or 'bzr ls --ignored --unknown' to catch both ignored and otherwise-unknown files)
[10:16] <siti> hmm, I wonder if there would be a way to standardise this across multiple DVCS ;)
[10:17] <siti> ok
[10:17] <siti> I really want this to happen because currently (for me anyway and I bet many others) making debian packages is well horrible, dpatch/quilt are very annoying to work with
[10:18] <siti> quilt is a bit better, but it's very much the maintainer model
[10:22] <Hobbsee> pitti: ping
[10:22] <siti> Mithrandir: well, what do you think would need to happen for something like this (what ever DVCS) to be adopted
[10:22] <siti> because I really want it to happen...
[10:23] <pitti> hey Hobbsee 
[10:23] <Mithrandir> siti: lots of infrastructural work, probably.
[10:23] <Hobbsee> pitti: are you busy?  got a concept question about apport
[10:23] <Hobbsee> pitti: https://bugs.launchpad.net/bugs/123394 - how it handles during dist-upgrades, and memory usage.
[10:23] <ubotu> Launchpad bug 123394 in apport "adept notifier crashes and bogs down after a dist-upgrade" [Undecided,New]  
[10:27] <pitti> Hobbsee: hm, weird; the apport UI should load the problem reports one by one
[10:28] <pitti> Hobbsee: I'm not sure what exactly adept-notifier does, but I guess it just checks for new reports and calls apport-qt
[10:28] <pitti> Hobbsee: the root of the problem might be that libapt reports followup package errors; it should only report the first one
[10:28] <pitti> mvo: ^
[10:29] <mvo> pitti: it should ignore follow-up errors, let me check the report
[10:31] <mvo> pitti: the bugreport does not look like it is releated to libapts problem reporting
[10:35] <pitti> mvo: hm, ok; but something must have written those reports?
[10:37] <mvo> pitti: right. I have seen parts of kde crash during a upgrade, so it maybe just those?
[10:37] <pitti> mvo: ah, right, maybe
[10:38] <pitti> we'd need to actually see those crash reports
[10:38] <mvo> yeah
[10:38] <Hobbsee> what i'm wondering about is "is htere a safeguard to make sure apport + the upgrader doesnt take up all available ram?"
[10:39] <pitti> Hobbsee: apport> depends; it will limit core dumps to 3/4 of available RAM
[10:39] <pitti> Hobbsee: apport> but if there are multiple processes crashing in parallel, there will be multiple apports, too
[10:43] <dholbach> good morning
[10:45] <cjwatson> siti: if you're using a distributed VCS it really doesn't make sense to use dpatch or quilt *as well*, IMHO.
[10:45] <siti> cjwatson: yep
[10:47] <siti> I think we will see productivity gains for packages similar to what you would see with software if both moved from emailing patches -> DVCS
[10:51] <siretart> pitti: dammit, I forgot some uploads. Yes, will look into them!
[10:51] <siretart> Hobbsee: both
[10:51] <pitti> siretart: StevenK did some (all?) now; thanks for looking into it
[10:51] <StevenK> I did all of the libavformat0d ones.
[10:52] <StevenK> That should have also taken care of libavcodec0d.
[10:54] <StevenK> Nice how logs for currently building packages isn't available anymore.
[10:55] <Hobbsee> siretart: right
[10:56] <siretart> StevenK: thanks!
[10:56] <StevenK> siretart: No problem, and sorry if I stepped on your toes.
[10:56] <siretart> StevenK: not at all!
[10:56] <siretart> Hobbsee: is there any problem with that self dependency?
[10:57] <siretart> Hobbsee: I think it is a bug (or at least interesting behavior) of dpkg-shlibdeps: it perhaps should check that it is about to generate a self dependency and just drop it
[10:57] <StevenK> Hey, way cool. libgoffice-0-3 appears in it's own cruft list.
[10:59] <Hobbsee> siretart: dunno.  i would have thought all dependancies on itself were bad...
[11:00] <siretart> Hobbsee: I haven't noticed problems because of this. in any case, I think that should be fixed in dpkg-shlibdeps (if necessary at all)
[11:01] <Hobbsee> siretart: fair enough, i just noticed it by accident
[11:02] <StevenK> Oh, that's right. I can't fix kwave because the error defeats my C++ skills.
[11:03] <pitti> seb128: apport change for the assert unwinding committed
[11:03] <pitti> seb128: that doesn't change the subject structure for now, though
[11:03] <seb128> pitti: rock on ;)
[11:03] <seb128> ok
[11:03] <pitti> seb128: it'll end up as 'signal 5 in gnome_vfs_volume_unset_drive_private' ATM
[11:03] <seb128> that's good enough
[11:04] <seb128> triagers will stop to mark bugs duplicates because they crashed in "g_logv"
[11:04] <pitti> seb128: hm, seems it's urgent enough to upload it now?
[11:05] <seb128> pitti: would be nice
[11:05] <seb128> no hurry, but having bug title with an actual function name would be better
[11:09] <pitti> seb128: uploaded
[11:10] <seb128> pitti: danke
[11:17] <pitti> cjwatson: could you please mark https://code.launchpad.net/~kamion/langpack-o-matic/depends-fixes as 'merged'?
[11:19] <cjwatson> pitti: done
[11:19] <pitti> thanks
[11:22] <pygi> Hobbsee, why did you try to eat me? 
[11:43] <saispo> 3/win 24
[11:46] <Hobbsee> hiya sabdfl 
[11:47] <Hobbsee> pygi: because you clearly were wanting it?  I could have fed you to the troll in #ubuntu-offtopic if you like, or the one in #ubuntu-ops
[11:47] <pygi> Hobbsee, I wasn't even here, I was sleeping!
[11:47] <Hobbsee> pygi: sure sure.  you were just pretending.
[11:47] <pygi> :'(
[11:52] <pygi> (the mess)
[11:52] <Hobbsee> heh
[11:55] <Fujitsu> Hobbsee: What about the poor kernelspace channels?
[11:56] <Hobbsee> Fujitsu: the trolls tend not to go there
[11:56] <Hobbsee> Fujitsu: and even if they do, they dont bring 50 spambots with them.
[11:56] <Hobbsee> it's much easier to kb one into the middle of next millenium than 50
[11:57] <Fujitsu> Ah. Fun.
[11:59] <Hobbsee> Fujitsu: of course!
[12:19] <StevenK> seb128: Is Gstreamer 0.8 going to stay around for Gutsy, or are we going to kill it?
[12:19] <Hobbsee> Fujitsu: and 2 more gone from #ubuntu
[12:20] <seb128> StevenK: it's on my list of things to clean, why?
[12:21] <StevenK> seb128: gstreamer0.8-flac depends on an old version of libflac.
[12:21] <StevenK> seb128: Trying to determine if it'd be a waste to spend time on it.
[12:21] <seb128> yeah, don't bother with 0.8
[12:21] <Mithrandir> we still have gst 0.8 in the archive?
[12:22] <StevenK> Mithrandir: It seems so.
[12:22] <Mithrandir> sounds crackful
[12:23] <seb128> Mithrandir: there is still some apps using it
[12:25] <Mithrandir> seb128: goobox and gstreamer-editor?
[12:25] <Fujitsu> goobox looks to be about it.
[12:25] <Fujitsu> What Mithrandir said.
[12:26] <seb128> Mithrandir: yeah, it's good enough to be cleaned now
[12:26] <seb128> Mithrandir: that's what I said, it's on my "to clean" list for gutsy ;)
[12:27] <Mithrandir> seb128: you and your lists. ;-)  Always on top of things. :-P
[12:27] <dholbach> there was a goobox update that crevette is working on
[12:27] <seb128> lol
[12:27] <dholbach> could possibly be they switched to gst0.10
[12:27] <Fujitsu> dholbach: They would be fairly deranged to not have siwtched.
[12:27] <Fujitsu> *switched
[12:28] <Fujitsu> goobox in experimental uses 0.10, so I presume they've switched.
[12:29] <dholbach> oh nice
[12:29] <dholbach> thanks Fujitsu, I prodded crevette about it
[12:29] <Fujitsu> Is gstreamer-editor any use if there's nothing using 0.8?
[12:29] <seb128> dholbach: he knows about it
[12:30] <seb128> Fujitsu: no
[12:30] <seb128> Fujitsu: that's just a matter to update goobox and clean the gst0.8 stack, I'll do it this week
[12:30] <Fujitsu> seb128: Great :)
[12:41] <linux_user400354> where can i find the default wallpapers from breezy? ive searched at packages.ubuntu.com and didnt find it.
[12:42] <pitti> Hobbsee: harsh
[12:42] <Hobbsee> flipping hell....
[12:43] <Hobbsee> pitti: when he asks the same question in 4+ places...
[12:43] <Hobbsee> pitti: besides, i've already kickbanned over 10 people today - the trolls are out in force.
[01:03] <StevenK> pitti: Regenerate the cruft lists when you feel so inclined, thanks.
[01:03] <pitti> StevenK: grinding
[01:04] <StevenK> Poor drescher.
[01:09] <cjwatson> isn't that list cronned?
[01:10] <pitti> not yet, I didn't manage to do that any more on Friday
[01:16] <cjwatson> StevenK: Shouldn't requestsync use 'rmadison -u ubuntu' now? It's probably more reliable (i.e. less dependent on the local system) than 'apt-cache madison'.
[01:17] <cjwatson> StevenK: likewise it seems like it should use rmadison rather than looking at http://qa.debian.org/madison.php manually
[01:17] <StevenK> Good idea, I'll hack that in shortly.
[01:20] <siretart> pitti: I've tried to upload g-wrap, but my upload seems to have been eaten several times. I did uploads afterward, that did got accepted, but g-wrap is constantly getting eaten. Who can I bug about this?
[01:20] <pitti> siretart: let me take a look
[01:21] <Mithrandir> siretart: seems like you've triggered a bug in soyuz
[01:21] <Mithrandir> 11:00:24 ERROR   Exception while accepting: ERROR:  permission denied for relation bugtracker
[01:22] <pitti> StevenK: /cruft/ updated
[01:22] <siretart> Mithrandir: uh, shall I reupload without the Launchpad-Bugs-Fixed: magic?
[01:23] <Mithrandir> siretart: probably a good workaround
[01:23] <Mithrandir> I'll nag cprov when he wakes up
[01:23] <siretart> willdo so. thanks
[01:26] <siretart> Mithrandir: yepp, removing that line helped here
[01:31] <StevenK> pitti: Thanks.
[01:32] <StevenK> pitti: It was as I suspected - libavformat0d and libavcodec0d can be killed when you wish.
[01:32] <pitti> StevenK: cool, thanks
[01:33] <pitti> done
[01:34] <pitti> StevenK: would you mind looking into what Debian did with curl? Did they add a shim to provide the old ABI as well, or did they revert an upstream version, and is there some discussion somewhere?
[01:34] <pitti> StevenK: if the new package in Debian is sane, we should sync/merge it ASAP and re-rebuild the libcurl4 dependencies, I guess
[01:35] <StevenK> Where's geser, he's been following this madness closely.
[01:37] <ion_> Does it look like PulseAudio is going to be installed by default in gutsy, btw?
[01:37] <minghua> I roughly remember Debian's curl situation
[01:38] <minghua> upstream bumped SONAME version, but no actually ABI change
[01:38] <minghua> Debian first changed package name, then changed back
[01:39] <minghua> now libcurl3-* ships libcurl*.so.4, and a symlink of libcurl*.so.3 pointing to .so.4
[01:39] <minghua> All the details should be on debian-release list
[01:42] <pitti> \o/ http://ppa.dogfood.launchpad.net/ubuntu-langpack/ -> it's working! /me bounces
[01:45] <StevenK> pitti: Okay, it seems that Debian has reverted to libcurl3 package names... It does mean that libcurl4 is now cruft, and not libcurl3, if we sync/merge the latest curl.
[01:45] <pitti> minghua: " now libcurl3-* ships libcurl*.so.4" -> eewww
[01:45] <pitti> StevenK: well, I guess we have to bite that bullet
[01:46] <StevenK> As a symlink to libcurl*.so.3, even more disgusting.
[01:46] <Mithrandir> pitti: if you could NEW hildon-control-panel and hildon-control-panel-l10n, that'd be good.
[01:46] <pitti> we only have 10ish rdepends of libcurl4, that seems manageable
[01:46] <StevenK> pitti: Would you like me to push that onto on my queue after cjwatson's suggested changes to requestsync?
[01:46] <pitti> StevenK: that would be great
[01:46] <pitti> StevenK: the ubuntu diff should be trivial (removing a build dep AFAIR)
[01:47] <pitti> Mithrandir: is 30 minutes ok? I still have a lot of state in my head which I'd like to get written down
[01:47] <StevenK> pitti: Okay, cool.
[01:47] <Mithrandir> pitti: sure, no problem.
[01:48] <pitti> ihatemyinternet: hey, it's our internet, too! :)
[01:48] <ihatemyinternet> pitti: ;-) ihatemyisp is probably more accurate
[02:12] <Daviey> ping bdmurray, BenC, or dholbach
[02:14] <seb128> Daviey: you might want to give some context, they are not around
[02:14] <Daviey> Just want to talk to them about a 'hug' :)
[02:15] <seb128> k
[02:15] <seb128> dholbach should be around in not too long
[02:15] <Hobbsee> mvo: are you around?
[02:16] <Hobbsee> [22:16]  [Whois]  mvo has been idle for 59 minutes and 9 seconds.
[02:16] <Hobbsee> seems not.
[02:16] <Daviey> thanks Hobbsee 
[02:16] <seb128> Hobbsee: you might want to ask your question ;)
[02:17] <Hobbsee> seb128: i could.  there are multiple things though :)
[02:17] <siti> lol
[02:17] <mvo> Hobbsee: yes
[02:17] <mvo> Hobbsee: but hacking
[02:18] <Hobbsee> mvo: right.  so i should save my question about ubuntu-restricted-extras, and recommends?
[02:18] <mvo> Hobbsee: it is not working with recommends currently?
[02:19] <Hobbsee> mvo: no.  it's getting fed the recommends correctly, but it's not installing them by default.
[02:19] <Hobbsee> mvo: it's a metapackage, in section metapackages, so i'm not sure hwy
[02:19] <Hobbsee> (thru apt, at least)
[02:20] <mvo> Hobbsee: oh, that is indeed strange, does it make a difference if you run it with "apt-get install --install-recommends ubuntu-restricted" ?
[02:20] <mvo> Hobbsee: what does apt-get install ubuntu-restricted -o Debug::pkgDepCache::AutoInstall=true" say?
[02:22] <Hobbsee> mvo: with the first, the recommends get installed too, as expected, and the second doesnt seem to produce any different output to normal
[02:22] <Hobbsee> as in, shows the recommends, but that they're not going to be installed
[02:24] <Hobbsee> (can show you the output if you want, but there's nothing terribly interesting)
[02:24] <mvo> Hobbsee: is that for the package that is currently in the gutsy archive? or is that for your local one? I would like to reproduce this here
[02:24] <Hobbsee> mvo: it's in the gutsy archive
[02:25] <Hobbsee> mainly because i was lazy and didnt want to set up a repository here.
[02:25] <Hobbsee> and the previously uploaded version didnt compile (oops, dunno how i forgot to test that one)
[02:26] <Mithrandir> :-P
[02:26] <StevenK> Ouch!
[02:26] <StevenK> That *stings*
[02:26] <Mithrandir> oops, I just hit the moon
[02:26] <Hobbsee>  * HOBBSEE IS ON THE MOON, DAMMIT!!!!
[02:26] <Mithrandir> hard to fling them over the horizon.
[02:27] <Hobbsee> or is it fire?  or police?
[02:28] <StevenK> :-P
[02:32] <sladen> pitti: would there be a way to tag the failed stacktraces.  hopefully they can be rerun whenever gdb/each is actually fixed
[02:33] <Hobbsee> sladen: Hobbsee change notifications?
[02:33] <pitti> sladen: sure, we can just invent a tag
[02:34] <sladen> pitti: that's the easy bit, having apport/launchpad add it is the hard bit
[02:34] <mvo> Hobbsee: the problem is that the section is multiverse/metapackages
[02:34] <mvo> Hobbsee: for apt that is different from just "metapackages"
[02:34] <sladen> pitti: I guess if it's an apport thing it's easy, but it's a launchpad needed change they we could be here until next year
[02:34] <Hobbsee> mvo: ah, point
[02:34] <pitti> sladen: ah, I see
[02:34] <pitti> sladen: no, it would be changed in the retracer bot
[02:34] <pitti> sladen: the entire reprocessing magic is not LP code, but my own
[02:35] <sladen> pitti: apt-get source apport ?
[02:35] <Hobbsee> mvo: therefore...you cant have any metapackages installing recommends that arent in apt. that sounds like a bug.
[02:35] <mvo> Hobbsee: its not hard to make apt more clever, I'm just busy currently with other tasks :/
[02:35] <pitti> sladen: bzr get https://code.launchpad.net/~ubuntu-core-dev/apport/ubuntu
[02:35] <Hobbsee> mvo: no problem
[02:35] <pitti> sladen: (as apt-get source will tell you as well nowadays \o/)
[02:39] <StevenK> pitti: curl uploaded.
[02:39] <pitti> StevenK: yay
[02:40] <StevenK> pitti: So now you get to rescue it out of binary NEW in about 2 hours. :-)
[02:40] <pitti> StevenK: oh? which binary package does it introduce?
[02:41] <StevenK> Oh, wait. If libcurl3 hasn't been NBS'd, you don't.
[02:43] <pitti> StevenK: no, not with that insane list of rdepends :)
[02:46] <StevenK> Heh
[02:47] <StevenK> Actually, you'll need to regenerate cruft in 2 hours to deal with libcurl4, not 3.
[02:58] <dholbach> hi Daviey
[03:38] <StevenK> pitti: cmus now Build-Depends on libfaad2-dev. Should we just drop it to multiverse?
[03:39] <pitti> StevenK: sounds fine
[03:39] <StevenK> pitti: Can you throw the magical switch then, please?
[03:40] <pitti> thrown
[03:40] <StevenK> Danke
[03:44] <StevenK> pitti: After the curl mess is sorted out, it looks like the largest transition is libflac.
[03:52] <mdz_> pitti: I just had apport decline to file a bug because it said it was already reported 'in the bug displayed in the web browser', but nothing was displayed, and I got an error from firefox about it already running
[03:52] <mdz_> pitti: was it trying to launch the browser as root or something?
[03:52] <mdz_> (this was a system daemon crash)
[03:52] <pitti> mdz_: apport GUI itself is ran as root, but it tries very hard to start firefox as user
[03:53] <pitti> mdz_: every other browser works without a hitch, but firefox just fails with that sometimes
[03:53] <pitti> I didn't yet find out why
[03:53] <pitti> I pass the necessary environment variables etc.
[03:53] <pitti> mdz_: do you already have a firefox open?
[03:54] <mdz_> pitti: I just tried again, and it still doesn't display in the browser, and no error this time
[03:54] <mdz_> pitti: yes
[03:54] <persia> StevenK: ObjectiveC?
[03:54] <StevenK> persia: Does that use .m and .h?
[03:54] <pitti> mdz_: so it opened ffox, but doesn't load any page?
[03:54] <pitti> mdz_: might be a broken bug pattern; which package is this for?
[03:55] <StevenK> error: expected specifier-qualifier-list before 'FLAC__FileDecoder' -- and I'm neatly unsure what that means. :-)
[03:55] <StevenK> I've never needed to touch ObjC.
[03:56] <persia> StevenK: Yep.  That's Objective C.
[03:58] <StevenK> Hrm. I think curl's binaries are going to miss this publisher run.
[04:02] <StevenK> persia: And what about the compiler error? :-)
[04:04] <persia> StevenK: I appear to be having difficulties with my disks currently, so I can't reproduce.  I'm searching now (is this a good time to mention that my one (1) engagement with Objective C was to reverse-engineer a replacement in Java?)
[04:06] <siti> mvo: https://bugs.launchpad.net/ubuntu/+source/gdebi/+bug/87029  I am getting this problem in gutsy
[04:06] <ubotu> Launchpad bug 87029 in gdebi "[apport]  gdebi-gtk crashed with error in finish_dpkg()" [Medium,Fix released]  
[04:06] <StevenK> persia: Ohh.
[04:07] <persia> StevenK: It appears that a specifier-qualifier-list is something of the form "const long" (that's actually the qualifier followed by the specifier).
[04:08] <persia> StevenK: http://objc.toodarkpark.net/objctoc.html might be interesting reading
[04:08] <mvo> siti: let me check
[04:09] <StevenK> persia: But how am I supposed to qualify something that is a struct?
[04:10] <persia> StevenK: use the struct-or-union-specifier or the struct-or-union-identifier, I think.
[04:10] <persia> StevenK: In comprehensible language: put the name of whatever it is before the thing that returns it.
[04:11] <StevenK> That isn't very comprehensible. :-)
[04:14] <persia> StevenK: Sorry.  It made more sense to me than the first thing I said.  From what I understand, it's complaining that FLAC__FileDecoder is untyped, and wants a type.  Either the syntax in the area is generally wrong, there is an issue with namespaces, or you need to insert the struct-or-union-specifier for the return type of FLAC__FileDecoder before the declaration (or something).  I would probably make more sense with a pastebin to look at :)
[04:17] <StevenK> Way cool.
[04:17] <StevenK> Launchpad is reporting that cmus 2.1.0-2ubuntu1 is both Published and PendingRemoval in multiverse.
[04:17] <pitti> StevenK: hm, is there a new version of it?
[04:18] <StevenK> Not that I can see.
[04:21] <StevenK> persia: I think the *real* reason is that the FLAC header files no longer declare FLAC__FileDecoder.
[04:23] <persia> StevenK: That would certainly do it.  Unless you're especially daring, I'd suggest hunting an expert in Objective C (or reading the manual not a few times) prior to attempting a patch.
[04:24] <StevenK> It seems I'm especially daring, given I've just fired off another test build
[04:24] <persia> StevenK: Best of luck :)
[04:26] <StevenK> Oh, it helps if you pass the right .dsc to sbuild. :-P
[04:34] <iwj> StevenK: The __ suggests to me that it's an internal name, not intended for use by things outside the flac libraries.
[04:38] <StevenK> iwj: It seems everything in FLAC starts with FLAC__.
[04:41] <iwj> Weird.
[04:41] <StevenK> iwj: Weird seems to describe FLAC quite well. :-)
[04:42] <StevenK> Why I'm up fixing a package that Build-Depends on it at quarter to 1 escapes me, though.
[04:51] <pitti> StevenK: ah, indeed curl is in NEW due to libcurl3-dbg *grumpf*
[04:51] <pitti> StevenK: built on all 5 arches, NEWed
[04:52] <StevenK> Hah! I win!
[04:52] <StevenK> Ahem. Or something.
[04:52] <StevenK> pitti: Thanks.
[04:52] <pitti> np
[04:52] <StevenK> Just in time for the next publisher run.
[04:52] <ScottK> pitti: Woud this be a decent time to harass you about my MIR?
[04:53] <pitti> ScottK: let me open a tab for that to do it tonight
[04:53] <Hobbsee> hehe
[04:53] <pitti> ScottK: pinentry, right?
[04:53] <ScottK> pitti: Yes - https://wiki.kubuntu.org/MainInclusionReportPinentry
[04:53] <Hobbsee> pitti: it's after breakfast.  keep putting it off :P
[04:55] <ScottK> pitti: Thanks.
[04:57] <agoliveira> Hi guys. I think I found another bug regarding instalation. I dont' have the means to send a log now but for what I can see, something called ma-search-users crashed badly with a glibc error "corrupted double-linked link". Anyone wants to talk about it?
[04:57] <agoliveira> I mean, Gutsy Tribe 2
[04:58] <cjwatson> agoliveira: that would be evand
[04:58] <evand> arr
[04:58] <agoliveira> Oh, indeed.
[04:59] <agoliveira> evand: So, what to check it out?
[04:59] <geser> StevenK: Hi, I see that curl was merged
[05:00] <evand> agoliveira: can you file a bug and attach /var/log/syslog from the livecd.  Let me know the number and we'll go from there.
[05:00] <geser> any reason why curl 7.16.2-6 wasn't merged?
[05:01] <agoliveira> evand: Sure.
[05:01] <evand> thanks
[05:03] <agoliveira> evand: Hmmm... the bug does not show up in /var/log/syslog. I saw it because I had to run ubiquity in a terminal. Does this log go anywhere else?
[05:03] <Hobbsee> geser: i think StevenK has something to do with it
[05:03] <StevenK> geser: Because MoM didn't tell me about it?
[05:04] <cjwatson> evand: while randomly looking at ma-search-users.c I noticed that it calls opendir but never calls closedir
[05:05] <geser> StevenK: the little bit I know about that libcurl mess, I attached as a comment to bug #122775 (which can be closed now btw)
[05:05] <ubotu> Launchpad bug 122775 in curl "merge curl 7.16-5 from Debian unstable (main)" [Undecided,New]  https://launchpad.net/bugs/122775
[05:05] <evand> hilarity.  That code is a bit of a mess in general, I'll take a look and push out a slightly less buggy version.
[05:05] <evand> thanks for the heads up
[05:06] <geser> StevenK: so that means that the packages upload for a libcurl3 -> libcurl4 transition can now be uploaded again for libcurl4 -> libcurl3?
[05:07] <StevenK> geser: Yes.
[05:08] <StevenK> geser: I will merge -6 when MoM/ftp.d.o catch up
[05:08] <StevenK> And of course packages.d.o doesn't know about -6. POS.
[05:10] <geser> StevenK: I usually use the PTS for such thinks and it knows about -6
[05:11] <StevenK> geser: Yes, but I can't read the changelog with the PTS.
[05:11] <geser> it would be good to merge -6 soon, as else pacakging still depending on libcurl3 get broken as the symlink from libcurl3*.so to libcurl4*.so is missing
[05:12] <geser> unfortunately the PTS is missing currently the recent changelog (ends at -5)
[05:14] <evand> agoliveira: on second thought, can you run through with --debug.  It might give me a slightly better idea of why it's failing. (Attach /var/log/installer/debug)
[05:17] <StevenK> pitti, geser: curl 7.16.2-6ubuntu1 uploaded.
[05:17] <pitti> yay
[05:17] <StevenK> geser: Now, shush
[05:18] <agoliveira> evend: I have to go lunch in a few minutes but I'll try that as soon as I return, ok?
[05:18] <evand> sounds good, thanks again
[05:19] <agoliveira> My pleasure
[05:19] <StevenK> pitti: Let's leave re-running the cruft until you re-surface tomorrow, methinks.
[05:38] <geser> StevenK: thanks, I will start upload packages from universe for the libcurl4 -> libcurl3 transition in a few days (probably the second half of this week)
[05:57] <geser> StevenK: btw apt-cache lists currently only 43 rdepends for libcurl3* and 90 for libcurl4*, so we had already done the most part of the libcurl3 -> libcurl4 transition
[05:57] <pitti> geser: urgh? apt-cache rdepends libcurl4|wc -l says 19 here, and that includes some noise
[05:58] <pitti> geser: ah, erk, -gnutls
[06:08] <geser> pitti: libcurl4 (no SSL, dropped again), libcurl4-gnutls and libcurl4-openssl
[06:08] <pitti> geser: right, I didn't consider that
[06:11] <geser> let's hope this libcurl mess is finished soon
[06:21] <mvo> hey glatzor_
[06:23] <glatzor_> servus mvo
[06:27] <agoliveira> evand: I just did what you said. The error I found shows up in the console so I captured it too just as it happens. After that I canceld the instalation as I recall it does not happen again. Do you want me to open a bug and attach it or send the files for you directly?
[06:27] <agoliveira> s/said/asked
[06:27] <evand> bug please
[06:27] <agoliveira> Sure do.
[06:30] <Daviey> BenC: Hey, are you about?
[06:30] <BenC> Daviey: yeah
[06:30] <Daviey> Hey, can i pm quickly?
[06:31] <BenC> Sure
[06:37] <agoliveira> evand: Here it goes - https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/123584
[06:37] <ubotu> Launchpad bug 123584 in ubiquity "ma-search-users crash when installing Gutsy Tribe 2" [Undecided,New]  
[06:39] <evand> thanks agoliveira.  I'll follow up from there.
[06:39] <agoliveira> evand: No problem. If you need anything else, just ping me.
[06:39] <ogra> agoliveira, nah, you are :) sayiung bug #123584 would suffice to trigger it ;)
[06:39] <ubotu> Launchpad bug 123584 in ubiquity "ma-search-users crash when installing Gutsy Tribe 2" [Undecided,New]  https://launchpad.net/bugs/123584
[06:40] <mikmorg> cjwatson: ping
[07:01] <cjwatson> mikmorg: hi
[10:21] <LaserJock> cjwatson: are you available for a little debian-cd chat?
[10:21] <cjwatson> LaserJock: now's not a great time
[10:21] <LaserJock> cjwatson: ok, that's fine
[10:21] <cjwatson> sometime, sure
[10:21] <LaserJock> cjwatson: I've got a bzr branch on LP with changes for Edubuntu's addon cd
[10:22] <LaserJock> another time is fine
[10:46] <mikmorg> cjwatson: Hello again.
[11:36] <siretart> Mithrandir or cjwatson: please reject the gnochm upload to feisty-proposed. Clearly a mistake on my part, sorry :(
[11:37] <micahcowan> Would anyone like to sponsor my fix to vim/screen interaction (plus a typo in filetypes.vim)? bug 113227. cjwatson, particularly (since you uploaded the last vim).
[11:37] <ubotu> Launchpad bug 113227 in vim "Incomplete/broken mouse handling in screen" [Wishlist,In progress]  https://launchpad.net/bugs/113227
[11:43] <cjwatson> mikmorg: hello, only here briefly though. UK evenings tend not to be good times for me
[11:44] <cjwatson> micahcowan: opened in my web browser for attention tomorow
[11:44] <cjwatson> tomorrow
[11:44] <micahcowan> cool, thanks much! :)
[11:44] <Chipzz> micahcowan: I don't think that fix is correct actually
[11:45] <Chipzz> micahcowan: I think this comment is actually correct (or half-correct):
[11:45] <Chipzz> https://bugs.launchpad.net/ubuntu/+source/screen/+bug/113227/comments/4
[11:45] <ubotu> Launchpad bug 113227 in vim "Incomplete/broken mouse handling in screen" [Wishlist,In progress]  
[11:46] <Chipzz> micahcowan: IIRC there are 3 mouse protocols
[11:46] <micahcowan> Chipzz, it has been approved upstream (not yet in upstream source, though)
[11:46] <micahcowan> Was discussed fairly well on the ML (which has been down for the last few weeks, but you can see the archives).
[11:46] <Chipzz> micahcowan: I think the actual bug is that the termcap/terminfo for screen is incorrect
[11:47] <micahcowan> Chipzz, there is no means for termcap/terminfo to specify xterm-style mouse support.
[11:47] <micahcowan> Nor the other two protocols. Only the basic kmous.
[11:48] <micahcowan> Why did you give the link to my own comment?
[11:51] <Chipzz> one sec
[11:53] <Chipzz> take a look at:
[11:53] <Chipzz> http://chipzz.safehex.be/docs/rxvt-2.16/rxvt.ref
[11:53] <Chipzz> and
[11:53] <Chipzz> http://chipzz.safehex.be/docs/rxvt-2.16/dec_vt_mouse.html
[11:58] <micahcowan> Chipzz, what about them? I'm familiar with the escape sequences. My point is, terminfo/termcap do not have methods for specifying them nor their support.
[11:59] <micahcowan> They should, and that would be the better fix (I'm going to try to work with the ncurses team to make that happen), but in the meantime, checking the TERM value is how most apps are doing it (and how vim has been). My fix just ensures that screen is one of those accepted values (since it handles mouse sequences corretly, both in the case that it's being used by an "xterm", and when it's not).
[12:01] <micahcowan> Some ncurses-based apps also use the presence of kmous to indicate that the terminal also supports appropriate DEC/xterm-style mouse sequences, even though they really don't have anything to do with eachother. screen does lack kmous (no idea whether screen supports kmous); however, its presence would do nothing to resolve this issue, as vim doesn't check kmous to determine this.
[12:01] <Chipzz> hrrrm ok
[12:01] <Chipzz> I thought there was a way to check it
[12:02] <micahcowan> Yeah, I wish :/
[12:02] <agoliveira> kylem: yo! ping?
[12:02] <micahcowan> There's a couple other things that it would be nice for terminfo to support, such as the use of modifiers with special keys (like the cursor keys).
[12:04] <Chipzz> problem is, won't do you much good in the end
[12:04] <Chipzz> ncurses is built on top of termcap/terminfo
[12:04] <Chipzz> and even you fix it in ncurses
[12:04] <Chipzz> you still have a whole list of other OS's which it won't work on
[12:05] <Chipzz> legacy, legacy, and then some more legacy :P
[12:06] <kylem> agoliveira, ?
[12:07] <agoliveira> kylem: Hi. Stupid nOOb question: I want to try to create a device driver for an USB device in my spare time, any pointers on how to reverse engineer such a beast? There's no docs on it that I'm aware of.
[12:08] <agoliveira> kylen: ... if you ever tried that, I don't know ;)
[12:09] <kylem> if you have a windows driver, it's pretty easy
[12:10] <agoliveira> kylem: Yes, there is.
[12:10] <kylem> you can log all the usb urbs submitted to the device fairly trivially, there's some usb snooping drivers for windows/linux/macosx.
[12:10] <kylem> usb is actually one of the most pleasant protocols to reverse engineer, heh
[12:11] <agoliveira> kylem: Hmmm... so I think that's the way I go. ANy pointers to suitable docs?
[12:13] <kylem> agoliveira, http://sourceforge.net/projects/usbsnoop/
[12:14] <agoliveira> Cool. In case you're wondering, it's a simple stupid thing. I got one of this http://www.notebookreview.com/default.asp?newsID=3403 and I want to create a device for the small OLED display on the top.
[12:18] <agoliveira> kylem: Thanks man. I'm closing the shop for today. Bye.
[12:19] <kylem> night.