[12:06] <bur[n] er> anyone know what other package besides xorg-common needs to be reconfigured when upgrading from hoary to breezy?
[12:14] <bur[n] er> disregard that msg... reconfigure xserver-xorg was all that was needed
[12:14] <doko> lamont: pitti was looking at #13486
[12:15] <lamont> doko: he looked at it long enough to discover that -O1 works around the gcc issue...
[12:15] <lamont> I don't think he was looking at it still....
[12:36] <mdz> doko: gcc-3.3 wants to move to universe
[12:36] <mdz>  o gcj-3.3 libgcj4 libgcj4-awt libgcj4-common libgcj4-dev            {gcc-3.3}
[12:36] <mdz> not the source, but those binaries
[12:38] <Keybuk> a fresh breezy install actually feels quite nice
[12:38] <doko> mdz: these binaries aren't built anymore, they should vanish automatically?
[12:39] <doko> mdz: but wait, please keep it in main, we currently need gnat-3.3 as a build dep.
[12:40] <elmo> mdz: I just did a rene run that might help
[12:40] <elmo> anastacia always wants to demote NBS stuff

[12:44] <mdz> Keybuk: did you get accosted by the autofiends?
[12:44] <aroman> I'm trying to track down a bug in breezy related to i810 and Xorg. I'm getting an error: *** glibc detected *** double free or corruption (fasttop): 0x<address> ***. Now I'm not sure how to go about this bug. Is there any way to enable more messages from Xorg?
[12:44] <Keybuk> no :)  I may have upset them a bit though
[12:44] <mdz> Keybuk: yeah, colony 3 actually feels pretty good, belying the pain of its production
[12:45] <Keybuk> mdz: only through pain can art and beauty be acheived
[12:45] <mdz> aroman: Xorg logs detailed messages to /var/log/Xorg.0.log
[12:45] <aroman> mdz, more detailed than that :P
[12:45] <mdz> Keybuk: and art can be reduced to pain
[12:45] <mdz> Keybuk: or is it beauty?
[12:46] <Keybuk> dunno, we should consult jbailey; he's the expert on combining the three
[12:46] <aroman> is there a way to find out which version of gcc a binary has been compiled with?
[12:46] <mdz> aroman: what do you want, a log of every function call?
[12:46] <aroman> mdz, wouldn't mind that :)
[12:46] <Keybuk> I strace'd X once, it actually worked
[12:46] <mdz> aroman: then no, there isn't a way to get that
[12:46] <aroman> mdz, darn.
[12:46] <jbailey> Keybuk: Err... 
[12:47] <doko> aroman: hmm, look at the build logs ...
[12:47] <doko> http://people.ubuntu.com/~lamont/buildLogs/
[12:47] <aroman> k
[12:48] <jbailey> aroman: Some binaries have the information in a comments header.
[12:48] <Mez> keybuk: did you go tonight?
[12:48] <Nafallo> doko: I might have a bug for you ;-)
[12:48] <Nafallo> doko: Ubuntu isn't recognised in the default spellchecks in Ubuntu ;-)
[12:49] <aroman> jbailey, in the binary itself?
[12:49] <aroman> cuz.. it seems like Xorg is being built with gcc-4.0
[12:49] <aroman> but 2.6.12-latest is being built with gcc-3.4
[12:50] <aroman> now.. the i915 driver is in the kernel distribution, and I'm not sure how well they work together
[12:50] <Mez> jbailey, I think Scott might have had one too many
[12:50] <Keybuk> Mez: yeah, wasn't a huge turn-out, but was fun
[12:51] <doko> Nafallo: no, it's a bug that "Debian" is recognized.
[12:51] <Nafallo> doko: hehe
[12:51] <Nafallo> depends on how you look at it ;-)
[12:52] <aroman> X -version will say it's compiled with 3.4.5
[12:52] <aroman> hrm
[12:52] <jbailey> Mez: No, he's just begging for one more... And I won't let him have it. =)
[12:53] <SloMoSnail> aroman: X -version tells you the compiler used for the kernel... but not the one used for X ;)
[12:53] <jbailey> aroman: Hmm, someone asked this the other day, and I thought we figured out that strip didn't rip it out of the comments field.
[12:53] <jbailey> But I don't see it in any of my binaries here.
[12:54] <mjg59> jdub: Pls forward admin details kthxbi
[12:54] <jdub> mjg59: aha
[12:54] <jdub> mjg59: righto
[12:54] <aroman> SloMo, aha...
[12:54] <mjg59> jdub: Thanks!
[12:55] <aroman> jbailey, so what? it's a strip bug :S 
[12:55] <Nafallo> jdub: mailinglist? :-)
[12:55] <jbailey> aroman: =)
[12:56] <aroman> jbailey, I haven't done much "real" development. I know C/C++, but I haven't used all these programs like strip, etc... what exactly does strip do? afaik it strips debug information..
[12:57] <jbailey> Oh, hmm.  Current Debian unstable on ia64 doesn't strip the data out.  Current breezy on ppc does.
[12:57] <jbailey> It was in a Debian channel, so I was likely using the ia64 for the testing.
[12:58] <jbailey> aroman: By default, even without specifying -g, The compiler, linker, and assembler may put extra bits into your binary.
[12:58] <aroman> and strip removes the unneccesary code
[12:58] <jbailey> strip removes all the things that it knows you don't need to run the program.
[12:58] <aroman> ok
[12:58] <jbailey> Right.
[12:58] <aroman> that makes sense
[12:59] <aroman> well I'm having a weird X problem... basically X just dies without any error in the log.
[12:59] <jbailey> aroman: If you want to see a simple example, write a hello world file in C, and do a "gcc -S hello.c"
[12:59] <jbailey> And then look at what's in hello.
[12:59] <jbailey> s
[12:59] <aroman> XIO: fatal IO error 104 (Connection reset by peer) on X server ":0.0"
[01:00] <jbailey> aroman: Are you currently running as the same user you logged in as?
[01:00] <jbailey> My first guess is that you just don't have rights to the socket.
[01:00] <aroman> um
[01:00] <aroman> X doesn't start at all..
[01:01] <aroman> and I tried running startx or even just X as root
[01:01] <aroman> so this is with i810 driver (i915 in the kernel)
[01:01] <aroman> with vesa as my driver it works
[01:01] <aroman> so I do believe it is a i810 driver problem
[01:01] <jbailey> Trying 'strace startx'.  It'll give you a lot of information but somewhere in the 4 or 5 pages before it dies, you should see some syscall return an error.
[01:02] <jbailey> Ah, hmm.
[01:02] <jbailey> I don't know enough about X to guess that.
[01:02] <jbailey> Connection reset by peer is usually a permissions error of some sort for me.  Beyond that, I just sit in the corner and look sad until someone helps me.
[01:02] <aroman> by the way I am running latest hoary...
[01:03] <aroman> well I had this error waaaay back a few years in redhat 7.2 or 8... and it was an X problem which got fixed in an update...
[01:03] <aroman> and it was not a permissions problem
[01:03] <aroman> strace is interesting
[01:03] <jbailey> With X, anything is possible. =)
[01:04] <jbailey> It's sort of like a million monkeys got together to code it...  Except that they were really bright monkeys who only had obscure hardware to work with.
[01:07] <aroman> what the frau!? 
[01:08] <aroman> this is a line from sudo strace X 2> xlog && less xlog
[01:08] <aroman> open("/dev/tty". O_RDWR|O_NONBLOCK|O_NOCTTY) =  -1 ENXIO (No such device or address)
[01:09] <aroman> and after this the glibc error gets printed
[01:09] <aroman> now.. why is it trying to get /dev/tty? 
[01:09] <jbailey> Better question is "why is it failing"
[01:09] <aroman> yeah..
[01:09] <jbailey> Certainly on my system /dev/tty is 666
[01:09] <aroman> the device is there
[01:09] <aroman> and EVERYONE has rw permissions..
[01:10] <aroman> same
[01:10] <aroman> 666
[01:10] <aroman> crw-rw-rw
[01:10] <aroman> I guess c because it's a device node
[01:10] <aroman> hmm this is really interesting..
[01:11] <jbailey> aroman: "character device"
[01:11] <aroman> jbailey, ah
[01:11] <jbailey> It means that you can send things to it one byte at a time and expect it to be useful.  It also tends to be sequential read and writes only.
[01:11] <jbailey> The other option you'll usually see in devices is 'b' for block device.
[01:12] <aroman> god I need to eat something... I'll be back in 5 minutes...thanks for the insight jbailey... talk to you in a few
[01:12] <jbailey> aroman: Anytime. =)
[01:14] <mjg59> jdub: No mail from you...
[01:16] <jdub> hrm
[01:23] <Keybuk> jbailey, aroman: you mean other than the fact you can't open /dev/tty O_NONBLOCK?
[01:24] <Keybuk> at least, I don't think you can
[01:24] <jbailey> No idea.  The definetly gets into the realm of maybe crawling a bit too far into X's source code. =)
[01:24] <mdz> doko: all of the lapack/atlas3 stuff is still pending in anastacia
[01:26] <mdz> elmo: whoa, stuff is showing up in anastacia that isn't even present in breezy
[01:26] <elmo> after a cron.sync run?
[01:27] <mdz> elmo: not directly after, no
[01:27] <mdz> I did one perhaps an hour ago
[01:28] <elmo> example?
[01:28] <mdz>  o libosmesa4-xorg libosmesa4-xorg-dbg libosmesa4-xorg-dev              {xorg}
[01:28] <mdz>    [Reverse-Depends: Supported seed, libosmesa4-xorg-dev] 
[01:28] <mdz>  o libebook1.2-3                                       {evolution-data-server}
[01:28] <mdz>    [Reverse-Depends: contact-lookup-applet] 
[01:28] <elmo> yeah, that will show up
[01:28] <elmo> she doesn't know to check the database to see if the binaries still exist
[01:28] <mdz> is that because you ran rene after I ran cron.sync?
[01:28] <elmo> ("SQL is hard, let's go shopping")
[01:28] <elmo> yeah
[01:28] <mdz> ok
[01:29] <elmo> (I could fix it, but I'm lazy and just keep rerunning cron.sync)
[01:30] <Nafallo> wow! more and more proof those are real women. elmo quotes them :-P.
[01:37] <Keybuk> Nafallo: all of the dak suite are named after someone in particular
[01:40] <Nafallo> Keybuk: yay! I thought that was just a rumour :-).
[01:46] <aroman> k that took longer than 5 minutes...
[01:47] <aroman> so.. back to my error...
[01:47] <aroman> Xorg with i810 driver and i915 kernel module
[01:47] <aroman> open("/dev/tty". O_RDWR|O_NONBLOCK|O_NOCTTY) =  -1 ENXIO (No such device or address)
[01:48] <aroman> this is a line of sudo strace X
[01:48] <aroman> after that I get *** glibc detected *** double free or corruption (fasttop): 0x<address> ***
[01:49] <aroman> why am I not able to open /dev/tty? it's permissions are 666, not to mention the fact I am running X as root
[01:54] <aroman> what's interesting, is that using the vesa module, it succeeds in opening "/dev/tty", O_RDWR
[01:54] <aroman> there must be something related to i810 (xorg), i915 (kernel drm) and xorg
[01:54] <mdz> the tty error is most likely unrelated to your problem
[01:55] <aroman> mdz, it's what I figured by now, but after that error, X just dies without any error in the log
[01:56] <aroman> I DO get a bunch of vm86old(<address>) = -1 ENOSYS (Function not implemented)
[01:57] <mdz> aroman: install the debug version of the X server, run it under gdb, and get a backtrace from the crash
[01:57] <aroman> mdz, apt-get install xorg-debug? 
[01:58] <aroman> xserver-xorg-dbg
[02:02] <aroman> total success lol
[02:02] <aroman> I ran sudo gdb Xorg-dbg and that froze the machine
[02:13] <jbailey> Hmm, The right apple key is called "Right win key" in gnome.  That's probably not right. =)
[02:17] <carstenh> iirc it it mapped to the same keycode, so gnome could not know this :)
[03:16] <jbailey> carstenh: Right.  Perhaps my default locale should be en_CA@Mac =)
[03:17] <carstenh> :)
[03:20] <jdub> mjg59: remote possibility of ping?
[03:20] <mjg59> jdub: Hi
[03:20] <mjg59> (Remote but answered)
[03:20] <jdub> holy cow! :)
[03:20] <jdub> so i'm diddling with usplash
[03:20] <mjg59> Yup
[03:20] <jdub> and helping cliff out a bit doing artwork
[03:21] <jdub> why are we restricted to 12 colours?
[03:21] <jdub> i would like 16 colours
[03:21] <jdub> and a pony
[03:21] <mjg59> You can have 16 colours
[03:21] <mjg59> As long as one is black, one is red and one is green
[03:21] <mjg59> And we have one to use as a progress bar
[03:21] <mjg59> And one to use as a background for the text
[03:22] <jdub> ok, your image had 12 colours
[03:22] <mjg59> Black is negotiable - it can be any colour, as long as it contrasts with the text background colour and doesn't look suck
[03:22] <mjg59> My image is arbitrary and shit
[03:22] <jdub> and i just tried a 16 colour image - it looked horrendous
[03:22] <mjg59> The code may need some hacking to work with usplash
[03:22] <mjg59> Basically, get me a picture that looks OK and I'll make usplash work with it
[03:23] <mjg59> I don't entirely understand the BOGL code, and there's no documentation
[03:23] <jdub> which palette indices do the black/red/green/progress/textbg colours have to be?
[03:23] <mjg59> They don't
[03:23] <mjg59> I can sort those
[03:23] <mjg59> But it helps if they're in a block
[03:23] <jdub> hrm
[03:23] <jdub> can you make specific indices for them?
[03:24] <mjg59> If you'd like
[03:24] <jdub> so we can muck with images without having to change the code?
[03:24] <mjg59> Pick numbers, I'll make the code work
[03:24] <jdub> what are the red and green used for, btw?
[03:24] <mjg59> Success and failure in the output messages
[03:24] <mjg59> Green is arguable, we probably need red
[03:24] <jdub> how will the progress bar be used?
[03:25] <jdub> like winxp, arbitrary cycling?
[03:25] <mjg59> That's an interesting question, and one that I will answer in the fullness of time
[03:25] <jdub> heh
[03:25] <aroman> I say we should do something like the original bootsplash could do
[03:25] <mjg59> aroman: Which is?
[03:26] <aroman> ie. for each service started, display a message, well not really... more like.. starting network, starting hotplug, etc. and progress until all the services are started. I would like to see the progress bar fill up so that at 100%, the system is up and running
[03:26] <mjg59> aroman: That's possible
[03:26] <aroman> mjg59, and I would be interested to work on it..
[03:26] <jdub> but you will never see 100%
[03:26] <mjg59> But Ubuntu continues starting services after gdm starts, which means that the graphical splash will have vanished
[03:26] <jdub> because gdm will start :-)
[03:27] <jdub> haha
[03:27] <mjg59> Basically:
[03:27] <mjg59> If we want it to go to 100%, the speed at which it increases will not be constant
[03:27] <aroman> if I could solve this damn i810 bug! :(
[03:27] <mjg59> Which is a pain
[03:27] <mjg59> aroman: Have you checked Bugzilla for an existing bug?
[03:27] <mdz> infinity: so, you and daniels have created much amusement for Keybuk and myself
[03:28] <aroman> mjg59, yeah.. it's there, no answers... apparently this bug has been since 22 build of Xorg
[03:28] <mjg59> aroman: Which bug number?
[03:28] <aroman> h/o
[03:28] <aroman> http://bugzilla.ubuntu.com/show_bug.cgi?id=12418
[03:28] <aroman> so 12418
[03:29] <jdub> mjg59: the colour requirement you skipped was text area foreground colour, right?
[03:29] <aroman> so anyways, with usplash... I mean we could make the progress bar fill up to 100, where 100 means gdm started, or if gdm is not enabled, it means you can now log in...
[03:30] <mjg59> jdub: That was what I wanted black for
[03:30] <jdub> mjg59: oh
[03:30] <mjg59> jdub: Otherwise pick an arbitrary colour
[03:30] <mjg59> Basically, I need:
[03:30] <jdub> so the colour roles are:
[03:30] <mjg59> 1) Something to write with
[03:30] <mjg59> 2) Something to write on
[03:30] <mjg59> 3) Something to indicate failure
[03:30] <Keybuk> we don't need green
[03:30] <aroman> I like green...
[03:30] <mjg59> 4) Something to indicate success (may be the same as (1))
[03:30] <Keybuk> having green to indicate success, and red to indicate failure
[03:31] <mjg59> 5) Something to draw a progress bar with (may be the same as one of the above)
[03:31] <Keybuk> when the saturation and value of the green and red are exactly the same
[03:31] <Keybuk> IS A FUCKING STUPID IDEA
[03:31] <desrt> failure = bold
[03:31] <mjg59> It's acceptable for these colours to be used within the artwork as well
[03:31] <Keybuk> colour is useful to go "LOOK AT ME!"
[03:31] <mjg59> aroman: Are you using kubuntu or ubuntu?
[03:31] <Keybuk> so white for everything except red for failure makes the failures stand out
[03:31] <aroman> mjg59, ubuntu
[03:31] <desrt> i'd like to ask everyone: isn't gentoo great?
[03:31] <mjg59> aroman: Ok, fine
[03:32] <mjg59> aroman: Just checking - that bug report is against the wrong package, then. I'll fix it now.
[03:32] <aroman> mjg59, that bug isn't in kdm, it's in X/i810/i915, somewhere there
[03:32] <Keybuk> we did this argument a year ago, and left nathaniel at the bottom of a lake, let's not do it again ;)
[03:32] <lamont> desrt: wrong channel.  then again, #ubuntu is the wrong channel for that too
[03:32] <jdub> Keybuk: and yet we still got lumped with blingstart
[03:32] <desrt> lamont; totally on topic.  we're talking about flashy colours on startup.
[03:33] <lamont> meh
[03:33] <Keybuk> jdub: ya know, isn't it almost a year to the day since Oxford?
[03:33] <jdub> is it?
[03:33] <Keybuk> and nathaniel claimed he'd have it working by the end of that conf
[03:33] <aroman> yes, gentoo does have pretty nice splash support
[03:33] <mjg59> Pretty much
[03:33] <Keybuk> when was oxford?
[03:33] <mjg59> August last year
[03:33] <Keybuk> 9th-20th august
[03:33] <jdub> golly
[03:34] <lamont> are we nearly there yet?
[03:34] <mjg59> So I've been hacking on laptops for you for about a year
[03:34] <Keybuk> http://www.netsplit.com/events/2004/canonical-oxford/canonical-oxford-018.html
[03:34] <jdub> that feels so long ago but so recent all at once
[03:34] <mjg59> What am I doing with my life?
[03:34] <Keybuk> THIS MAN FAILED TO BRING YOU USPLASH
[03:34] <lamont> jdub: definitely
[03:35] <mjg59> Hurrah for me, eh?
[03:36] <lamont> mjg59: you totally rock
[03:36] <aroman> hurrah for mjg59 !!
[03:36] <lamont> huzzah!
[03:36] <aroman> so it seems the only way to get ethernet on my laptop is to apply the syskonnect patch and re-compile the kernel..
[03:37] <aroman> the module just won't build by itself..
[03:37] <jdub> mjg59: ok, i just flat out don't believe that we need the text output :-)
[03:37] <HrdwrBoB> aroman: recompile it then just shove the module in the right place
[03:37] <jdub> mjg59: and it is a waste of my colours!
[03:37] <mjg59> aroman: File a bug against the kernel, we'll sort it for release
[03:37] <mjg59> jdub: Acknowledged
[03:37] <mjg59> jdub: (And ignored. Screw you, hippy!)
[03:37] <jdub> heh
[03:38] <jdub> i think this is worth nutting out, though
[03:38] <mjg59> http://www.netsplit.com/events/2004/canonical-oxford/ - weren't we young and innocent then?
[03:38] <mjg59> jdub: I think that it's a bit close to release to change the startup experience entirely
[03:38] <mjg59> jdub: I'd go with weaning people off in two stages
[03:38] <jdub> oh man
[03:39] <jdub> you manipulative evil man
[03:39] <mjg59> First we give them graphics
[03:39] <mjg59> Then we take away their text
[03:39] <jdub> using sensible release sensitive answers to blow away my usability/sass concerns
[03:40] <mjg59> http://www.netsplit.com/events/2004/canonical-oxford/canonical-oxford-028.jpg - I have no recollection as to why that flip chart says "Moo"
[03:40] <jdub> that's infuriating and wonderful all at the same time
[03:40] <mjg59> Maybe that was the night before the Antitrust drinking experience
[03:40] <elmo> mjg59: I probably wrote it
[03:42] <mdz> it looks like elmo's handwriting
[03:42] <elmo> my handwriting isn't readable
[03:42] <elmo> but I tend to write Moo on things like flipcharts as I'm startlingly unoriginal
[03:43] <Keybuk> mjg59: the scarier ones are the pictures from the original London meeting
[03:43] <Keybuk> mdz had hair, and thom looks practically pubescent
[03:45] <mdz> Keybuk: where are those?
[03:46] <Lathiat> ahh, yay for rebuilds, 600M mirror sync this morning. :)
[03:46] <mdz> Keybuk: you not only took all of 5 photos at debconf5, you somehow managed to stay out of everyone else's photos too
[03:46] <Keybuk> mdz: lamont had some
[03:46] <mdz> oh, you have photos of the mataro tv press conference
[03:46] <Keybuk> lamont: bring out the London photos
[03:46] <mdz> that certainly was amusing
[03:47] <mdz> Keybuk: is your camera's flash bright orange or something?
[03:48] <lamont> Keybuk: I'll drag out the lot sometime soon-ish - maybe some nice person will go through and catalog them for me.
[03:49] <Keybuk> at least find the one with mdz-with-hair and the thom one :p
[03:49] <lamont> the thom one?
[03:49] <elmo> http://people.ubuntu.com/~james/tmp/
[03:49] <aroman> also, on 2.6.12-6-386 vga=791 does not work anymore I get a blank screen
[03:49] <aroman> has there been a transition from the default vesafb driver?
[03:49] <Keybuk> mdz: dunno, but the ccd does tend vaguely towards red
[03:49] <Keybuk> I think I'm supposed to colour-correct it afterwards, but gimp sucks at that
[03:50] <Keybuk> elmo: sweeet
[03:50] <mdz> aroman: usplash doesn't support vesafb
[03:50] <aroman> mdz, even with splash disabled vga=791 gives me a blank screen
[03:51] <mdz> aroman: then don't do that ;-)
[03:51] <mjg59> mdz: Though it shouldn't give a blank screen
[03:51] <mjg59> vesafb ought to load instead
[03:51] <aroman> but I want 1024x768 in my console :P
[03:51] <mdz> elmo: thom looks pretty disgusted with what lamont just said
[03:51] <Lathiat> then don't want? ;)
[03:51] <aroman> Lathiat, no can do!
[03:51] <aroman> :P
[03:51] <mdz> and yet Keybuk is very amused
[03:51] <elmo> mdz: yeah, but scott looks pretty happy
[03:52] <mdz> that's classic
[03:52] <tseng> jeff looks totally 70s
[03:53] <Lathiat> with his toilet^H^H^H^H^H^H laptop?
[03:53] <tseng> the hair
[03:53] <mjg59> Why is that picture so speckly?
[03:54] <elmo> not sure; the rest of the set aren't
[03:54] <mjg59> mdz looks pretty 70s
[03:55] <mdz> I have michael j. fox hair in that photo
[03:55] <mjg59> It's as if you're about to start a porn film, or something
[03:55] <Keybuk> mdz: but the same sweater
[03:55] <whiprush_> who is that to the left of Keybuk 
[03:55] <mdz> lamont
[03:55] <jsgotangco> morning
[03:56] <tseng> whiprush_: the backpack is almost certainly stuff with lamont-gear
[03:56] <mjg59> That backpack contains equipment to set fire to a shed and put it out again
[03:56] <mjg59> Goddamnit, I want it to be winter
[03:56] <tseng> i hope those glasses dont entirely belong to jeff
[03:57] <elmo> mjg59: amen
[03:57] <elmo> this weather is the suck
[03:59] <mdz> mjg59: winter?  why?
[03:59] <mjg59> Winter is more attractive
[03:59] <mjg59> And I'm sick of being too hot
[03:59] <mdz> UK weather just cycles between different flavours of miserable
[03:59] <jsgotangco> lol
[04:00] <mjg59> mdz: Bah. You should visit Cambridge in winter.
[04:00] <mdz> I was in cambridge at this time of year, and it was fairly pleasant
[04:00] <mdz> considering
[04:00] <mjg59> Yeah
[04:00] <mjg59> Come here in 5 months
[04:01] <Keybuk> it was raining earlier this evening on way back from lug
[04:01] <Keybuk> it was quite pleasant, as it was still warm
[04:01] <mdz> lamont: where are these alleged london photos of yours?
[04:02] <lamont> mdz: on my desktop machine...
[04:05] <aroman> mjg59, ok I've attached some files to the i810 bug
[04:06] <aroman> output of startx, xorg.conf and Xorg.0.log
[04:09] <mjg59> aroman: Cool, thanks
[04:09] <lamont> mdz: some of the photos copying to people.u.c/~lamont/pictures
[04:10] <lamont> note that it'll take a while to copy them.
[04:10] <lamont> included are the wonderful escher-esuqe trip through the hotel from that april-in-london meeting, etc.
[04:10] <aroman> mjg59, no problem.. I want to get involved in solving this bug... 
[04:10] <mjg59> aroman: That's the entirity of the Xorg.0.log?
[04:11] <lamont> ew.  42 pics at ~ 10 min /pic...  there's gonna be a bit of a wait.
[04:11] <aroman> mjg59, yup
[04:11] <mjg59> Looks like it's crashing on BIOS interaction
[04:11] <mjg59> Weird
[04:11] <aroman> it's crashing somewhere... but it's not crashing in a normal fashion...
[04:11] <mjg59> aroman: Ok, looks solidly like a driver bug
[04:11] <aroman> because that glibc warning/error doesn't go in the log
[04:11] <aroman> just in the console
[04:12] <aroman> so X dies before it gets the chance to print the message in the log... I think
[04:13] <mjg59> "Use the OS setup option ACPI=Disabled at the initial OS setup screen.
[04:13] <mjg59> Failure to do so may cause system damage."
[04:13] <mjg59> Yes. That would be because YOU'VE BROKEN THE FAN CONTROL CODE IN THE DSDT.
[04:14] <Keybuk> mjg59: who's that?
[04:14] <mjg59> HP
[04:14] <Keybuk> heh
[04:14] <mjg59> It's from their official document on how to install NLD on their business range
[04:15] <Keybuk> odd, fan control works pretty well on mine
[04:16] <mjg59> Yeah, it's the newer ones
[04:17] <Keybuk> so HP have forgotten how to make good laptops
[04:17] <Keybuk> Dell never knew how
[04:17] <Keybuk> and IBM got bored and sold Thinkpad
[04:17] <Keybuk> who do we buy laptops off now? :-/
[04:17] <jdub> lenovo
[04:18] <jdub> for all your thinkpad needs
[04:18] <Lathiat> mjg59: haha, nice
[04:20] <LaserJock> mdz: yesterday I asked about problems I had installing Colony 3 where the install would stop at "downloading 4 of 8 (0s remaining)" or something like that
[04:21] <daniels>    * x11proto-gl-dev inexplicably declared Replaces: libglu1-mesa-dev, which
[04:21] <daniels>      caused VERY CONFUSING THINGS to happen
[04:21] <daniels> ere
[04:21] <daniels> mdz: inexplicably -> they both shipped the same file, and that's what normally happens?
[04:21] <desrt> word.
[04:21] <desrt> Replaces: is quite nice for not having to manually uninstall the old one yourself
[04:22] <jlj> infinity: what's the status of new network-manager package?
[04:22] <LaserJock> well today I checked out my ram and it one of my sticks had errors when I did a memtest
[04:22] <LaserJock> so I took that stick out and installed colony3 using the other one and everything went fine
[04:24] <LaserJock> the wierd thing is that I have installed other distros and Hoary on the same machine and everything was fine
[04:24] <aroman> mjg59, I've also created the kernel bug entry - sk98lin doesn't compile
[04:24] <aroman> mjg59,  bugid: 13793
[04:24] <LaserJock> the only time I have had any problems is with the 081405 snapshot and Colony3
[04:24] <Keybuk> daniels: x11proto-gl-dev -3 provided glu.h and Replaces: libglu1-mesa-dev
[04:25] <Keybuk> libglu1-mesa-dev gets glu.h, and gets upgraded; but because of the replaces, x11proto-gl-dev still owns the file
[04:25] <Keybuk> x11proto-gl-dev gets upgraded to -4, which doesn't ship the file, so glu.h goes bye-bye
[04:29] <daniels> Keybuk: the Replaces also disappeared in -4
[04:30] <Keybuk> but it's too late by that point
[04:30] <Keybuk> because libglu1-mesa-dev didn't conflict x11proto-gl-dev, there was no guarantee of the order they'd get installed
[04:30] <Keybuk> so if libglu1-mesa-dev got upgraded first (which was happening to people) then they lost glu.h
[04:31] <daniels> cute
[04:36] <sbalneav> Evening all.
[04:37] <aroman> all right, I'm off for tonight
[04:37] <aroman> good night people!
[04:37] <aroman> until tomorrow
[04:38] <Keybuk> yay, Braniac is coming back next week
[04:42] <mdz> daniels: the files moved from x11proto-gl-dev to libglu1-mesa-dev
[04:42] <mdz> daniels: therefore libglu1-mesa-dev should have replaced x11proto-gl-dev, not the other way around
[04:44] <Keybuk> infinity did the right thing to the wrong thing
[04:53] <jdub> http://www.schneier.com/blog/archives/2005/08/new_cryptanalyt.html
[05:02] <infinity> Keybuk : I didn't "do the wrong thing to the wrong thing", we just forgot to swap it back when the canonical file moved around (again)
[05:02] <Keybuk> heh, fair enough
[05:02] <infinity> Thankfully, those pesky headers should have stopped moving by now. :)
[05:02] <Keybuk> uh, huh; because everyone lost them <g>
[05:03] <infinity> They'll magically get them back on the next upgrades. :)
[05:03] <infinity> That's the joy of losing non-conffiles.
[05:03] <infinity> They just.. Reappear! :)
[05:03] <daniels> the idea was that x11proto-gl would provide the canonical copies of glx.h and glu.h
[05:03] <daniels> that decision later got reversed by upstream
[05:03] <daniels> *shrug*
[05:04] <Keybuk> it was fun figuring out what happened
[05:04] <infinity> Yeah, I've done that to myself once before.  But I think it also involved other scary things, like directory->symlink migration on top of backwards replaces.
[05:04] <infinity> And someone tapping on my forehead with a pencil while I tried to figure it out.
[05:05] <Keybuk> not quite Swordfish
[05:05] <Lathiat> heh
[05:05] <infinity> No, nothing is as bad as being forced to watch Swordfish while figuring out a problem.
[05:05] <infinity> (Or at all)
[05:06] <Keybuk> it's not that bad, it scores 8 on a scale of 1 to Antitrust
[05:06] <daniels> infinity: antitrust is worse than swordfish.
[05:06] <Keybuk> jinx
[05:06] <daniels> you can't jinx someone because you expressed the same sentiment
[05:06] <Keybuk> can
[05:11] <mdz> mjg59: I am no longer able to get my T42 to crash no matter how much power.sh crack I enable
[05:11] <mdz> infinity: are you saying that it moved once before that?  I didn't think libglu1-mesa-dev existed until more recently
[05:12] <daniels> mdz: libglu1-mesa-dev has existed since the dawn of time
[05:12] <daniels> mdz: in the beginning, libglu1-mesa-dev and xlibmesa-glu-dev provided glu.h (note that this is not libgl1, or glx.h)
[05:13] <daniels> then, when we changed xlibmesa*, it was libglu1-mesa-dev and libglu1-xorg-dev providing it
[05:13] <daniels> then upstream said that it belonged in the proto/GL module
[05:13] <daniels> so x11proto-gl-dev got it, and Replaced those two
[05:13] <daniels> then we killed off libglu1-xorg-dev, so it was just x11proto-gl-dev and libglu1-mesa-dev
[05:13] <daniels> then we had to reverse that, so x11proto-gl-dev no longer shipped it, and the canonical copy lived in libglu1-mesa-dev
[05:14] <rob^> has the graphical installer has been deferred until after breezy?
[05:14] <mdz> this is why replaces should be versioned
[05:14] <rob^> great english skills today..
[05:14] <mdz> rob^: yes
[05:15] <rob^> mdz, ok thanks
[05:19] <infinity> mdz : Yes, obviously.  It would have become a versioned replaces as soon as the copy left libglu1-mesa-dev.. But then it never did, cause it moved back there.  Queue ominous music.  Oh well, it's all sorted now.
[05:20] <mdz> infinity: yes, I think so.  revving libglu1-mesa-dev should hopefully also fix things for people who lost the file
[05:22] <mdz> infinity: to me, not knowing that the file had already moved once, it looked like you preemptively added a replaces for the place where the file was going to go in the future :-)
[05:24] <infinity> mdz : yeah, it's a bit muddy. :)
[05:24] <mdz> hence "inexplicably"
[05:24] <daniels> this one was sort of handwaving-through-the-clouds-of-bong-smoke
[05:30] <infinity>  /build/buildd/pike7.6-7.6.27/src/interpret.c:1287: error: PIC register 'ebx' clobbered in 'asm'
[05:31] <elmo> infinity: any reason not to do a test-rebuild of main again soon?
[05:32] <infinity> elmo : None that I can see, though if we're even going to get new kernels on the PPC buildds, -autotest would be a nice way to stress-test them.
[05:32] <elmo> oh, meh, right
[05:32] <elmo> and there's the small matter of breezy's amd64 kernel lameness
[05:32] <infinity> elmo : As I told you before though, I'd have no issues with you scrapping the w-b databse for -autotest every two or three days and letting it start from scratch.
[05:33] <infinity> I could do it too, but walking the whole DB to --forget everything is a bit inefficient. :)
[05:33] <infinity> (Oh, wait, that doesn't work anyway, if katie thinks the autotest binaries are installed somewhere)
[05:34] <daniels> infinity: so the pike people have to write better inline asm
[05:34] <infinity> daniels : Evidently, but it's not failed before.
[05:34] <infinity> daniels : So, uhh.  Joy.
[05:34] <daniels> \o/
[05:34] <infinity> Also, the pike build system is crazy non-deterministic fun.  If a file fails to compile, it magically drops optimisation and tries again.
[05:35] <elmo> hahahahahaha
[05:35] <daniels> ... sweet mother of christ.
[05:35] <elmo> christ, why do we have that madness in main anyway?
[05:35] <daniels> i should do that for mesa
[05:35] <infinity> Ellifiknow.
[05:35] <daniels> the crazy per-arch optimised stuff (it does a normal build followed by a gentoo build which goes in /usr/lib/i686/...) should do that
[05:36] <daniels> start at -O9 -funroll-loops -funroll-all-loops -march=p4 -mtune=p4 -fomit-frame-pointer, etc
[05:36] <daniels> then run the mesa regression test suite against it
[05:36] <daniels> and step down until we get something that passes
[05:37] <infinity> You'd be more likely to trip on miscompilations than ICEs, leaving you with compiled -- but useless -- binaries.
[05:37] <daniels> infinity: hence the mesa regression test suite
[05:37] <infinity> Oh, wait.  I missed the part where you mentioend the testsuite.
[05:37] <infinity> Yes.
[05:37] <infinity> You scare me.  Stop now.
[05:37] <daniels> made.  of.  awesome.
[05:38] <daniels> (but hey, who needs stupid options when gcc goes 'oh, hey, volatile accesses! we can optimise those out, right? no-one needs those.' anyway)
[05:38] <infinity> You're not bitter, are you?
[05:38] <daniels> noooooooo.
[05:38] <daniels> not at all.
[05:39] <daniels> i don't hate my users, individually and collectively.
[05:39] <elmo> for swig apparently
[05:39] <elmo> go swig
[05:40] <Lathiat> daniels: Can you compile xorg with -ffast-math ?
[05:40] <infinity> Oh, this isn't new.  pike has been failing on i386... Since... Since the binutils_2.16 upgrade, it looks like.
[05:40] <elmo> dude talk to the doko
[05:41] <infinity> elmo : We could disable swig's ability to generate pike bindings.  Like anyone really uses that feature anyway.
[05:41] <infinity> I'd be willing to bet that 99% of swig usage is for generating python bindings.
[05:41] <elmo> well that's just the first hit; let's see what melanie thinks
[05:41] <infinity> Just a guess.
[05:42] <elmo> hmm, it really is only swig
[05:42] <infinity> Yeah, swig pulls in every interpreter it's capable of generating bindings for.
[05:42] <daniels> Lathiat: you ... aren't serious ... are you?
[05:42] <Lathiat> daniels: of course i am!
[05:43] <infinity> Many/most of which we want in main anyway, but I see no argument for supporting pike.
[05:43] <daniels> infinity: i bet the latest hj lu binutils will fix this problem
[05:43] <daniels> wooooooooo choking on the crackpipe
[05:43] <infinity> Given that we don't support pike's main use cases (roxen/caudium), supporting pike itself is a bit pointless.
[05:51] <Lathiat> heh whys that
[05:52] <infinity> Or.. Just kdebase.  That's not so bad.
[06:09] <fabbione> morning
[06:09] <AndyFitz> g'day fabbione
[06:13] <fabbione> mdz: ubuntu-meta is up
[06:14] <mdz> fabbione: thanks
[06:15] <fabbione> no problem
[06:31] <AndyFitz> ogra, reckon we could get ubuntu-title.ttf packaged ?
[06:39] <fabbione> mdz, elmo: who is going to be around at 12:00/13:00 UTC? 
[06:39] <mdz> fabbione: the smart money is on elmo
[06:39] <mdz> that's 0500/0600 for me
[06:39] <fabbione> elmo: ?
[06:40] <fabbione> mdz: i will need somebody with NEW power ;)
[06:47] <Treenaks> is there a way to create striked-through text?
[06:47] <mdz> daniels: so it's that point in the release cycle where I try to get my keyboard to work right again
[06:48] <mdz> daniels: my left alt key works as a perfectly normal modifier, but my right alt key has an identity disorder
[06:48] <daniels> mdz: ok, here's the two-step procedure
[06:48] <daniels> step one: throw keyboard out window
[06:48] <daniels> step two: purchase sensible keyboard
[06:48] <mdz> there is nothing wrong with my hardware
[06:48] <daniels> mdz: oh, it's still doing the whole level 3 thing?
[06:48] <mdz> it works well under Linspire
[06:48] <infinity> s/tittler/titters/
[06:48] <mdz> daniels: no, it's weirder than that
[06:49] <daniels> mdz: you know, I *did* have respect for you
[06:49] <mdz> daniels: so I press left_alt+tab, and I get the GNOME window switching dialog
[06:49] <mdz> daniels: then I release alt, and the dialog stays up
[06:49] <mdz> until I press another key, at which point it goes away, and leaves me in the same window where I started
[06:49] <daniels> mdz: awesome!
[06:49] <mdz> regardless of whether I tabbed over to another window
[06:50] <mdz> this is with setxkbmap -variant dvorak us
[06:50] <Lathiat> mdz: i get the same thing
[06:50] <Lathiat> mdz: you sure its not a gnome thing?
[06:50] <Lathiat> like the same thign happens with ctrl+alt+arrow
[06:50] <mdz> yes, it does
[06:50] <mdz> but it seems unlikely to be a gnome thing, to me
[06:50] <daniels> mdz: wfm
[06:50] <Lathiat> i kinda find it handy
[06:50] <mdz> Lathiat: what keyboard configuration do you use?
[06:50] <Lathiat> can let go of alt
[06:50] <Lathiat> and scroll back and forwards
[06:50] <Lathiat> mdz: standard us
[06:50] <daniels> mdz: regardless of us or us(dvorak)
[06:51] <mdz> daniels: _really_?
[06:51] <mdz> it's like right alt is sticky
[06:51] <Lathiat> yeh
[06:51] <Lathiat> its quite possibly a gnome thing
[06:51] <daniels> mdz: er, only with right alt?
[06:51] <daniels> mdz: '05:49 < mdz> daniels: so I press left_alt+tab, and I get the GNOME window switching dialog
[06:51] <Lathiat> since it can tell the difference between the two
[06:51] <mdz> daniels: correct
[06:51] <Lathiat> e.g. i can set mmy second alt to be a compose key
[06:51] <Lathiat> or whatever
[06:51] <mdz> daniels: sorry, s/left/right/
[06:53] <daniels> ok, happens to me too
[06:53] <daniels> i guess it's a metacity thing, since xkb appears to be doing the right thing in this case
[06:54] <daniels> mdz: setxkbmap -option compose:ralt
[06:54] <Lathiat> not sticky in compose mode
[06:54] <Lathiat> 
[06:54] <mdz> daniels: that has made my right alt key into a compose key
[06:54] <infinity> mdz : Permission to break UVF for pike7.6 (7.6.27 -> 7.6.33) to fix the FTBFS?  (pike7.6 has no dependencies in main except for swig, which I'll rebuild and make sure it's okay)
[06:54] <daniels> Lathiat: i temporarily disabled compose
[06:54] <mdz> daniels: which means that it no longer exhibits the sticky behaviour, but...it's a compose key
[06:54] <mdz> infinity: ok
[06:54] <daniels> mdz: well, yeah.  problem solved!
[06:55] <daniels> mdz: and you gain the ability to talk to people named sren, kster, etc
[06:55] <mdz> daniels: it happens to you with a us layout too?
[06:55] <infinity> Is there value in that?
[06:55] <infinity> ;)
[06:55] <Lathiat> and go ?
[06:55] <Lathiat> err,  :)
[06:55] <mdz> k
[06:56] <Treenaks> mdz: I have a US layout
[06:56] <daniels> yarr, metacity's code is crack
[06:56] <daniels> mdz: yeah, once I disable compose
[06:56] <Treenaks> mdz: what should the problem be?
[06:56] <mdz> Treenaks: the problem is that my right alt key behaves differently from my left alt key
[06:56] <Treenaks> mdz: right-alt+tab is weirdly sticky here too
[06:57] <Treenaks> (during alt-tabbing at least)
[06:57] <mdz> I consider this a bug and a regression
[06:57] <daniels> mdz: yes
[06:57] <Treenaks> mdz: I never use my ralt, but yeah, this behaviour is sucky
[06:57] <Lathiat> i dont think anyone uses their right alt
[06:57] <daniels> mdz: i *think* it's metacity's code, which I'm looking at the moment
[06:57] <Treenaks> (compose=menu here :))
[06:57] <daniels> cue bleeding eyes
[06:57] <Lathiat> except when single handly changing virtual desktops
[06:57] <mdz> daniels: haven't found anything in the metacity changelog thus far
[06:57] <daniels> Treenaks: pc101 4 lyf
[06:58] <mdz> when does seb128 wake up?
[06:58] <mdz> Lathiat: I do
[06:58] <daniels> mdz: it's 0658, so probably not any time soon
[06:58] <Lathiat> 
[06:58] <Lathiat> hrm
[06:58] <Lathiat> whats the ctrl+shift+<no> for yarr
[06:58] <Treenaks> daniels: pc102 actually, but yeah :)
[06:58] <mdz> daniels: how do I undo compose:ralt?
[06:58] <Treenaks> 104 even
[06:59] <infinity> -options ''
[06:59] <Lathiat> mdz: setxkbmap -option ralt:alt
[06:59] <mdz> ah, setxbkmap -option ''
[06:59] <Lathiat> or that
[06:59] <infinity> s/options/option/ yes.
[06:59] <mdz> the man page beat infinity by a split second
[07:01] <daniels> Treenaks: i only have 80 or thereabouts physical keys
[07:02] <daniels> Treenaks: pc104 is pc101 + windows + menu + windows.  which I don't have on this laptop.
[07:02] <Lathiat> i have just one windows key
[07:02] <mdz> yay IBM for omitting the windows keys
[07:03] <daniels> mdz: exactly, it lets you have a compose key on your ralt and still have no useless keys -- result!
[07:03] <Lathiat> i have 87 keys
[07:03] <Lathiat> and 7 'media' keys
[07:04] <Lathiat> altho there is a virtual numap
[07:04] <daniels> hm, metacity has broken b-ds as well.  joy.
[07:04] <Lathiat> so taht adds 14
[07:04] <Lathiat> so, 101. :)
[07:04] <mdz> I wouldn't mind having my right windows key be a compose key
[07:04] <Lathiat> mdz: yeh, thats what i'd use as compose if i had one
[07:04] <Lathiat> i suppose using my left windows key isnt such a bad idea
[07:04] <daniels> mdz: -option compose:rwin
[07:05] <mdz> -option compose:rwin doesn't seem to do that
[07:05] <daniels> mdz: or compose:menu for the menu one
[07:05] <mdz> nor does compose:menu do the expected thing
[07:06] <daniels> fun
[07:06] <daniels> seems to wfm on standard pc104/us
[07:06] <mdz> I suppose this al works nicely on us layouts
[07:06] <daniels> probably
[07:06] <daniels> although dvorak shouldn't be modifying that
[07:06] <daniels> anyway, time for breakfast
[07:06] <Lathiat>  when i open the keyboard capplet.. and the font one.. i get some windowwith a warning title open but goes away to quick to see whats in it, any idea what it is?
[07:06] <mdz> I should investigate dinner at some point
[07:07] <mdz> Lathiat: I got that too, with the font capplet
[07:07] <Lathiat> mdz: yeh, my keyboard one does it too, has for quite a while and its annoying
[07:07] <Treenaks> daniels: I plugged an external keyboard into my laptop, because the keyboard on it sucks so hard
[07:08] <Lathiat> i want to get an external keyboard so i can raise my screen up to me eye height
[07:08] <Treenaks> daniels: (and it's actually a 110-key board, counting the multimedia cruft)
[07:08] <Lathiat> but some stupid ebay seller hasnt shipped by usb->ps/2 convertor yet
[07:08] <daniels> mdz: i386?
[07:08] <Lathiat> mdz: indeed
[07:08] <mdz> daniels: yes
[07:09] <fabbione> mdz: xev is on my list of 2353 mini pkgs to do
[07:09] <fabbione> mdz: probably today will be in
[07:09] <daniels> mdz: http://amnesiac.heapspace.net/~daniels/misc/xev
[07:10] <mdz> daniels: with -option compose:rwin, my rwin key becomes some sort of dead key, but it doesn't do anything with 2 subsequent keypresses
[07:10] <mdz> it wants three
[07:10] <mdz> rwin+'+o produces three 'o's
[07:11] <daniels> ffs, heisenbug
[07:11] <mdz> 100% reproducible
[07:11] <daniels> when I compile metacity with a bit of debugging information in, it's sweet
[07:11] <Lathiat> haha
[07:11] <mdz> does it fix the alt key problem?
[07:11] <Lathiat> leave it in then :)
[07:11] <daniels> mdz: yes
[07:11] <mdz> awesome
[07:12] <daniels> THE SAME VERSION THAT'S IN THE THINGY
[07:12] <daniels> BUT WITH ONE EXTRA PRINTF
[07:12] <daniels> unless sebarino sucks
[07:12] <daniels> and metacity doesn't support xkb
[07:12] <mdz> compose:lwin doesn't work at all
[07:12] <mdz> compose:menu has the same problem as compose:rwin
[07:13] <daniels> right
[07:14] <mdz> -option compose:rwin doesn't work for me with a us layout either
[07:14] <mdz> is it only me?
[07:14] <mdz> my X server hasnt' been restarted in a long time
[07:14] <daniels> daniels@ephemera:~/canonical/metacity% grep Xkb =metacity; sudo dpkg -i metacity_2.11.2-0ubuntu1_i386.deb > /dev/null && grep Xkb =metacity
[07:14] <daniels> zsh: exit 1     grep Xkb =metacity
[07:14] <daniels> Binary file /usr/bin/metacity matches
[07:14] <daniels> the former being the archive version, the latter being my version
[07:14] <mdz> root      5205  2.6 11.6 334556 241184 ?       R    Jul25 928:20 /usr/X11R6/bin/X :0 -audit 0 -auth /var/lib/gdm/:0.Xauth vt7
[07:14] <daniels> mdz: i don't have a windows key, else I'd help
[07:14] <mdz> I'm going to restart my X server
[07:14] <mdz> hopefully I have a working xorg.conf
[07:15] <Lathiat> famous last words
[07:16] <daniels> ehm.  metacity's configure.in is on crack.
[07:17] <Lathiat> surprise!
[07:17] <mdz> YES
[07:17] <mdz> daniels: /etc/X11/X was a broken symlink to /usr/bin/X11/X
[07:18] <mdz> didn't fix compose:rwin though
[07:20] <mdz> oh, and now my scroll wheel doesn't work. BONUS
[07:20] <Lathiat> yeh
[07:20] <Lathiat> need to add a ZAxisMapping
[07:21] <daniels> mdz: yeah, x-common fun with moving directories to symlinks
[07:21] <\sh> and this is a real "serious wake up call" mythtv and inline asm ftbfs on amd64 *yawn*
[07:21] <\sh> morning gentlemen
[07:22] <Lathiat> heh
[07:22] <Lathiat> im surprised it built on x86
[07:22] <Lathiat> i tried and kept failing
[07:23] <mdz> Lathiat: I have zaxisampping in xorg.conf, as always
[07:23] <\sh> Lathiat: tseng called me a "hero" for that ... but when it's build on amd64, I would like to call myself \sh
[07:23] <Lathiat> mdz: oh? mine wasnt there.
[07:23] <Lathiat> \sh: heh
[07:24] <Lathiat> \sh: .. did it ever build on amd64?
[07:25] <\sh> and powerpc is much more worse "no opcode 'emms'"
[07:28] <\sh> mdz: possible to compile mythtv against an external libavcodec?
[07:29] <mdz> \sh: not in general, no
[07:29] <mdz> \sh: the person to talk to is Chutt on #mythtv
[07:29] <\sh> mdz: so i have to get the patches for avcodec from somewhere else
[07:29] <mdz> I had mythtv building on amd64, powerpc and i386 a few months ago
[07:29] <mdz> and working playback on all three
[07:30] <mdz> 0.18.1 has some amd64 fixes
[07:30] <mdz> Chutt is now using an amd64, so it'll work out of the box there in the future
[07:31] <\sh> mdz: good to know..so I will give 0.18.1 a try later on...
[07:31] <mdz> \sh: I have 0.18.1 source packages in my repository on dijkstra
[07:32] <\sh> mdz: send me the link ;)
[07:38] <\sh> laters...
[07:38] <jsgotangco> later
[07:39] <AndyFitz> The Ubuntu font has beenLGPL'ed :-)   http://andy.fitzsimon.com.au/Ubuntu-Title.ttf
[07:41] <jsgotangco> its beautiful!
[07:41] <jsgotangco> the w and m now blend well
[07:41] <carstenh> fixed++ :)
[07:49] <infinity> AndyFitz : Thanks for the license change.
[07:50] <AndyFitz> infinity.: no worries LGPL is the most simple and sane for a font 
[07:52] <bob2> woo, godd stuff AndyFitz 
[07:54] <infinity> AndyFitz : Assuming you want GPL-style protection, yes, LGPL seems the best fit for something as widely-used as a font.
[07:54] <AndyFitz> now making your own ubuntu derived project is 10% easier
[07:55] <siretart> bob2: around?
[07:55] <bob2> indeedy
[07:56] <siretart> may I query?
[07:56] <bob2> if it's about lyx, I'm uploading to sid tommorow
[07:56] <siretart> aah. great!
[07:56] <siretart> I'm trying to get a working lyx into universe
[07:57] <bob2> what's wrong wit hthe current one? (aside from being old)
[07:57] <siretart> if you upload tomorrow, that'll be a great birthday present :)
[07:57] <siretart> the current one doensn't build at all, because of gcc issues
[07:57] <bob2> oh, right
[07:57] <siretart> neiter gcc-3.4 nor gcc-4.0
[07:57] <bob2> hm
[07:57] <bob2> that's probably RC in Debian then
[07:57] <bob2> oops
[07:58] <siretart> bob2: I tried to check out the arch repository on alioth, but failed
[07:58] <bob2> hm?
[07:58] <bob2> that's odd
[08:00] <siretart> get: invalid revision spec (pkg-lyx-devel@lists.alioth.debian.org--devel)
[08:00] <bob2> right
[08:00] <bob2> that's the name of the archive, not something you can check out :)
[08:01] <siretart> hmm.. what was the right command then?
[08:01] <bob2> one sec
[08:01] <bob2> 'baz get  pkg-lyx-devel@lists.alioth.debian.org--devel/lyx--debian--1.0 debian/'
[08:02] <bob2> tho it's pretty obsolete
[08:02] <siretart> ah..
[08:02] <siretart> If you want, I'd like to test your upload to sid in breezy
[08:03] <siretart> that is, on my breezy installation, of course..
[08:11] <daniels> mdz: new libx11 and metacity blatted at the archive
[08:12] <mdz> daniels: thanks for merging Isaac's fix
[08:12] <daniels> no worries
[08:12] <daniels> i need to commit that upstream also
[08:14] <pitti> Good morning
[08:14] <bob2> siretart: thanks, I'll email you when I've put it somewhere
[08:15] <siretart> bob2: that'll be great!
[08:17] <pitti> mdz: do you want separate reports for every universe word list l-support now depends on? these are trivial packages...
[09:04] <pitti> mjg59: ping
[09:05] <\sh> re
[09:20] <ogra> mdz, still around ?
[09:24] <Mithrandir> mdz: yes, partimage for amd64 is a large amount of work.
[09:31] <\sh> Mithrandir: please install in breezy chroot on ravel: libmysqlclient14-dev liblame-dev liblircclient-dev libxxf86vm-dev libdvb-dev g++-3.4 gcc-3.4 thx :)
[09:34] <Mithrandir> E: Couldn't find package thx
[09:34] <Mithrandir> yeah, and E: Couldn't find package liblame-dev
[09:34] <Mithrandir> is it multiverse?
[09:35] <robitaille> Mithrandir: yes
[09:35] <\sh> yepp
[09:35] <Mithrandir> \sh: did you investigate why qt ftbfs?
[09:36] <Mithrandir> \sh: installed
[09:36] <\sh> Mithrandir: yesterdays fix from me and infinities upload fixed 
[09:36] <\sh> qt
[09:36] <Mithrandir> ok, goodie
[09:36] <\sh> missing QRegion region; 
[09:36] <\sh> Mithrandir: thx :)
[09:36] <jdub> Mithrandir: does suspend work at all on amd64?
[09:37] <ogra> jdub, it did on hoary
[09:37] <jdub> ok, so it's intended that it work
[09:37] <jdub> it would be other factors that made it not work
[09:37] <jdub> right?
[09:37] <Mithrandir> jdub: no idea, my amd64 has SATA which makes Things Blow Up
[09:38] <ogra> jdub, i dont know about amd64 desktop systems... 
[09:39] <ogra> (mine is a laptop)
[09:39] <jdub> right, the user has sata :-)
[09:40] <\sh> ogra: what thoughts are coming to you when I say: libavcodec on amd64?
[09:40] <ogra> \sh, none... i'm still totally shocked that my iso dissapeared...
[09:41] <\sh> hmmm
[09:41] <ogra> and dont know what to do now, since i called for testers for this one
[09:41] <ogra> the two that are there are totally unusable....
[09:41] <Mithrandir> jdub: I think fabio is putting some patch to support suspend on SATA now-ish.
[09:41] <\sh> mdz: mythtv 0.18.1 is ftbfs on amd64 as well, at the same point as 0.18
[09:41] <ogra> and i dont understand why 20050818 isnt there anymore
[09:42] <fabbione> Mithrandir, jdub: yes.. but they are experimental. they work on some systems.. not all of them
[09:42] <fabbione> Mithrandir, jdub: that's for sure better than none of them working
[09:42] <fabbione> jdub: later this evening they will be in archive
[09:42] <Treenaks> Mithrandir: suspend-on-sata? Is that like wake-on-lan, but backwards?
[09:42] <fabbione> upload planned for 12:00/13:00 UTC more or less
[09:43] <jdub> yeah, noticed that change
[09:43] <Treenaks> Mithrandir: "Hey, disk activity, let's suspend"
[09:46] <\sh> Dear Linus, include updated sk98lin drivers asap so I can work with Linux out of the box
[09:46] <Treenaks> \sh: you forgot "kthxbye"
[09:46] <fabbione> \sh it will not happen.. they are working a rewrite of it
[09:47] <\sh> fabbione: ok..what about this idea...I will create a additional package for it...or a separate kernel package for some marvell yukon ethernet machines?
[09:47] <\sh> at least for breezy
[09:48] <\sh> fabbione: forget it
[09:49] <fabbione> \sh the point is simple.. that driver won't be in the stock kernel. Making another kernel for you is not an option.
[09:49] <\sh> fabbione: module compiling doesn't work with this package...only "patching kernel, recompile kernel, copy module from kernel tree to usbdevice, from usbdevice to r200"
[09:49] <fabbione> so you can only make a module
[09:50] <\sh> fabbione: would it be an option for integrating the module into the installer kernel?
[09:50] <fabbione> jamesh: no
[09:50] <fabbione> mah
[09:50] <fabbione> \sh no..
[09:50] <\sh> see
[09:50] <fabbione> that means having the module in the stock kernel
[09:51] <\sh> and this is for cd/dvd less machines with this bloody hardware a "no breezy out of the box at all" .. no pxe install
[09:53] <fabbione> \sh i am not going to compromise on that as i already explained to you a few times by now
[09:53] <fabbione> that driver is crap
[09:53] <fabbione> and we can't commit to support it
[09:53] <michele> fabbione, looks like everybody wants you to put crappy patches in the kernel ;)
[09:53] <michele> sounds like a conspiration
[09:54] <\sh> ah come on...
[09:54] <daniels> don't forget drm
[09:55] <\sh> I know I'm teasing fabbione a bit...but, think about this: the normal user doesn't care about "crappy code patches inlines" whatever...they know only: "it's working" or "it's not working" ... and I want to find out what's the best solution for "providing at least unofficial support"
[09:56] <fabbione> \sh the normal user cares when at the first security upgrade everything breaks down because the driver doesn't compile anymore
[09:57] <michele> \sh, don't worry, I was making fun of myself... I teased him for the last two days ;)
[09:57] <\sh> fabbione: the user shouldn't compile the driver at all 
[09:58] <\sh> I should write a wiki page on how to adjust the install iso ,-)
[09:58] <fabbione> \sh warning: you are pushing my patience over a certain limit.. 
[09:59] <fabbione> we have been over this topic over and over
[09:59] <fabbione> you got an answer with a proper explanation
[09:59] <\sh> fabbione: forget what I'm saying...:)
[10:08] <mvo> ping jamesh 
[10:08] <seb128> hey mvo
[10:08] <mvo> hey seb128 
[10:08] <mvo> good morning!
[10:09] <daniels> morning sebarino
[10:09] <seb128> hey daniels
[10:09] <daniels> seb128: just as a heads-up, I uploaded metacity earlier
[10:10] <daniels> seb128: libx11-dev was missing an x11proto-kb-dev depends, so metacity couldn't use XKBlib.h and thus disabled XKB support
[10:10] <seb128> daniels: np. What did you change?
[10:10] <daniels> seb128: so I just changed the B-Ds
[10:10] <seb128> oh, nice catch :)
[10:10] <daniels> seb128: removed the xlibs-pic B-D as well since that was a bit bong
[10:10] <daniels> seb128: try this
[10:10] <seb128> what sort of bug does that create?
[10:10] <daniels> seb128: press left-alt+tab
[10:10] <daniels> seb128: now, on a US keyboard, press right-alt+tab
[10:10] <seb128> right one does nothing
[10:11] <daniels> ralt isn't caught as a modifier by all the core stuff, so the keypress throws up the window switch
[10:11] <daniels> but the keyrelease doesn't break the grab
[10:11] <daniels> i figured it was probably missing xkb support when I recompiled to add a printf and it magically worked
[10:12] <seb128> cool
[10:13] <seb128> infinity: around?
[10:15] <Mithrandir> I think gamin is leaking memory:
[10:15] <Mithrandir> 20792 tfheen    15   0  539m 529m  844 S  1.0 35.2  18:15.51 gam_server
[10:16] <fabbione> Mithrandir: and are you surprised?
[10:17] <fabbione> Mithrandir: try to use the kill -SIGUSR2 and look in /tmp for the debugging log
[10:17] <fabbione> you should be able to see in sometime why
[10:17] <fabbione> mostlikely it's not releasing the lists of monitored dir/files
[10:18] <rob^> how come lame is included with breezy, but gstreamer0.8-lame isn't?
[10:18] <jdub> rob^: lame is in multiverse (evil), and the gstreamer package is not built against it
[10:19] <Mithrandir> $ wc -l /tmp/gamin_debug_pqNZ4B
[10:19] <Mithrandir> 7280 /tmp/gamin_debug_pqNZ4B
[10:19] <rob^> does gstreamer0.8-lame have copyright/patent issues?
[10:19] <seb128> Mithrandir: http://bugzilla.gnome.org/show_bug.cgi?id=313846
[10:20] <seb128> rob^: it has
[10:20] <pitti> seb128: why should it?
[10:20] <pitti> seb128: lame has, but an interface to lame?
[10:20] <seb128> pitti: gstreamer0
[10:21] <seb128> pitti: gstreamer0.8-lame doesn't do anything without lame ...
[10:21] <seb128> pitti: I mean we can't ship it
[10:21] <jdub> rob^: we can't ship lame due to mp3 patent concerns
[10:21] <pitti> seb128: right; well, that's the same issue as with gstreamer-mad and libmad
[10:21] <rob^> jdub, is in the repo though
[10:21] <seb128> pitti: yeah
[10:21] <rob^> its
[10:21] <pitti>    libmad0 | 0.15.1b-2.1 | http://archive.ubuntu.com breezy/main Packages
[10:21] <Mithrandir> seb128: internal server error.
[10:21] <seb128> it's multiverse
[10:21] <jdub> rob^: and we don't build stuff in main against stuff in universe or multiverse
[10:22] <seb128> Mithrandir: http://bugzilla.ubuntu.com/show_bug.cgi?id=13449
[10:22] <jdub> (or we'd shift it into main, which we simply cannot do with lame)
[10:22] <pitti> seb128: then we have to demote libmad0, otherwise we can as well ship g-mad
[10:22] <pitti> same for xine
[10:22] <rob^> cant you just include gstreamer0.8-lame in universe?
[10:22] <jdub> pitti: yes, these are also concerning
[10:22] <pitti> jdub: libmad0 and libxine1 are still in main
[10:22] <rob^> or multiverse
[10:22] <seb128> pitti: we don't ship gst-mad, it's universe
[10:23] <jdub> rob^: we could build it completely separately in multiverse, but that would require all kinds of work we're not interested in doing
[10:23] <rob^> ah ok
[10:23] <rob^> np
[10:23] <Mithrandir> seb128: ok, anything I can do to help getting it fixed?
[10:23] <jdub> rob^: because the gstreamer plugins are mostly built from one source (at the moment)
[10:23] <pitti> seb128: I know, but since we ship libmad0 anyway, we can as well ship g-mad; g-mad doesn't have issues, libmad0 has
[10:23] <seb128> Mithrandir: ping desrt about that, he talk with GNOME guys about it
[10:23] <jdub> pitti: libmad0 in main is an aberration
[10:23] <seb128> pitti: right, but my understanding was that we should not ship limbad, not other way
[10:24] <pitti> seb128: right, just teasing :-)
[10:24] <Mithrandir> desrt: any idea what's up with ubuntu #13449?  I see the same issue here
[10:24] <rob^> ok, thanks jdub 
[10:24] <seb128> ;)
[10:24] <jdub> (at one point we fixed libxine)
[10:24] <seb128> (at one point gst will be so good than nobody will care about xine :p)
[10:25] <jdub> yes, libmad0 is only a suggests
[10:25] <pitti> seb128: can't be; the package is patented, not copyrighted
[10:25] <pitti> seb128: so if gst is as good as xine, then it contains the same algorithms and we can't ship it as well
[10:26] <seb128> k, so I got 495 bug mails since aug 1st ...DOH
[10:26] <seb128> (only from bugzilla.ubuntu I mean)
[10:26] <seb128> (and these are just gnome bugs, no ubuntu-bugs)
[10:26] <jdub> seb128: http://people.ubuntu.com/~cjwatson/germinate-output/breezy/rdepends/libmad/libmad0
[10:26] <jdub> ha ha
[10:27] <seb128> DOH
[10:27] <pitti> yes, gst-plugins b-deps on it
[10:27] <pitti> xine-lib b-deps on it as well
[10:28] <pitti> same procedure for totem
[10:29] <pitti> seb128: so AFAICS the only solution is to split off a main-capable package of totem and gst?
[10:29] <seb128> what about totem? totem-xine?
[10:29] <pitti> seb128: yes, same thing
[10:29] <seb128> the solution for totem would be to do 2 sources packages, one for gst, one for xine
[10:29] <seb128> but that kind of suck
[10:29] <jdub> perhaps it's worth doing separate builds of gst plugins then
[10:30] <jdub> xine is harder
[10:30] <seb128> there is some people who were speaking about gst-plugins0.8-universe
[10:30] <pitti> jdub: xine is not hard, it's just nasty
[10:30] <jdub> pitti: build it with the correct deps, ship a package with only the mad loader?
[10:30] <pitti> jdub: just copy the source package and don't build everything
[10:30] <jdub> xine-mad... excellent :-)
[10:31] <pitti> xine-universe
[10:31] <jdub> oh please can we call it mad? :)
[10:31] <pitti> erm, no
[10:31] <pitti> not xine
[10:31] <pitti> totem-universe
[10:31] <jdub> why totem?
[10:31] <jdub> it just depends on xine
[10:31] <pitti> or better totem-gstreamer and totem-xine source pkgs
[10:32] <pitti> jdub: xine-lib has to completely be dropped to universe then
[10:32] <jdub> if we ship a xine-mad package that only includes xineplug_decode_mad.so, totem will just work
[10:32] <pitti> s/completely be/be completely/
[10:32] <jdub> hrm, i don't think so
[10:32] <pitti> jdub: and how do you want to build that?
[10:32] <jdub> make xine build without mad -> easy
[10:32] <pitti> jdub: you certainly need xine hearers
[10:32] <jdub> make a xine-mad source to build just that .so
[10:33] <jdub> same solution for gstreamer
[10:33] <pitti> jdub: but we don't need to split xine-lib, but totem
[10:33] <seb128> what about:
[10:33] <jdub> totem is already split sufficiently to allow for this
[10:33] <seb128> - ship everything to main
[10:33] <seb128> - go to jail
[10:33] <seb128> - enjoy? hum ... :p
[10:33] <pitti> seb128: we Europeans won't go to jail
[10:33] <jdub> seb128: i will not be elmo's bitch :-)
[10:33] <pitti> we still have reasonably sane laws here
[10:33] <seb128> HA HA
[10:34] <daniels> we need to split xine into xine and xine-universe
[10:34] <seb128> pitti: right ... so let's do this :)
[10:34] <daniels> jdub: you already are
[10:34] <pitti> daniels: why?
[10:34] <jdub> we can make both the xine and gst-plugins packages in main not reference mad at all
[10:34] <pitti> jdub: again, we need to split totem and completely demote xine into universe
[10:34] <pitti> no need for splitting xine
[10:34] <jdub> and build alternate versions in universe or multiverse that only include the mad plugins
[10:34] <daniels> pitti: split totem?
[10:34] <daniels> i thought we were shoving xine into main to use that as our preferred backend for breezy
[10:34] <jdub> pitti: we don't need to push xine into universe
[10:34] <pitti> daniels: totem-gst for main, and totem-xine for uiverse
[10:35] <daniels> pitti: errr ... our plan was to do xine as default for breezy
[10:35] <pitti> daniels: EPATENTS
[10:35] <daniels> assuming I find some time in my life for things that are not X
[10:35] <daniels> pitti: exactly, which is wy I need to sit down and split xine into xine and xine-universe
[10:35] <daniels> this is all part of VideoPlaybackRoadmap :P
[10:35] <pitti> daniels: I'm using totem-xine, it's so much better than t-gst, but we can't do that in main
[10:35] <daniels> why not?
[10:35] <seb128> daniels: the plan was to try both options once the patents part dropped
[10:35] <jdub> we need to extract the evil shit
[10:35] <pitti> daniels: all the mpeg stuff is patented
[10:36] <seb128> pitti: just drop them
[10:36] <jdub> pitti: and we can do that at the xine level
[10:36] <daniels> pitti: right, which is why you either just drop it altogether, or build the mpeg stuff separately in universe
[10:36] <pitti> seb128: after dropping mpeg and mp3 from xine, what would be left to make it preferable to gst?
[10:36] <daniels> pitti: a/v sync
[10:36] <jdub> pitti: it works better for video
[10:36] <pitti> ok
[10:36] <daniels> pitti: and you can build the plugins in universe anyway
[10:36] <jdub> pitti: and you can always install the evil stuff from universe/multiverse anyway :)
[10:36] <pitti> but please don't drop it completelya
[10:36] <daniels> (which I think is a complete crock, but whatever)
[10:37] <seb128> pitti: what daniels said ... but I'm not sure of how good is esdsink with BBB's fixes
[10:37] <pitti> *sigh* patents
[10:37] <daniels> esd???
[10:37] <jdub> daniels: (note that explicitly building them elsewhere demonstrates knowledge == triple damages)
[10:37] <pitti> just move everybody to europe...
[10:37] <daniels> jdub: yes, hence 'complete crock'
[10:37] <pitti> or move the european laws to the U.S.
[10:56] <jdub> jbailey: are we going to do basic initramfs root building on the buildd for breezy?
[10:56] <jdub> hrm, and latest kernel has hard depends on initramfs-tools
[10:56] <jdub> bum
[10:57] <jdub> ...
[11:15] <mvo> I need some advice how to phrase the dialog that updates to the new language-pack structure when language-selector is run. I have now: "Update language support?\nSome packages for full language support are not installed on your system. Do you want to install them now?"
[11:17] <poningru> that sounds good
[11:17] <poningru> see the thing is the buttons for that come out as yes/no
[11:17] <poningru> you may wanna try to make them as install/ dont install
[11:18] <mvo> right, that sounds good
[11:18] <mvo> thanks
[11:19] <poningru> np
[11:20] <mvo> "install now", "ignore" as captions for the buttons?
[11:21] <poningru> well the ignore may be confusing
[11:21] <poningru> but is acceptable
[11:21] <poningru> depends on what the dialog is coming after
[11:21] <mvo> I would like to avoid having "install" on both buttons ("Install now", "not install"). maybe "keep"?
[11:22] <poningru> right
[11:22] <poningru> no ignore sounds fine for this case
[11:24] <jsgotangco> install/ignore looks fine to me
[11:25] <jsgotangco> keep is also good hmmm i prefer that
[11:31] <poningru> wouldnt you say ignore seems more professional than keep? 
[11:31] <poningru> keep seems a little childish when compared to ignore
[11:31] <Nafallo> skip?
[11:32] <Nafallo> morning btw :-)
[11:33] <mvo> I like "ignore" better too because there isn't anything to "keep" 
[11:34] <Nafallo> but you "skip" the installation ;-)
[11:35] <Nafallo> you may probably say you ignore it to, but that doesn't sound as much as a choice as skip/install IMO.
[11:36] <Mithrandir> mvo: what does l-s do if you choose don't install/skip/ignore?
[11:37] <mvo> Mithrandir: bug you again next time
[11:37] <Nafallo> then it's ignore
[11:38] <Mithrandir> mvo: yes, but what is the next step it'll walk you through?
[11:38] <mvo> the dialog opens and you can selector the various languages you want to see supported
[11:39] <Mithrandir> I would probably use "Upgrade support" and "Remind me next time" then.
[11:39] <Mithrandir> or something along those lines, remind me next time is a bit long
[11:40] <mvo> "Remind me again" ?
[11:40] <mvo> well, not really shorter :/
[11:40] <jsgotangco> "Remind me again" does sound very friendly
[11:43] <mvo> thanks, I'll use it then
[11:47] <poningru> yeah I really like Remind me again
[11:51] <mvo> Mithrandir: from what package?
[11:51] <Mithrandir> kino
[11:52] <Mithrandir> it's from starting it, opening the "open file" dialog and then quitting.
[11:52] <mvo> *ick*
[11:52] <mvo> that does not sound like fun :/
[11:52] <niran> i really should save more often, or at least use something more stable than drpython.
[12:08] <Lathiat> pitti: any of those kernel vulns remote or local exploitable for root?
[12:09] <Treenaks> Lathiat: root doesn't need exploits :P
[12:09] <Lathiat> err
[12:09] <Lathiat> you know what i mean ;p
[12:09] <Lathiat> altho with selinux that isnt necesarily true ;p
[12:09] <Lathiat> i suppose a crash usually means writing things where they shouldnt be so its possible
[12:18] <mjg59> mdz: That's with laptop-mode completely disabled?
[12:19] <mjg59> pitti: Hi
[12:19] <pitti> Lathiat: no, mostly DoS
[12:19] <pitti> mjg59: Hi! just tested usplash on my amd64 box with nvidia card, works fine :-)
[12:20] <mjg59> pitti: Rock
[12:20] <pitti> mjg59: however, is it already supposed to work on the live CD out of the box?
[12:20] <mjg59> Not yet
[12:20] <pitti> mjg59: ok
[12:20] <Nafallo> mjg59: amd64 + ati mobility radeon 9700 works btw :-)
[12:20] <pitti> mjg59: on the live CD, I did not get any output after "Starting Ubuntu...", just a long pause
[12:20] <pitti> that is really irritating
[12:24] <JaneW> chmj: have you totally taking over the Bluetooth SoC bounty mentoring?
[12:25] <chmj> JaneW: yes 
[12:26] <chmj> JaneW: pitti still available to help 
[12:27] <pitti> yes
[12:35] <pef> hello
[01:06] <tepsipakki> does someone here know why *gnome.org is down?
[01:14] <siretart> daniels: hi. I think there is something weird with xmkmf or makedepend. either xmkmf shoud require the command 'makedepend' or makedepend should provide 'gccmakedep'
[01:17] <siretart> daniels: if I symlink gccmakedep to makedepend, libforms1 can be built
[01:19] <fabbione> siretart: be carefull
[01:19] <fabbione> gccmakedepend is not the same
[01:20] <siretart> oh
[01:20] <fabbione> i don't recall details, but they behave differently
[01:21] <fabbione> siretart: thanks.. that's ok.. i remember hitting my head on it a year ago
[01:21] <fabbione> that's why i don't recall ;)
[01:21] <fabbione> or i don't want to remember :P
[01:21] <siretart> well, for libforms1, makedepends seems to be sufficient, but xmkmf creates makefiles requiring gccmakedep
[01:21] <siretart> thats the actual problem I see. after symlinking it, it works
[01:21] <siretart> s/works/builds/
[01:22] <fabbione> siretart: iirc gccmakedepends needs to be part of Imake
[01:22] <fabbione> let me check..
[01:23] <siretart> hm. there is no gccmakedep in imake
[01:23] <Nafallo> seb128_: why do everything on my system still want libcairo1 installed?
[01:23] <fabbione> it's either imake or xmkmf...
[01:23] <fabbione> in my very old pkgs they were all together
[01:24] <\sh> update from hoary to actual breezy is not working nicely..and removing firefox from hoary and trying to install firefox on breezy doesn't work ,-)
[01:24] <siretart> fabbione: it used to be in xutils, but this package is empty for now
[01:24] <fabbione> siretart: yes i know...
[01:24] <Nafallo> s/do/does/
[01:24] <seb128_> Nafallo: "everything"?
[01:24] <fabbione> siretart: we are in the process to start a big X bug squash party from monday..
[01:25] <fabbione> siretart: just open a bug so that we don't forget
[01:25] <seb128_> Nafallo: sudo apt-get remove libcairo2
[01:25] <siretart> fabbione: ok. will do. thanks
[01:25] <fabbione> siretart: np
[01:26] <jbailey> jdub: What do you mean initramfs root building on the buildd?
[01:26] <jbailey> jdub: Having trouble parsing that out. =)
[01:26] <Nafallo> seb128_: synaptic wants to remove the same things as with libcairo1.
[01:26] <Nafallo> seb128_: synaptic bug or something? :-)
[01:27] <seb128_> Nafallo: so things use cairo2 :)
[01:27] <seb128_> Nafallo: no is just you beeing too impatient, and having both doesn't hurt
[01:27] <Nafallo> seb128_: yepp, and deps libcairo2 :-P
[01:28] <seb128_> Nafallo: there was like 200 packages built with cairo1, let's wait to get them build
[01:28] <seb128_> s/build/rebuild
[01:28] <Nafallo> seb128_: hmm, oki :-). I thought stuff would know they should use new cairo (which they do), so I can't see why they want to old one :-P.
[01:29] <seb128_> Nafallo: that doesn't work this way
[01:29] <seb128_> Nafallo: ldd /usr/bin/nautilus | grep cairo
[01:29] <Nafallo> hm, oki. so it does use both :-P
[01:29] <seb128_> Nafallo: you have to rebuild to change a new soname
[01:30] <seb128_> Nafallo: nautilus does use cairo1, it has not rebuilt yet, cairo2 comes from gtk
[01:30] <seb128_> Nafallo: the soname is a build time stuff, not a dynamic running one
[01:30] <Lathiat> hrm
[01:30] <Lathiat> gedit seems to be linked against libcairo 1 and 2
[01:30] <seb128_> ie: you have to rebuild everything on a soname change, it's not working automagically
[01:31] <seb128_> Lathiat: gedit is built with cairo1, gtk with cairo2, cf what I was saying
[01:31] <Lathiat> oh right
[01:31] <Nafallo> seb128_: hmm, oki. I go with that explaination for a while :-)
[01:31] <Nafallo> atleast till I find new arguments ;-)
[01:32] <seb128_> Nafallo: you will not find any good argument on this, soname changes force you to rebuild
[01:32] <Nafallo> seb128_: yepp. but both wants to remove for instance beagle, which is already updated.
[01:32] <seb128> both what?
[01:33] <Nafallo> both libcairo1 and 2 :-)
[01:33] <seb128> keep both
[01:33] <seb128> that doesn't hurt
[01:33] <siretart> fabbione: filed as #13812
[01:33] <Nafallo> I'll sure will :-)
[01:33] <seb128> you need this few ko on your hdd or what?
[01:33] <fabbione> siretart: against what package?
[01:33] <siretart> fabbione: against xutils. they used to be there
[01:33] <Nafallo> seb128: hehe, just bugging me to see Installed (local or obsolute). I want stuff to be clean ;-).
[01:33] <siretart> fabbione: btw, the sparc won't go online this week. Joerg want to put them in another case
[01:34] <seb128> Nafallo: rebuild 180 packages yourself and you will be here
[01:34] <fabbione> siretart: ok. no problem for the sparcs.. there is no rush
[01:34] <seb128> Nafallo: or wait 1 or 2 days
[01:34] <fabbione> siretart: i am busy enough atm :P
[01:34] <Nafallo> seb128: *s*
[01:34] <siretart> sure ;)
[01:34] <sedak> fabbione 
[01:34] <Nafallo> seb128: I rather work on libdps1 then ;-)
[01:35] <sedak> i am considered of adding a driver in the kernel
[01:35] <pitti> ogra__: schoolbell has debian/ files in the orig.tar.gz; is that upstream's fault or an accident?
[01:35] <sedak> can i pm you ?
[01:35] <ogra__> pitti, i dont think anybody of us touched the schoolbell packages
[01:35] <fabbione> sedak: the kernel is basically closed for breezy. open a wishlist bug with all the info and we will look at it for breezy+1
[01:36] <pitti> ogra__: Version: 1.1.1-1ubuntu2
[01:36] <fabbione> sedak: for curiosity... what driver is that?
[01:36] <ogra__> pitti, except for dep changes in the python transition
[01:36] <pitti> ogra__: you touched it last, before there was a sync
[01:36] <sedak> it's http://bugzilla.ubuntu.com/show_bug.cgi?id=12679
[01:36] <sedak> i've already made a separated package
[01:36] <sedak> i need to ask siretart to the right to upload
[01:37] <sedak> siretart ?
[01:37] <sedak> you're here ?
[01:37] <siretart> sedak: yes, im here
[01:38] <infinity> seb128 : Still need a mess of give-backs?
[01:38] <fabbione> siretart: i doubt it can make it for breezy
[01:38] <seb128> infinity: yeah, due to libpixman.la
[01:38] <infinity> seb128 : Do you have a list, or do I need to hunt?
[01:38] <seb128> wait
[01:38] <siretart> fabbione: I really would like to have lyx in breezy!
[01:39] <fabbione> siretart: lyx ?
[01:40] <ogra__> pitti, i definately didnt make it native :)
[01:40] <daniels> siretart: xmkmf/makedepend> i dunno.  not everything using xmkmf needs makedepend, but it could be nice.
[01:40] <ogra__> pitti, must have been upstream... 
[01:40] <Nafallo> siretart: lyx in swedish is luxury, so sure :-)
[01:40] <seb128> infinity: gnome-desktop eel2 nautilus nautilus-cd-burner totem
[01:41] <fabbione> daniels: look around my young padowa
[01:41] <seb128> infinity: these packages with this order, and after that you can massive give back ... do you need the whole list for other stuff?
[01:41] <daniels> fabbione: hmm?
[01:41] <siretart> fabbione: lyx is a grafical word processor, using latex
[01:42] <fabbione> daniels: never mind :)
[01:42] <fabbione> siretart: and why do you ask me?
[01:42] <seb128> does somebody know where hides http://arch.ubuntu.com/gnome@arch.ubuntu.com/ ?
[01:42] <fabbione> siretart: i will upload my first set of pkgs smaller than 1MB today .. for the first time in my life..
[01:42] <fabbione> it's going to be an experience
[01:42] <daniels> fabbione: you love modularisation
[01:43] <seb128> jamesh: do you know what's going on with http://arch.ubuntu.com/gnome@arch.ubuntu.com/ ?
[01:43] <pitti> seb128: ok, cool, we have (or had) the whole gnome cvs imported?
[01:43] <Nafallo> hi daniels :-)
[01:43] <fabbione> daniels: i love sed more :)
[01:43] <seb128> pitti: maybe not the whole, but a good part yeah ...
[01:43] <jamesh> seb128: in what sense?
[01:43] <pitti> jamesh: 404 sense
[01:43] <seb128> jamesh: 404 not found
[01:43] <jamesh> hmm
[01:43] <ogra__> pitti, 1.0-1 has this too
[01:43] <seb128> jamesh: that used to work ... :)
[01:44] <pitti> ogra__: ok, nm
[01:44] <jamesh> seb128: try bazaar.ubuntu.com
[01:44] <ogra__> pitti, so its upstreams fault... i'll talk to jinty, he made the package afaik
[01:44] <seb128> jamesh: this one works, thanks
[01:44] <siretart> daniels: I attached a debdiff to libforms1 to bug #13812 I just filed. It just needs this gccmakedepend issue. now I'm stuck because I don't where to get gccmakedep :(
[01:44] <pitti> ogra__: either upstream's or the packager's
[01:45] <ogra__> pitti, they are the same
[01:45] <jamesh> seb128: no idea why arch.ubuntu.com stopped working though
[01:45] <pitti> ah :-)
[01:45] <ogra__> pitti, jinty is release manager for schooltool and friends as well as the packager ;)
[01:45] <mvo> jamesh: did you got my launchpad-integration mail?
[01:46] <seb128> jamesh: no big deal, the other one works fine :)
[01:48] <infinity> seb128 : Do those need to be done serially?
[01:49] <jamesh> mvo: just looked at it quickly.  I reckon we'll need a bit of extra API for it to actually work from Python though (the "this is my source package name" API)
[01:49] <jamesh> mvo: or else we'll be directing everyone to /distros/ubuntu/breezy/+sources/python/whatever
[01:49] <mvo> jamesh: good point, I'll have a look
[01:50] <ogra__> pitti, was ripping out mozilla of nvu a requirement or a suggestion ?
[01:50] <pitti> ogra__: please don't introduce another copy of moz code
[01:50] <seb128> infinity: yep
[01:51] <ogra__> pitti, hmm... ok....
[01:51] <pitti> ogra__: we try to get rid of the code, not get more of it
[01:51] <infinity> seb128 : Right.  You sucks.  That is all.
[01:51] <infinity> s/sucks/suck/
[01:51] <pitti> ogra__: can you please try to build with m-firefox-dev?
[01:51] <daniels> siretart: you don't want gccmakedep, I don't understand why you're trying to use it
[01:51] <ogra__> pitti, have you seen the package ? 
[01:51] <seb128> infinity: not my fault if we ship these .la files !
[01:51] <daniels> siretart: the xorg monolithic tree uses 'makedepend', which is packaged under the cleverly-concealing name, 'makedepend'.
[01:51] <pitti> ogra__: yes; a PITA
[01:51] <ogra__> jup
[01:52] <daniels> siretart: i assume everything else using xmkmf uses this also, which would make sense, because xutils previously provided makedepend, not gccmakedep.
[01:52] <infinity> seb128 : So take daniel's lead and stop shipping them. :)
[01:52] <daniels> siretart: if you want to use gcc's depend thing, you can, but makedepend is there also.
[01:53] <seb128> infinity: we should just have a dh_something hacked to drop them
[01:54] <daniels> dh_stabintheface
[01:54] <infinity> I'd support that.  dh_destroy should be a facist script that just removes anything we don't like.
[01:54] <daniels> which eliminates all of the things that make daniels angry
[01:54] <daniels> 'libtool?  not any more!'
[01:54] <daniels> 'xprint, not bloody likely'
[01:55] <Treenaks> dh_fascist ?
[01:55] <Nafallo> dh_daniels
[01:55] <Nafallo> :-)
[01:55] <ogra__> pitti, you should probably leave schooltool for now..., i forwarded you a mail... but it doesnt look like they'll make it in time 
[01:56] <Treenaks> Nafallo: what should that do? s/^\(Maintainer:\).*/\1 Daniel Stone <>/ in debian/control ?
[01:56] <pitti> ogra__: ok, then we can also defer schoolbell?
[01:56] <daniels> Treenaks: hrranghr
[01:56] <Treenaks> daniels: oh wait, control.in of course
[01:58] <Nafallo> Treenaks: that and dh_fascist things ;-)
[01:58] <ogra__> pitti, the mail only talks about schooltool.... and as i said, i doubt they'll make it in time... i told jinty the packages have to be ready and pbuilder/lintian clean before the 30th, but i didnt hear back anything yet
[01:59] <daniels> lintian?
[01:59] <ogra__> pitti, see the mail from sabdfl about it... but i dont want to delay edubuntu just because schooltool is late
[01:59] <daniels> dh_stabintheface would remove lintian also
[01:59] <jbailey> Is there an easy way to tell what package versions are on the livecd without booting into it?
[01:59] <daniels> jbailey: download it, mount it, mount the cloop, dpkg -l
[01:59] <ogra__> (edubuntu is delayed enough already)
[01:59] <Nafallo> daniels: merge all that in dh_daniels please ;-)
[02:00] <pitti> ogra__: which mail from Mark? can you forward it to me, please?
[02:00] <jbailey> daniels: Ah well, thanks. 
[02:01] <pitti> erm, gcc changelog: "Build-depend on libcairo2-dev"; seb128, did you infiltrate gcc already? gtk bugs in our compiler?
[02:01] <pitti> gcc - now with font smoothing
[02:01] <Nafallo> lol
[02:01] <seb128> pitti: gcc uses gtk :)
[02:02] <pitti> seb128: does gcc now have a gtk gui???
[02:02] <pitti> scream
[02:02] <infinity> pitti : java.
[02:02] <seb128> pitti: no, but cairo is used for gcj
[02:02] <ogra__> pitti, i just did... thats the mail i talk about all the time ...
[02:03] <pitti> ogra: ah, I see
[02:03] <ogra> :)
[02:05] <\sh> daniels: xlibmesa-gl-dev is now libgl1-mesa-dev ?
[02:05] <\sh> or libgl1-xorg-mesa?
[02:06] <daniels> \sh: libgl1-mesa-dev.  libgl*-xorg* is dead.
[02:07] <\sh> daniels: and according to this xlibmesa-glu-dev is libglu1-mesa-dev
[02:07] <infinity> \sh : libgl1-mesa-dev for GL and libglu1-mesa-dev for GLU.
[02:07] <\sh> infinity: thx ...
[02:08] <infinity> \sh : I'm still uploading fixes all night.  They have t obe done in a reasonably sane order.
[02:08] <\sh> infinity: well..I just check the unmetdeps list..and I found a girlfriend again which I fixed before mesa trans.
[02:09] <infinity> ... You fixed your girlfriend before the mesa transition?
[02:09] <\sh> hehhee
[02:09] <infinity> I'm so lost.
[02:09] <\sh> well..yes...girlfriend is arkrpg ,-)
[02:10] <Treenaks> \sh: so you're a l33t g4m3r now?
[02:10] <infinity> Anyhow, if you do plan on fixing stuff, can you make sure to remove ALL references to xlibmesa-* and libgl*-xorg*?
[02:10] <Nafallo> \sh: your girlfriend needed to be fixed :-). did she build cleanly on all arches?
[02:10] <daniels> infinity: ... why ... on earth ... does kdebase ... want xmkmf ...
[02:10] <infinity> \sh : Otherwise, just wait for me to do them all.
[02:10] <infinity> daniels : Fucked if I know, dude.  CHeck the build log.
[02:10] <Nafallo> \sh: does that mean you ported virtual valerie? ;-)
[02:11] <ogra> pitti, ?? only needed for schooltool, but that will not be ready for Breezy; deferred ??
[02:11] <\sh> infinity: u don't want to do universe, right? 
[02:11] <ogra> pitti, we need it for edubuntu, the question is which version ...
[02:11] <pitti> ogra: ah, you can use it on its own? 
[02:11] <pitti> ogra: ah, ok
[02:11] <ogra> nope
[02:11] <daniels> infinity: urgh
[02:12] <seb128> pitti: is there a known locale issue?
[02:12] <infinity> \sh : I don't know if I WANT to, but I was going to anyway.
[02:12] <pitti> seb128: no?
[02:12] <\sh> infinity: ok...then we will w8 for u :) 
[02:12] <pitti> ogra: nope == you can't use it on its own?
[02:12] <ogra> pitti, its just a question if we do a last minute update, if this would need a new main inclusion report, please leave it until we know if they make it
[02:13] <pitti> ok
[02:16] <ogra> pitti, i just want to avoid doubled work
[02:17] <sedak> fabbione, i've uploaded the module package on revu (rtl8180-kernel)
[02:18] <fabbione> revu??
[02:18] <sedak> i don't know what to do next ... if someone want to check it, it may be a good idea as it is my first package
[02:18] <fabbione> what's that?
[02:18] <sedak> i don't really know :-)
[02:18] <Nafallo> fabbione: MOTUs reviewsystem :-)
[02:18] <fabbione> Nafallo: ok
[02:18] <siretart> daniels: I don't really want gccmakedep. but xmkmf is creating makefile which call that binary!
[02:19] <fabbione> sedak: -> #ubuntu-motu
[02:19] <Nafallo> fabbione: http://siretart.tauware.de/revu/ for a look ;-)
[02:19] <fabbione> Nafallo: i am quite busy..
[02:19] <sedak> ok
[02:19] <Nafallo> fabbione: oki :-)
[02:19] <siretart> fabbione: he talks about this one: http://siretart.tauware.de/revu/details.py?upid=422
[02:20] <ogra> fabbione, its a kernel nic module, the question i wanted sedak to ask you was if its possible to get it included in the kernel rather then having a separate package in universe thats why i sent him here... 
[02:21] <zul> fabbione: if need be i can create a patch for it
[02:22] <fabbione> ogra: breezy kernel is closed
[02:22] <seb128> mvo: is gnome-language-selector changing any config file?
[02:22] <fabbione> only bug fixing
[02:22] <ogra> fabbione, oh
[02:22] <ogra> ok
[02:22] <zul> ok
[02:22] <mvo> seb128: yes the language in /etc/envirorment 
[02:22] <seb128> mvo: I've used it to install the french input package yesterday and today I've LANG set to fr_FR
[02:22] <seb128> instead of fr_FR.UTF-8
[02:22] <mvo> *ick*,
[02:23] <mvo> seb128: that's a bug, I'll fix it (that it does not set utf-8)
[02:23] <fabbione> zul: thanks, but we to go bug fixing now..
[02:23] <pitti> mvo: does l-s touch ~/.profile or call language-env?
[02:23] <seb128> mvo: thanks to it I've
[02:23] <seb128> locale: Cannot set LC_CTYPE to default locale: No such file or directory
[02:23] <seb128> locale: Cannot set LC_MESSAGES to default locale: No such file or directory
[02:23] <seb128> locale: Cannot set LC_ALL to default locale: No such file or directory
[02:23] <seb128> mvo: fr_FR@euro and fr_FR.UTF-8 are generated but not fr_FR ... so it broke my locales :)
[02:23] <mvo> pitti: no, it does not touch ~/.profile 
[02:23] <Nafallo> hmm, I got that bug :-)
[02:23] <pitti> mvo: do you call language-env?
[02:24] <mvo> pitti: no
[02:24] <mvo> should I?
[02:24] <pitti> mvo: seb128 has some l-env magic comments and the wrong LANG= between in his .profile
[02:24] <pitti> mvo: no, you shouldn't
[02:24] <Treenaks> seb128: hey, I've heard that about nl_NL vs nl_NL.UTF-8 as well
[02:24] <Treenaks> pitti: , sorry
[02:24] <Nafallo> I did dpkg-reconfigure -plow locales to fix is though :-P
[02:25] <seb128> mvo, pitti: the .profile maybe comes from me, and it's fr_FR@euro which is not b0rked
[02:25] <seb128> but
[02:25] <seb128> $ grep LANG /etc/environment
[02:25] <seb128> LANGUAGE="fr_FR:fr"
[02:25] <seb128> LANG=fr_FR.UTF-8
[02:25] <seb128> $ sudo gnome-language-selector
[02:25] <seb128> $ grep LANG /etc/environment
[02:25] <seb128> LANGUAGE="fr_FR:fr"
[02:25] <seb128> LANG="fr_FR"
[02:25] <seb128> 
[02:26] <seb128> that's it
[02:26] <seb128> pitti, mvo: thanks
[02:27] <seb128> mvo: should I bugzilla that?
[02:27] <mvo> seb128: if you want, but I have made a not for myself already
[02:27] <daniels> siretart: oh.  libforms1 sucks.
[02:28] <seb128> mvo: ok, that's fine if you know about it, no need of bugzilla :)
[02:28] <mvo> ok, thanks for noticing :)
[02:28] <seb128> np
[02:29] <tseng> seb128: will you take care of problems from cairo transition?
[02:29] <tseng> seb128: for example muine ppc ftbfs on libpanel
[02:29] <siretart> daniels: thats possible. how do I make it to use makedepend?
[02:30] <daniels> siretart: wait for xmkmf 0.99.0-3, which forces makedepend for projects which can't be arsed picking which one they want :P
[02:30] <siretart> daniels: excellent! :)
[02:31] <seb128> tseng: yeah
[02:31] <tseng> seb128: thanks.
[02:31] <fabbione> Nafallo: You don't have permission to access /revu/incoming/rtl8180-kernel-0508191415/rtl8180-kernel_0.22.0cvs050817-1_source.changes on this server.
[02:31] <fabbione> so how one is supposed to check stuff, if you can't download it?
[02:31] <siretart> fabbione: thats on purpose. that will be fixed in revu2
[02:31] <seb128> tseng: np. This one should autoretry anyway
[02:31] <siretart> fabbione: you don't need the changes file, the other files are downloadable
[02:32] <fabbione> yes if i only want to check what's changed....
[02:32] <fabbione> and i don't need to download *
[02:32] <siretart> fabbione: yeah, I could have just rm'ed the file, or sed'ded the gpg sigs. but that was more convinient for me to debug this prototype
[02:33] <siretart> fabbione: you can use the debdiff link from the overview page
[02:33] <siretart> and sorry, debdiffing to version in archive will be in revu2
[02:33] <fabbione> even if you sed the gpg sign you obtain nothing useful
[02:34] <fabbione> once i can download .dsc .diff.gz .orig i can rebuild and get a really signed .changes
[02:34] <siretart> I wanted to prevent that $RANDOM_GUY grabs signed packages from review and upload it to ubuntu. 
[02:34] <siretart> there are also ppl with keys in the ubuntu keyring uploading to revu
[02:35] <fabbione> siretart: just strip the signature :)
[02:35] <fabbione> or mangle it to death
[02:35] <sedak> fabbione, i am changing the package a little anyway
[02:35] <fabbione> ok
[02:35] <siretart> fabbione: we are currently working on a rewrite: revu2. there we will do something like that. (most probably throwing changes file away and regenerating it, we will see)
[02:36] <sedak> i'll tell you when i think it'll be the final temporary version :-)
[02:36] <fabbione> siretart: just strip the signature :)
[02:36] <fabbione> it's easier
[02:36] <fabbione> i love "final temporary version" :)
[02:37] <Nafallo> hehe
[02:37] <siretart> :)
[02:40] <JaneW> are there any willing LTSP testers around?
[02:41] <lu|dinner> define 'ltsp tester' :)
[02:41] <JaneW> http://wiki.ltsp.org/twiki/bin/view/Ltsp/LtspBreezyTest
[02:41] <tseng> morn luis
[02:41] <luis_> morning, tseng
[02:41] <JaneW> and http://edubuntu.org/EdubuntuTesting
[02:42] <luis_> I don't really have time this morning,  but I'll read through; I might have time this afternoon.
[02:42] <JaneW> luis_: finally https://wiki.ubuntu.com/ThinClientIntegration (enough?)
[02:42] <JaneW> luis_: cool we desperately need more testers
[02:43] <luis_> I've been interested in playing with ltsp for a long time
[02:43] <luis_> this seems like a good excuse :)
[02:43] <luis_> (since I have no other really good reason to do it :)
[02:43] <ogra> luis_, you need two machines for that, one that acts as ltsp server and one that can PXEboot as thin client
[02:44] <Lathiat> matching?
[02:44] <Lathiat> oh
[02:44] <luis_> ogra: yes, I know; I have three around the house :)
[02:44] <Lathiat> sorry
[02:44] <Lathiat> i read that out of nowhere, ignore me.
[02:44] <luis_> four, but one has not booted in at least three years :)
[02:49] <Nafallo> hmm, there is nothing new about the breezy+1 conference?
[02:52] <ogra> yay, pitti
[02:52] <ogra>   --with-system-nspr      Use system installed NSPR
[02:52] <ogra>   --with-nspr-prefix=PFX  Prefix where NSPR is installed
[02:52] <ogra>   --with-nspr-exec-prefix=PFX
[02:52] <ogra>                           Exec prefix where NSPR is installed
[02:52] <seb128> infinity: you can start kicking all the other stuff when nautilus is built, few stuff build-depends on nautilus-cd-burner/totem
[02:52] <ogra> hehe, looks easier then i thought
[02:52] <pitti> ogra: well, that's nspr, not the complete mozilla; but it's a start :-)
[02:52] <ogra> (nvu that is)
[02:53] <pitti> ogra: and using the system nspr is required anyway :-)
[02:53] <ogra> it doesnt use it now
[02:53] <luis_> JaneW: anyway, anything in particular you wanted poked on ltsp? or just 'I can pxe-boot the remote client from the server'?
[02:53] <luis_> 'and nothing immediately blows up'? :)
[02:54] <ogra> luis_, "you can use it and work with it...."
[02:54] <luis_> ok
[02:56] <JaneW> luis_: I am passing on the message for mdz, so you could ask him, else I think testing in general is required...

[02:57] <luis_> OK, I'll poke around, after I'm burnt out on the morning's task :)
[03:00] <Mithrandir> elmo: please sync libtextwrap
[03:03] <fabbione> elmo: please start the NEW the kernels...
[03:03] <fabbione> to NEW even
[03:03] <azeem> daniels: do you happen to have a list of all the new x.org source packages, or is there an easy way to get them
[03:03] <azeem> ?
[03:04] <daniels> azeem: ha ha ha ha ha ha
[03:04] <azeem> pfft
[03:05] <daniels> daniels@ephemera:~/canonical% find drm/ mesa/ xfonts/ xorg/ -name \*.orig.tar.gz | cut -f1 -d_ | wc -l
[03:05] <daniels> 122
[03:05] <daniels> azeem: plus about another 40 or so in drivers
[03:06] <daniels> azeem: so probably about 170 all up, plus all the apps fabio's doing ... maybe 200 altogether?
[03:06] <azeem> hrm
[03:16] <pitti> Hey jdthood , how are you
[03:17] <jdthood> pitti: Fine, thanks
[03:17] <seb128> jamesh: 
[03:17] <seb128> lib/Makefile.am:
[03:18] <seb128> -EXTRA_DIST = lpi.defs
[03:18] <seb128> +EXTRA_DIST = lpi.defs lpi.override
[03:19] <mvo> seb128, jamesh: added, thanks
[03:19] <seb128> mvo: added what to what? what archive should I package ... ? :)
[03:20] <mvo> seb128: heh :) added to my branch
[03:24] <mvo> seb128: I'm changing the name of the python module from "lpi" to "LaunchpadIntegration", please wait with a upload until that is finished 
[03:25] <seb128> mvo: sure, I'm just playing here, I'm waiting to have a python version of launchpad_integration_set_sourcepackagename :)
[03:25] <mvo> seb128: already done ;) please merge from my archive
[03:34] <seb128> pitti: do you know why there is no mozilla-firefox-dom-inspector current version for hoary?
[03:36] <\sh> hmmm...I will play this weekend with shtoom 
[03:36] <\sh> I think this is a nice work to do ... learning twisted, learning shtoom ,-)
[03:37] <pitti> seb128: it's not? did you look in universe? (it shold be there)
[03:38] <seb128> pitti: a pygtk guy asked here yesterday, he has hoary-security for main and universe 
[03:38] <seb128> pitti: he says there is no 1.6.0-0ubuntu0.0.1 for it, just 1.6.0-0ubuntu0.1 from warty-security
[03:38] <pitti> seb128: 
[03:38] <pitti> mozilla-firefox-dom-inspector | 1.0.2-0ubuntu5 | hoary/universe | amd64, i386, ia64, powerpc
[03:38] <pitti> mozilla-firefox-dom-inspector | 1.0.6-0ubuntu0.1 | hoary-security/universe | amd64, i386, ia64, powerpc, sparc
[03:39] <seb128> hum
[03:39] <pitti> http://archive.ubuntu.com/ubuntu/pool/universe/m/mozilla-firefox/
[03:39] <seb128> k, I'll ping him again
[03:39] <pitti> it's there
[03:39] <seb128> sorry for the noise
[03:39] <seb128> from where do you get the "mozilla-firefox-dom-inspector | 1.0.6-0ubuntu0.1 | hoary-security/universe | amd64, i386, ia64, powerpc, sparc" ? 
[03:41] <pitti> seb128: madison on jackass
[03:41] <seb128> thanks
[03:42] <pitti> seb128: if you have the deb lines for hoary-security, you can also do it locally with apt-cache madison
[03:42] <pitti> no worries :-)
[03:47] <pitti> ah, so my new glibc is not a complete disaster at least :)
[03:47] <seb128> cool
[04:01] <Nafallo> fabbione: ping
[04:02] <fabbione> Nafallo: very short pong.. i am going offline for a bit
[04:02] <Nafallo> fabbione: why did the new kernel sett my root=/dev/hda2 to sda?
[04:02] <Nafallo> set even
[04:02] <fabbione> never.. the kernel does NOT set root=
[04:02] <fabbione> that's initramfs
[04:03] <fabbione> or grub
[04:03] <fabbione> the kernel doesn't play with that stuff at all
[04:03] <fabbione> Nafallo: you want to talk with jbailey 
[04:03] <infinity> Argh, GCC, QT, glibc, kernel... Could we upload a few more giant packages now, so I can't get anything done?  Please?  Yay.
[04:03] <Nafallo> fabbione: thanx :-)
[04:04] <j^> seb128 will you update totem to  1.1.4 or is totem not considered part of gnome-desktop?
[04:04] <fabbione> infinity: is the kernel actually building anywhere?
[04:04] <fabbione> i mean.. even sparc is already uploaded.
[04:04] <fabbione> your buildd's are slow...
[04:04] <infinity> No, but it ate some buildds for a while.  And I felt like complaning.
[04:05] <fabbione> linux-source-2.6.12_2.6.12-7.11_sparc.changes is NEW
[04:05] <Nafallo> jbailey: why did the root=/dev/hda2 became root=/dev/sda2 in grubs menu.list after I upgraded the kernel? :-)
[04:05] <fabbione> infinity: so noone is actually building the kernel?
[04:06] <seb128> j^: I've updated to 1.1.4 the day of the tarball
[04:06] <Nafallo> jbailey, fabbione: dooh! I hate ssh :-P
[04:06] <fabbione> infinity: better you get the kernel done for when i am back :P
[04:07] <infinity> fabbione : powerpc and ia64 are.
[04:07] <j^> seb128 http://packages.ubuntu.com/breezy/gnome/totem still shows 1.1.3-0ubuntu4
[04:07] <fabbione> what's up with amd64/i386?
[04:08] <seb128> j^: Source Package: totem, Download: [dsc]  [totem_1.1.4.orig.tar.gz]  [totem_1.1.4-0ubuntu2.diff.gz]  
[04:08] <seb128> j^: the build issue is on my list
[04:08] <j^> seb128 http://people.ubuntu.com/~lamont/buildLogs/t/totem/1.1.4-0ubuntu2/ failed on all archs
[04:08] <j^> ah ok
[04:09] <infinity> fabbione : Uhh, and i386.
[04:09] <infinity> fabbione : Only amd64 is done.
[04:09] <fabbione> ok
[04:09] <fabbione> thanks
[04:09] <fabbione> later
[04:10] <whiprush_> j^: did you mention the other day that you had patches for network-manager someplace?
[04:11] <j^> whiprush_ http://bootlab.org/~j/bazaar/
[04:12] <whiprush_> thanks
[04:12] <j^> debian-dir--network-manager--0.4 + network-manager--j--0
[04:13] <lucas> hi
[04:17] <lucas> I can't seem to find memtest on hoary's CDs. am I blind ? if I am not, what's the pseudo package to submit a wishlist bug about this ? debian-cd ?
[04:24] <torkel> whiprush_: I never heard anything from the guys working on the kerberos ticket manager. Did the project die?
[04:32] <Mithrandir> lucas: it's there, package name is memtest86+
[04:32] <Mithrandir> lucas: it's installed into the grub menu by default
[04:33] <lucas> I mean it's not available from the isolinux command line at boot
[04:33] <lucas> without installing ubuntu
[04:33] <lucas> (like it is in knoppix for example)
[04:33] <Mithrandir> it's on the live cd
[04:33] <Mithrandir> at least in breezy
[04:33] <lucas> ok
[04:33] <lucas> not in hoary it seems
[04:35] <lucas> erm
[04:36] <lucas> :-)
[04:38] <lucas> I really can't find it. The only available "boot methods" are live and live-expert
[04:39] <desrt> Mithrandir; no.  no idea.
[04:39] <desrt> Mithrandir; it's happened to several people, though
[04:40] <desrt> Mithrandir; kjartan maraas has found a leak in gamin but it's quite small in comparison
[04:41] <Mithrandir> desrt: is there anything I can do to help track it down?
[04:41] <desrt> Mithrandir; you can valgrind gam_server i guess
[04:42] <Mithrandir> desrt: I'm on amd64, unsure if that has anything to do with it (given that the initial report was for i386)
[04:42] <desrt> Mithrandir; as so far it's been quite difficult for anybody to get it to leak that memory in a controlled environment
[04:42] <Mithrandir> I'll see what I can get done.
[04:47] <mdke> what is the workaround to the current x-window-system problem in breezy?
[04:47] <mdke> (with libgl1-xorg-dri)
[04:48] <daniels> install libgl1-mesa-dri and libgl1-mesa instead
[04:48] <lamont> mdz: pics finished copying sometime during the night
[04:49] <mdke> daniels, okay thanks
[04:49] <mdke> daniels, i should also remove x-window-system-core?
[04:50] <mdke> oh sorry my bad
[04:50] <mdke> i see it
[04:50] <lamont> lucas: if you have a warty live iso, I expect it's there
[04:50] <lamont> (memtest, that is)
[04:51] <tseng> what is the url to the top of the arch supermirror?
[04:51] <mdke> daniels, actually I can't seem to get it to install those
[04:52] <mdke> i need to force something?
[04:58] <mdke> does anyone else know?
[05:03] <infinity> mdke : It's a transient dependency issue between several metapackages, I suspect.
[05:04] <mdke> could be
[05:04] <mdke> but i'd like to fix it because I can't seem to update anything else until i do
[05:05] <infinity> What's the error?
[05:05] <infinity> You should be able to give apt explicit instructions on how to proceed.
[05:06] <infinity> Like "apt-get --purge install libgl1-mesa libglu1-mesa libgl1-mesa-dri x-window-system-core- ubuntu-desktop-"
[05:06] <infinity> (Note the trailing "-" on the packages you want to remove)
[05:07] <infinity> You can reinstall ubuntu-desktop in a few days when we've unbroken the metapackage deps.
[05:08] <mdke> i'll try that command thanks
[05:08] <azeem> //(/ /w 28
[05:09] <infinity> mdke : Or something similar.  Note that I've not seen the exact error on your machine, so I'm just guessing at a reasonable solution.  Basically, if apt needs to change state of too many packages at once, it just gives up and begs you to tell it what to do.
[05:09] <mvo> seb128: do you think we should upload the new python enabled launchpad integration stuff? if so, I can upload a new g-a-i with that enabled 
[05:09] <infinity> mdke : So you need to tell it (in one run, hence the +/- notation) what to do.
[05:09] <seb128> mvo: where did that landed? to your archive?
[05:10] <infinity> mdke : Never an issue in a final release, i we're doing our jobs right, but during development, we confuse apt a LOT at times.
[05:10] <mvo> seb128: yes, you can merge from it
[05:10] <seb128> mvo: fine with me. Want to update the package or should I do it?
[05:10] <mdke> infinity, yeah that is good to know. I'm not 100% sure of all of the packages i need to remove, but I'll tinker around
[05:11] <seb128> infinity: how is the buildd kicking going?
[05:11] <infinity> mdke : Alternately, use a smarter frontend like aptitude or dselect, which may actually figure it out for you
[05:11] <infinity> seb128 : Okay, so far.
[05:11] <mdke> infinity, okay i'll try
[05:12] <seb128> infinity: you can massively retry from now since nautilus is built
[05:14] <infinity> seb128 : <nod>... Just waiting fron cron.daily to finish.
[05:14] <seb128> cool
[05:15] <infinity> The GCC/glibc/kernel/QT uploads kinda stuffed me up for bit.
[05:15] <infinity> I'm just going to find the silver lining and be thankful that no one uploaded OpenOffice too.
[05:16] <seb128> how does the retry stuff work? You have to give a list of package to the buildds?
[05:16] <infinity> In essence.
[05:16] <infinity> I cheat in sick and strange ways.
[05:16] <seb128> he he
[05:18] <dilinger> infinity: where are you living now, btw?
[05:18] <daniels> on the wrong side of town
[05:19] <infinity> Or the right one.
[05:19] <infinity> dilinger : Melbourne, last time I looked.
[05:22] <dieman> grr
[05:22] <dieman> i think the tds.net mirror is behind again
[05:36] <jasoncohen> pitti, i thought you weren't going to patch gnupg
[05:38] <pitti> jasoncohen: well, that open bug and the ubuntu-cve entry just annoyed me :-)
[05:38] <jasoncohen> pitti, heh, so it wore you down
[05:39] <pitti> jasoncohen: I couldn't stand bugzilla throwing it at me every day any more :-)
[05:39] <jasoncohen> pitti, when were the CVE's affecting the kernel found? i didn't see them the last time i checked the list of open CVEs
[05:40] <pitti> jasoncohen: well, they aren't "found", but requested
[05:40] <pitti> jasoncohen: I just requested some from mitre some days ago
[05:40] <pitti> jasoncohen: but mitre only updates their db once or twice a week
[05:40] <pitti> so they are lagging a bit
[05:40] <pitti> but many of the kernel CANs are there
[05:42] <jasoncohen> i still receive DSAs and it's pretty amusing that firefox and mozilla still haven't been patched by Debian. i think they fixed a few of the more serious issues only
[05:42] <pitti> jasoncohen: yes, only two of them
[05:42] <pitti> probably those they could backport
[05:42] <jasoncohen> so, are they just not going to patch the rest?
[05:42] <pitti> there was a discussion about whether just using new upstreams
[05:42] <pitti> no idea
[05:43] <pitti> I learned the hard way that patching can become impossible...
[05:43] <pitti> or, at least, exponentially hard
[05:43] <infinity> pitti : Did you see the Debian thunderbird update?... Absolute crack.
[05:44] <pitti> I remember the ffox one..
[05:44] <infinity> They backported EVERYTHING between 1.0.2 and 1.0.6 /except the version number/
[05:44] <pitti> infinity: I saw asacs prepared 1.0.6 package
[05:44] <infinity> So, it's 1.0.6, but reports 1.0.2.  Crazy.
[05:44] <pitti> infinity: haha
[05:44] <pitti> so essentially they used asac's package
[05:44] <pitti> lol
[05:44] <infinity> Not sure if that was asac's idea (his name's on the changelog) or joey's.
[05:45] <infinity> But.  Weird.
[05:45] <infinity> Very weird.
[05:45] <pitti> infinity: I talked with asac, he wanted it to be 1.0.6-0sarge1 or so
[05:45] <fabbione> infinity: how is going ppc build?
[05:47] <pitti> does anybody have reasonably knowledge about openssl?
[05:47] <infinity> fabbione : building powerpc-smp right now.
[05:47] <infinity> pitti : How deep in the API do you need to go?
[05:48] <pitti> infinity: not at all
[05:48] <infinity> pitti : I've wrapped my head around some bits used in php-openssl and apache-ssl.
[05:48] <pitti> infinity: I just wonder whether it will break anything if we change the default md algorithm to sha-1 (from md5)
[05:48] <fabbione> infinity: ok.. that doesn't tell me much :).. how many other flavours have been done already?
[05:48] <infinity> Changing the default is almost certainly a bad idea.
[05:48] <infinity> fabbione : No idea.
[05:48] <infinity> fabbione : I'm just tailin ght elog.
[05:49] <pitti> infinity: #13593; upstream does it in 0.9.8 since md5 became less secure
[05:49] <infinity> pitti : Are we that scared of md5 collisions?
[05:49] <fabbione> infinity: cd linux-source-2.6.12-2.6.12/debian/build && for i in build-*; do du -sk $i; done
[05:49] <pitti> infinity: the reporter of that bug has, apparently
[05:49] <fabbione> you can see by size :)
[05:49] <pitti> infinity: s/has/is/
[05:49] <infinity> pitti : In theory, because hashes are tagged uniquely, all the functions that read hashes should read them correctly regardless.
[05:50] <infinity> pitti : So the only potential for issues is with third party apps that expect something to have been an md5 and now it's not.
[05:51] <elmo> hum, how do I use the 'rescue' mode without having it mount the root filesystem?
[05:51] <pitti> infinity: right, but that's the part I don't understand - "(we didn't change 0.9.7 as we didn't want to break existing
[05:51] <pitti> implementations depending on the default digest being MD5)"
[05:51] <infinity> pitti : Oh, he's just asking for it for cert generation?.. That sounds more safe.
[05:52] <infinity> elmo : Get a woody boot disk? :)
[05:52] <infinity> elmo : Or kick to another VT before it gets to the "mount a root disk" option.
[05:53] <infinity> pitti : I thought we were talking a fundamental API/ABI change, requesting that all hash/digest functions return sha-1 instead of md5 if the hash is unspecified.
[05:53] <elmo> yeah, well, I'm having fun figuring out switch to another VT on powerpc d-i
[05:53] <infinity> pitti : It it's just openssl.cf he wants changed, I doubt it'll break TOO much.
[05:54] <daniels> elmo: get a real keyboard, hippy
[05:54] <pitti> infinity: ouch, no. It should still be supported, of course
[05:54] <infinity> elmo : OpenApple-F2?
[05:54] <pitti> infinity: ok, I just do that change in breezy, then let's see :-)
[05:54] <pitti> infinity: thansk
[05:54] <fabbione> elmo: can't you just use the normal installer in expert mode and stop at parman? go back to the menu and down to the "Execute a shell"?
[05:54] <daniels> elmo: (more seriously, I think the sense of Fn is inverted, so it's something like OpenApple-Fn-F1)
[05:54] <fabbione> s/parman/partman
[05:55] <ogra> fabbione, if he cant switch to console ? 
[05:55] <fabbione> ogra: that does not require to switch console
[05:55] <fabbione> it requires to boot the installer in expert mode and make it stop at partman
[05:55] <ogra> err, yes, i misread, sorry
[05:55] <fabbione> hit "Go Back"
[05:56] <fabbione> and go to the main menu, where there is "Execute a shell"
[05:56] <fabbione> same VT :)
[05:56] <ogra> :)
[05:57] <lamont> mdz: syncing syck_0.55-2 from debian will fix ia64's FTBFS  - thoughts?
[06:15] <mjg59> Simira: Latest kernel in Breezy should reboot on your hardware
[06:16] <\sh> re
[06:18] <sivang> hi all
[06:24] <fabbione> infinity: please make ppc go faster and install a ppc64 kernel
[06:33] <pitti> good bye, nice weekend!
[06:36] <mjg59> I installed a new kernel, rebooted and I have a message telling me I need to reboot
[06:36] <Lathiat> yeh its good isnt it
[06:36] <mjg59> I already have done!
[06:36] <Lathiat> that thing should be shot
[06:37] <mjg59> On boot, messages telling me I need to reboot ought to be cleaned out
[06:41] <KaiL> does somebody know about the nvidia driver issues with GeForce2 and older?
[06:42] <KaiL> because breezy has ONLY the newer driver (which does NOT work with this cards..)
[06:43] <mjg59> KaiL: Yes. nvidia appear to have shafted us.
[06:43] <fabbione> KaiL: you can't have 2 nvidia drivers at the same time.
[06:43] <fabbione> and you need to ask them why the old card is not supported
[06:43] <Burgundavia> mjg59, sorry I have been offline. My home internet died. There are a few things I want to chat to you about, issues with community people and getting good testing results
[06:43] <fabbione> it's a binary driver for which there is no code or support from us directly
[06:43] <KaiL> fabbione, additional packages with "legacy" in the name for the old one?
[06:44] <infinity> Probably something as simple as missing PCI IDs or something, since their drivers should be backward compatible all the way to the TNT.
[06:44] <fabbione> KaiL: no way...
[06:44] <KaiL> ...if the old driver works at least for 2.6.12.....
[06:44] <fabbione> that probably doesn't
[06:44] <KaiL> I'm just trying ;)
[06:44] <infinity> KaiL : Mail nVidia and complain.  Really.  It might help.
[06:45] <infinity> (Moreso than complaining to us will, that's for sure..)
[06:45] <KaiL> infinity, my solution is to buy an ATI...:)
[06:45] <mdz> mjg59: yes, but then I've also re-enabled everything and can't get it to crash
[06:45] <mjg59> mdz: Uhm. Weird.
[06:45] <fabbione> hey mdz
[06:45] <mjg59> Burgundavia: No problem
[06:47] <jk24> Hi, just one question : what mean "FTBFS" ? (on https://wiki.ubuntu.com/HoaryAptGetOrg?highlight=%28schurger%29)
[06:48] <fabbione> Fail To Build From Source
[06:49] <KaiL> hmm, which gcc is used for the breezy kernel?
[06:49] <KaiL> 3.4? 3.3?
[06:49] <fabbione> 3.4
[06:52] <fabbione> infinity: how is ppc doing???
[06:53] <Burgundavia> mjg59, ok, just to confirm, we are testing ootb support for laptops? If it needs a custom hack to get it going, it should be marked as no? (within reason of course)
[06:54] <elmo> fabbione: it's building packages
[06:54] <fabbione> elmo: ok great
[06:54] <jbailey> Nafallo: =)
[06:54] <fabbione> elmo: i did seed everything already.. so it should be easier for you to handle the stuff
[06:55] <mjg59> Burgundavia: It should be marked as no and the hack documented
[06:55] <Burgundavia> mjg59, ok
[06:55] <daniels> fabbione: what's wrong with lrm?
[06:55] <fabbione> daniels: kernel abi bump
[06:55] <mdz> fabbione: hi
[06:55] <daniels> ah
[06:56] <fabbione> mdz: i am afraid i won't make it in time to upload d-i and BenC did fall off the net
[06:56] <fabbione> lrm and linux-meta just uploaded
[06:56] <KaiL> uhm, which version was the "legacy"? ;)
[06:56] <hmrocha> KaiL, warty?
[06:57] <KaiL> hmrocha, nvidia "legacy" driver for GF2
[06:57] <hmrocha> KaiL, ok, i was out of context, sorry
[06:58] <mdz> fabbione: did you get a chance to talk to him first?
[06:58] <fabbione> mdz: yes..
[06:58] <mdz> fabbione: so he knows what to do?
[06:58] <fabbione> we had a long talk and stuff.. all of sudden he died
[06:59] <fabbione> mdz: yes, but if net doesn't come up for him, we can't block the abi transition at the last package :)
[06:59] <fabbione> mdz: the sources are ready and signed in my homedir on chinstrap
[06:59] <fabbione> it's only question of a lftp to jackass
[07:00] <Keybuk> seb128: isn't dragging a .ttf file into nautilus(fonts://) supposed to install them?
[07:01] <seb128> it does
[07:01] <seb128> but the view is not refreshed probably
[07:01] <seb128> the ~/.fonts has it?
[07:01] <Keybuk> ah, yes
[07:01] <\sh> hmmm...
[07:03] <siretart> daniels: I fixed now libforms1. with your latest xmkmf upload, I could build it again. all libraries moved from /usr/X11R6/lib to /usr/lib and /usr/X11R6/include to /usr/include
[07:03] <siretart> do you want to review that package or do you have enough trust in my? ;)
[07:03] <siretart> s/my/me/
[07:08] <siretart> last daniels 
[07:08] <siretart> sry
[07:09] <KaiL> 7174 works with 2.6.12, good
[07:09] <daniels> siretart: i'm too busy to review it anyway, so go nuts
[07:09] <KaiL> at least that works
[07:09] <siretart> daniels: ok. sorry bothering..
[07:09] <daniels> np
[07:09] <fabbione> YAY
[07:09] <fabbione> ppc is up
[07:09] <fabbione> right in time 
[07:10] <dieman> the driver i merged into hoary's X server has bugs :|
[07:10] <dieman> havent had a chance to look into it yet
[07:10] <daniels> dieman: hmm.  i think we have the latest (wrt i945) in breezy.  anything in particular?
[07:10] <dieman> sometimes when you logoff, X respawns into 640x480 but the screen is drawn as the native res (1280x1024) in this case
[07:10] <dieman> but it isn't a virtual screen
[07:10] <dieman> didn't have a chance to pull a log, so I need to run up there and recreate the issue
[07:11] <daniels> nice
[07:11] <daniels> if you do, bounce it to me and cc alanh@fairlite.demon.co.uk
[07:11] <dieman> ok
[07:11] <dieman> i just pulled that patch out of breezy, and im not using dri
[07:11] <dieman> since the kernel isn't new enough ;)
[07:11] <daniels> heh
[07:12] <Lathiat> daniels: what do i look out to figure out what kind of ati an ati is? (like r300 or what?)
[07:12] <daniels> Lathiat: lspci
[07:13] <Lathiat> daniels: well, ati firegl v3100 
[07:13] <Lathiat> havent got access to it atm
[07:14] <fabbione> elmo: i am going to do a batch upload of X crap, that will need NEW.. please send them right to universe.. nothing needs to go in main now.. these are cp -rp pkg1/debian pkg2/debian + sed -e... pkgs.. so nothing to complex to look at..
[07:15] <daniels> Lathiat: i think that's rv350 or therabouts, but yeah, lspci.  either that or trawl cvs.freedesktop.org/xorg/xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_chipset.h
[07:15] <Lathiat> its apparently a 'radeon'
[07:15] <daniels> lathi	firegl != radeon.
[07:16] <Lathiat> rv370 apparently
[07:16] <Lathiat> daniels: well, the page i just looked at lies then :)
[07:16] <Lathiat> open source 3d/dri ?
[07:16] <elmo> fabbione: eh, ok
[07:16] <daniels> Lathiat: a) firegls are certainly not radeons; based off the same chipset, sure, b) with the r300.sf.net driver, assuming it's agp, sure
[07:16] <Lathiat> daniels: cool
[07:17] <Lathiat> oops think i annoyed him
[07:28] <fabbione> mdz: ABI transition should be completed
[07:28] <fabbione> all kernerl bits and pieces should be in the archive at :33
[07:28] <mdz> fabbione: should be, or is?
[07:28] <KaiL> damn nvidia shit.....
[07:28] <fabbione> lrm will start to build at :35
[07:29] <fabbione> di uploaded and it will build at :35 too
[07:29] <fabbione> in theory at :03 of the next everything should be in place
[07:29] <mdz> d-i needs byhanding
[07:29] <fabbione> mdz: yes.. it still needs to build first
[07:29] <mdz> I don't see the d-i- upload in the queue
[07:29] <fabbione> mdz: i just did it
[07:29] <mdz> who uploaded it?
[07:29] <fabbione> me
[07:29] <fabbione> it will be processed now
[07:30] <mdz> oh, it hasn't even been accepted yet
[07:30] <fabbione> mdz: also.. the first batch of X crap is in upload phase
[07:30] <fabbione> now..
[07:30] <KaiL> uhm, why can I delete /lib/modules/2.6.12-6-386/volatile/nvidia.ko, reboot and the file is there again? ;)
[07:30] <fabbione> since it's like 15 hours that i am around...
[07:31] <fabbione> i would like to get some dinner
[07:31] <mdz> fabbione: did you talk with elmo about the byhand?  I'm not entirely comfortable doing that part
[07:31] <KaiL> is there some strange feature, I'm missing? :)
[07:31] <tseng> KaiL: because its unpacked from an initrd
[07:31] <fabbione> can i count on you guys to keep an eye on the last bits?
[07:31] <Lathiat> KaiL: the file is generated on boot
[07:31] <tseng> or, an initramfs I guess
[07:31] <fabbione> mdz: no.. i think elmo knows d-i
[07:31] <KaiL> hmm, and how to REALLY get rid of it? ;)
[07:31] <tseng> dont load it?
[07:31] <mdz> fabbione: about being around at the right time and getting it through quickly
[07:31] <Lathiat> KaiL: uninstall linux-restricted-modules or stop caring about it
[07:31] <KaiL> ...without removing the restricted overall
[07:31] <Lathiat> KaiL: why do you care?
[07:31] <fabbione> KaiL, Lathiat -> #ubuntu
[07:31] <Lathiat> fabbione: point
[07:32] <KaiL> Lathiat, the second is a bit problematic, as I'm needing an older one ;)
[07:32] <Lathiat> fabbione: sorry
[07:32] <tseng> please file a bug instead of just removing things
[07:32] <maswan> fabbione: did you need me for something?
[07:32] <KaiL> tseng, scroll up, the but goes to support@nvidia.com
[07:32] <KaiL> or better sales@? ;)
[07:33] <fabbione> maswan: buttercup keeps dieing at light speed, but i really have no time to look at it now.. leave it dead for now :)
[07:33] <Lathiat> KaiL: -> #ubuntu please
[07:33] <fabbione> maswan: otherwise i will need to stay here even longer
[07:33] <elmo> gargar
[07:33] <fabbione> mdz: no ... but i guess he is experienced enough to know that we are doing the usual abi bump dance by now...
[07:33] <maswan> fabbione: ACK. I'm on vacation until next week anyway. :)
[07:33] <elmo> fabbione: are these really all seperate upstream?
[07:33] <fabbione> maswan 
[07:33] <fabbione> maswan: enjoy!
[07:33] <maswan> fabbione: thanks!
[07:34] <fabbione> elmo: modular X.. all different source
[07:34] <elmo> this is so much crack
[07:34] <fabbione> elmo: and it's not even finished..
[07:34] <fabbione> still uploading and there is more to do
[07:34] <elmo> 60-130k source
[07:34] <elmo> x 18 or something
[07:34] <fabbione> elmo: mdz was asking if you can please be around for when d-i is built
[07:34] <fabbione> x51 more or less
[07:35] <fabbione> brb
[07:36] <elmo> someone should tell xorg upstream that really cool people split their projects BY .c file!!!1
[07:37] <Treenaks> elmo: who are you? Herbert Xu?
[07:38] <elmo> fabbione: argh, dude, you can't just cut'n'waste copyright files
[07:40] <fabbione> elmo: daniels told 2 minutes after.. i will fix that later..
[07:40] <fabbione> 2 minutes after i started the batch upload
[07:41] <fabbione> i tought they had a common copyright
[07:41] <fabbione> well up to you if you want to reject them is fine by me
[07:41] <fabbione> the upload is finished
[07:42] <elmo> if you don't mind I will, otherwise my inner ftp-master will cry quietly
[07:42] <fabbione> elmo: ok.. drop them all so that i can reupload the same version
[07:43] <infinity> But dude, we NEED xeyes in the archive like NOW.
[07:43] <fabbione> it will make my life easier
[07:43] <fabbione> infinity: s/NOW/YESTERDAY
[07:43] <mdz> s/NOW/LAST WEEK/
[07:45] <infinity> If I can convince upstream to drop xeyes (and other such useless demo apps) from the official 7.0 or 7.1 release, can we never ever ship them again?  Huh, huh? :)
[07:46] <fabbione> infinity: lrm FTBFS on ppc.. E: Couldn't find package linux-headers-2.6.12-7-powerpc
[07:47] <fabbione> i am afraid ppc didn't make the daily
[07:47] <infinity> That's why it's dep-wait, dude.
[07:47] <infinity> You're looking at the log that triggered the dep-wait.
[07:48] <fabbione> ah ok
[07:48] <fabbione> infinity:  i don't have real time access to build log..
[07:48] <fabbione> nothing i can do about it.. sorry
[07:48] <infinity> Rebuilding it now.
[07:48] <fabbione> perfect
[07:48] <fabbione> am64/i386 are go
[07:49] <fabbione> what about d-i ?
[07:52] <infinity> fabbione : Erm.. For some value of "now" that means "right after the two security builds clogging the two buildds that aren't building gcc-3.4 are done"
[07:53] <fabbione> linux-meta is ok everywhere
[07:53] <infinity> Should still make cron.daily, unless lrm takes forever to build.
[07:53] <fabbione> so we miss d-i and lrm on ppc
[07:53] <fabbione> the latter probably :)
[07:53] <infinity> If not, you get another half hour wait.  No big deal.
[07:53] <fabbione> i am off for dinner
[07:53] <fabbione> back in 20
[07:53] <infinity> (I'll gloss over the part where the 3 security build just uploaded were done by me, before really thinking)
[07:54] <infinity> s/build/builds/
[07:54] <infinity> *cough*
[08:01] <tseng> is kickstart the prefered method for automated installs?
[08:02] <Lathiat> i think kickstart is a layer over the debian-installer stuff which in theory is more flexible
[08:02] <Lathiat> preseed ?
[08:02] <tseng> right
[08:02] <Lathiat> ahh lovely 400M mirror sync
[08:03] <Lathiat> damn these rebuilds
[08:03] <tseng> https://wiki.ubuntu.com/AutomatedInstallations?highlight=%28install%29%7C%28automated%29
[08:03] <tseng> hm
[08:13] <elmo> mdz: the plan is still do dump oo1 right?
[08:13] <mdz> elmo: yes
[08:14] <mdz> doko: what happened with that ttf fix?
[08:14] <mdz> ttf-opensymbol
[08:19] <mdz> I guess I'll just switch it on in openoffice.org2 and see what happens
[08:20] <elmo> mdz: I was also wondering about reducing the number of gcc's in main, but I think we're kind of stuck with 3.3 :(
[08:21] <mdz> elmo: for what?
[08:22] <mdz> hmm, grub
[08:22] <mdz> libpng??
[08:22] <elmo> yeah, >> 3.3 miscompiled it
[08:22] <mdz> grub and libpng
[08:22] <elmo> doko was working on that, but I'm not sure if it's fixed yet
[08:22] <elmo> grub's changelog talks about 0.97 being the first to support >> 3.3, and we're only on 0.95
[08:23] <mdz> we will rival woody in our gcc content
[08:23] <infinity> Woody was.. 2.72, 2.95, 2.96, 3.0..?  I think we're worse.
[08:25] <\sh> ok...shtoom (SVN HEAD) works for sipgate.de now I can do some real nasty stuff
[08:25] <elmo> infinity: quick, which ppc buildd do you care least about?
[08:25] <infinity> elmo : The one you're about to reboot!
[08:26] <infinity> elmo : Take royal, I just killed the buildd there.
[08:26] <mdz> elmo: maybe we could at least stop building everything except the C compiler in 3.3
[08:28] <elmo> infinity: no, it wsn't a request for you to down stuff
[08:28] <elmo> I mean which one isn't d-i or live cd
[08:28] <infinity> elmo : Oh.  ross, I think.  Let me doublecheck.
[08:28] <elmo> thanks
[08:29] <infinity> Nope, I lie.  ross is DI.
[08:29] <infinity> adare is the "spare".
[08:29] <fabbione> YAY
[08:30] <fabbione> d-i is in..
[08:30] <fabbione> perfect..
[08:30] <mdz> jbailey: ping re: 12009
[08:30] <fabbione> mdz: transition completed for my side..
[08:31] <infinity> fabbione : lrm is uploaded, so it's in in 2 minues (+ cron.daily runtime)
[08:31] <infinity> Err, assuming it doesn't need NEW love.
[08:31] <fabbione> infinity: probably elmo did NEW it already :)
[08:31] <infinity> elmo : Giving one away..?
[08:31] <fabbione> infinity: i am happy that everything is built
[08:31] <infinity> elmo : Or trying to plan who to pick on for the new kernel?
[08:31] <fabbione> that's all i can do
[08:32] <elmo> infinity: I'm going to install the normal ubuntu kernel for now, until the FTBFS is fixed
[08:32] <elmo> but it'll probably break because I aliases /usr/bin/hotplug to /bin/false and stuff like that
[08:33] <infinity> I jst purge hotplug to get it out of my hair..
[08:33] <elmo> well, okay I purged it too, but that's not as fun
[08:35] <\sh> who is the "ubuntu UI master"? ,-)
[08:35] <Nafallo> \sh: you :-)
[08:35] <\sh> hahaha
[08:35] <Nafallo> \sh: and ogra :-)
[08:35] <\sh> yes
[08:36] <\sh> Nafallo: btw...shtoom is running with sipgate.de ,-)
[08:36] <\sh> Nafallo: I'm starting to package this nasty beast
[08:36] <\sh> (SVN HEAD)
[08:36] <Nafallo> \sh: ehm, dep twisted 2?
[08:36] <\sh> Nafallo: is already in breezy
[08:36] <\sh> or?
[08:37] <Nafallo> \sh: yesterday you told me it wasn't ;-)
[08:37] <Nafallo> or last night really :-)
[08:37] <\sh> Nafallo: didn't check ,-)
[08:37] <jbailey> mdz: If it's the bug he claims it is, it's already fixed in Hoary and Breezy.
[08:37] <Nafallo> \sh: dooh! we have it. I was onto package that yesterday :-P.
[08:38] <Nafallo> \sh: you fooled me! ;-)
[08:38] <jbailey> mdz: I need to boot off a live cd when my machine's not building to figure out why he's seeing it there.  It might have gone in as a hoary-update, I don't remember.
[08:38] <\sh> Nafallo: never trust a word ,-)
[08:38] <mdz> jbailey: you could ask the submitter
[08:40] <\sh> Nafallo: I'm packaging SER again...
[08:41] <mdz> ogra: that glade file you gave me seems to have a hardcoded date displayed in the dialog
[08:41] <Nafallo> \sh: so now you're packaging both? :-)
[08:41] <\sh> Nafallo: yes.. I need it first for hoary (to put it on the server somehow..or someone is sponsoring a machine for testing ,-))
[08:42] <infinity> elmo : Right, so when are you planning on blowing up adare?
[08:43] <Nafallo> \sh: have to go. screaming girlfriend. she want's to go shopping ;-).
[08:43] <\sh> Nafallo: have fun :)
[08:43] <elmo> infinity: uh, I thought it was ross?
[08:44] <elmo> oh, no, I missed your correction
[08:44] <elmo> infinity: tonight or tomorrow afternoon
[08:44] <elmo> basically, next time I hit the DC
[08:44] <elmo> infinity: oh, same question for amd64
[08:44] <elmo> I'll install breezy on their and serial consoleize it, so we can catch it exploding into little pieces
[08:44] <infinity> elmo : Two lines after I said it was ross, I corrected myself.
[08:45] <elmo> yeah, 19:44 < elmo> oh, no, I missed your correction
[08:45] <infinity> Wow, we can both read!
[08:46] <infinity> elmo : crested is the spare amd64.
[09:02] <ogra> mdz, thats something to fix in the gtk file...
[09:03] <{Seb}> will there be any more updates to the ubuntu kernel in breezy?
[09:03] <ogra> i'll send you an update after the weekent that updates the date/time entry
[09:03] <{Seb}> also, colony 3 is very fast and stable
[09:03] <ogra> weekend indeed
[09:05] <{Seb}> just the latest inotify patch isn't in the current kernel IIRC
[09:06] <fabbione> {Seb}: it's in the kernel we uploaded today
[09:06] <fabbione> please read the changelog
[09:07] <{Seb}> sorry
[09:07] <{Seb}> smashing :-)
[09:07] <{Seb}> also, the bluetooth in breezy kicks ass
[09:07] <{Seb}> my new Nokia phone worked first time
[09:07] <{Seb}> shame it can't be intergratd into the Send To in Nautilus
[09:07] <ogra> mdz, do you have an option to edut the cronscript that creates the edubuntu CD to not delete the old isos? it was very odd to recognize that my only installable CD was deleted last night and only 20050817 and a very broken 20050819 were left... took me the whole day to upload my local copy of 20050818 to rookery
[09:08] <ogra> s/edut/edit
[09:11] <mdz> ogra: it saves 3 days, I think
[09:11] <mdz> ogra: a) the daily CDs should be in working condition at this point in the release cycle, and b) if you need a stable milestone, it should be published as a milestone (like colony 3) so it doesn't disappear
[09:14] <ogra> mdz, yup, but edubuntu only has two daylies left... http://cdimage.ubuntu.com/edubuntu/daily/current/ and there was a lot broken on todays image a lot of very basic packages was rebuilt which left edubuntu-desktop uninstallable... 
[09:14] <ogra> (it complained about libcairo for example)
[09:15] <dabar> Does the live cd come with mp3 support, and can you install packages when using the live cd, do you guys know?
[09:15] <luis_> (1) no (2) yes
[09:16] <ogra> (3) thats an #ubuntu question
[09:16] <luis_> (4) good point.
[09:16] <ogra> :)
[09:17] <dabar> (5) i knew that.
[09:17] <dabar> Just like to bother you guys once in a while to make sure you are not asleep.
[09:17] <dabar> on point 2, onlhy if you make space on the hard drive?
[09:18] <tseng> it does into the tmpfs
[09:18] <tseng> *goes
[09:18] <dabar> so even with no hard drive space allocated you can install codecs for example? (last)
[09:21] <carstenh> jbailey: ping
[09:22] <jbailey> carstenh: pong
[09:23] <dabar> ok, thanks.
[09:26] <ogra> mdz, additionally the preseeding on the CD doesnt work and doesnt reflect the reality... is there any way for me to edit that ? 
[09:27] <ivoks> hi all
[09:28] <mdz> ogra: I don't know
[09:29] <ogra> mdz, so i have to wait for Kamion it seems... :/
[09:30] <_d4vid> hi all
[09:33] <mdz> ogra: I don't know what to tell you.  he was available for the entire release cycle until last week
[09:34] <ogra> mdz, yes
[09:44] <{Seb}> will the gnome foot be replaced by the ubuntu icon by the final breezy release?
[09:44] <{Seb}> on gnome-menu
[09:51] <_d4vid> play Bushido - Schmetterling
[09:52] <azeem> _d4vid: what a poor music taste
[10:09] <mdz> jbailey: I upgraded my ltsp chroot, regenerated the initramfs, and now it doesn't boot anymore
[10:09] <mdz>  /init: 1: cannot open: no such file
[10:09] <mdz> modprobe usage errors
[10:09] <mdz> kernel panic
[10:13] <jbailey> mdz: Can you boot an old kernel so that we can look inside of it?
[10:14] <mdz> jbailey: why boot an old kernel?
[10:14] <jbailey> jammcq and the other gentleman had each tested 0.21 in the ltsp environment before I uploaded.
[10:14] <mdz> I hadn't regenerated my initramfs in a while, so I don't know when it broke
[10:14] <jbailey> mdz: Right, ltsp. =)  Just so we can get to it.
[10:15] <mdz> jbailey: what do you need to see?  I can send you a copy of the initramfs if you like
[10:15] <jbailey> That would work too.
[10:15] <jbailey> Put it up on chinstrap?
[10:15] <mdz> uploading to people.ubuntu.com/~mdz/temp/initrd.img
[10:16] <mdz> ETA <1m
[10:17] <mdz> the first error is after Begin: Running /scripts/local-top\nDone.
[10:17] <mdz>  /init: 1: cannot open : No such file
[10:17] <mdz> then local-premount runs silently, then after it's "Done.", I get a modprobe usage error, then log-bottom and init-bottom run silently
[10:18] <mdz> then run-init complains and it panics
[10:18] <mdz> the version I've just uploaded is already hacked a bit to drop to a shell just before run-init so I could check things out
[10:18] <mdz> the run-init error is: run-init: opening console: No such file or directory
[10:19] <jbailey> Err.  That sounds like the new kernel isn't providing /dev/console automatically anymore.
[10:19] <mdz> it broke before I upgraded the kernel
[10:19] <mdz> I upgraded the kernel to see if that made any difference
[10:20] <Arnia> Hmm... anyone else had a problem with installing gnome-devel saying it will remove ubuntu-desktop?
[10:21] <Arnia> I've just moved across to Breezy so I can get some of the tools I've written ported to new versions of various packages, and when trying to set up my gnome development system I get threatened with the loss of libgl-xorg, libgl-xorg-dri, x-server-core and ubuntu-desktop
[10:23] <mdz> jbailey: dropping to a shell early, I see that /root is empty
[10:23] <mdz> and come to think of it, I never saw ipconfig run either
[10:24] <mdz> so it's simply not mounting root
[10:24] <jbailey> mdz: This is ltsp, right?  So it should be NFS?
[10:25] <jbailey> BOOT is set to local in the mkinitramfs.conf
[10:25] <mdz> jbailey: isn't that supposed to be automatic now?
[10:25] <jbailey> No, it infers all the drivers, but doesn't try to nfs mount unless it thinks it's trying for that.
[10:26] <jbailey> Is there a kernel command line variable that would tell it that it should do NFS instead?
[10:26] <mdz> shouldn't it also error out if it doesn't mount root?
[10:26] <jbailey> Yes, I'll have to look why it's not.
[10:27] <jbailey> Ah, there's a FIXME there.
[10:27] <mdz> jbailey: looks like you're missing quotes
[10:27] <mdz>         if [ ! -e ${ROOT} ] ; then
[10:28] <mdz> "if [ ! -e ] ; " evaluates to false
[10:28] <mdz> while "if [ ! -e '' ] ;" evaluates to true
[10:28] <jbailey> ah.  thanks, will fix that.
[10:28] <mdz> testing with BOOT=nfs
[10:29] <mdz> ok, now it's behaving as before
[10:29] <jbailey> Cool.
[10:30] <mdz> I'm trying to narrow down 12942
[10:31] <mdz> which is a blocker for ltsp
[10:32] <jbailey> Do I just need to tweak the timeout to be longer?
[10:32] <jbailey> Just on the assumption that it's not like it has anything better to do if it times out.
[10:33] <mdz> jbailey: I see no evidence of this having anything to do with nfsmount
[10:33] <mdz> it's the kernel which is timing out
[10:34] <mdz> and it retries, and still never works
[10:34] <jbailey> Ah.
[10:34] <mdz> but for extra extra fun, the new kernel BREAKS UNIONFS FABBIONE
[10:35] <mdz> this means that anyone who is following my colony 3 based LTSP testing instructions gets a non-working system
[10:45] <mdz> jbailey: adding a 'sleep 5' before nfsmount seems to alleviate the issue
[10:45] <mdz> jbailey: (issue = 12942)
[10:45] <mdz> it's some sort of race
[10:47] <jbailey> 'kay.  Is a 5 second sleep going to be too annoying, or should I only do that if nfsmount fails and then retry?
[11:49] <doko> mdz: still on my list for tomorrow
[11:50] <doko> elmo, mdz: will upload both png packages built with 4.0. I couldn't reproduce these with the current 4.0 anymore
[11:51] <mdz> doko: I'm doing a test build with ttf-opensymbol enabled, and will upload it if it works
[11:51] <elmo> doko: any chance grub will work, or are we screwed?
[11:51] <doko> grub: yes, upstream has fixes, but I did fear the merge ...
[11:51] <mdz> elmo: no netcat on jackass?
[11:52] <elmo> mdz: wha for?
[11:52] <mdz> elmo: to copy stuff from the morgue to rookery
[11:52] <elmo> eh
[11:52] <elmo> rsync jackass:: on the lan
[11:52] <mdz> elmo: why is it that morgue.ubuntu.com seems to end at 2005-04-22?
[11:52] <elmo> I stopped morgue syncing because it would run rookery out of sapce
[11:52] <mdz> ah
[11:53] <elmo> I don't know whether to only sync source or what
[11:53] <elmo> or only source for anything older than a month or something
[11:55] <doko> mdz: isn't ttf-opensymbol be built by OO.o2 as well? I can look at it tomorrow.
[11:55] <mdz> doko: no, it's disabled in the oo.o2 build
[11:56] <mdz> I have a test build running on concordia with it enabled; if it works, I'll upload
[11:56] <mdz> then it will supersede the oo.o one
[11:57] <doko> ok