=== Riddell [i=jr@kde/jriddell]  has joined #ubuntu-toolchain
=== infinity [n=adconrad@loki.0c3.net]  has joined #ubuntu-toolchain
[07:20] <doko> Riddell: see the mail to ubuntu-devel, after the flight-1 CD
=== chmj [n=chmj@wbs-146-183-254.telkomadsl.co.za]  has joined #ubuntu-toolchain
[09:14] <Riddell> doko: ok, well qt and above are all sitting here waiting on it so I look forward to being able to upload :)
=== jbailey [n=jbailey@modemcable139.249-203-24.mc.videotron.ca]  has joined #ubuntu-toolchain
[01:18] <jbailey> doko: Thinking of which.
[01:18] <jbailey> doko: kyle pointed out that monticito is dropped ia32 compatability in hardware.
[01:19] <jbailey> So I'm wondering if we should drop ia32-libs for ia64 for now, or if there's a software emulator we should be using for those cases?
[01:20] <doko> hmm, why keep it at all?
[01:21] <jbailey> Only for openoffice and such.
[01:21] <jbailey> But if we could drop it without consequence, it would save some hassle, I think.
[01:22] <doko> to drop it we need libc6-i386-dev first, *hint*, *hint* ;-P
[01:22] <fabbione> doko: hey dude
[01:22] <fabbione> hi jbailey 
[01:22] <fabbione> so when is the libstdc++++++ transition++ planned to start?
[01:23] <jbailey> Heya Fabio
[01:23] <fabbione> immediatly after flying-1 or will you give us time to stop the buildd?
[01:27] <doko> fabbione: sure, but I want to be able to build it probably before the weekend
[01:27] <fabbione> doko: for me it's enough to be able to stop the buildd before you upload the 3 billion pkgs :)
=== chmj [n=chmj@wbs-146-177-112.telkomadsl.co.za]  has joined #ubuntu-toolchain
[03:47] <lamont-away> so what's the word on new gcc?  should I start the hppa buildd back up for now?
[04:16] <doko> lamont-away: please do, we are waiting for the flight-1 CD
[06:21] <doko> jbailey: ping
[06:21] <doko> jbailey: any estimate for 2.3.6?
[06:27] <jbailey> doko: Upstream or in Ubuntu, and with which features?
[06:29] <doko> jbailey: Ubuntu, merges and libc6-i386
[06:29] <jbailey> merges will be post-flight-1, 2.3.6 from upstream will be this afternoon (with the X fix).
[06:29] <jbailey> libc6-i386 should probably be after, since it means fixing ia32-libs at the same time.
[06:31] <jbailey> doko: Anything else you need in before flight-1?
[06:33] <doko> jbailey: no, not at all
[06:33] <jbailey> Luvly.
[08:11] <lamont-away> uh, doko....
[08:11] <lamont-away> expect-tcl8.3 [!hppa !hurd-any !hurd-i386] 
[08:11] <lamont-away> from debian gcc-4.0...
[08:11] <lamont-away> which then ftbfs
[08:20] <doko> hmm, ok, it's already in the archive, fixed for the next one
[08:26] <elmo> oi, lamont
[08:26] <lamont-away> hi elmo
[08:27] <lamont-away> doko: that was gcc-4.0_4.0.2-4
[08:27] <elmo> lamont-away: did you manually upload binutils for hppa in Debian?
[08:27] <lamont-away> uh.  I don't remember doing that particularly
[08:28] <lamont-away> elmo: talking about expect b0rkage type issues or something else?
[08:28] <elmo> lamont-away: yes, I suspect so
[08:28] <elmo> both builds on the buildd failed with testsuite death
[08:28] <elmo> and yet a binary "appeared" in the archive
[08:28] <elmo> with no corresponding successful build
[08:28] <lamont-away> quite possible that I helped it along - but I don't think I manually built it, no.
[08:28] <lamont-away> who's sig on the .changes?
[08:29] <elmo> that'd be like.. effort.  can't you just admit it? :-P
[08:29] <lamont-away> (my normal involvement with expect and hppa's buildd comes in the form of 'killall -9 expect' once the build hangs
[08:29] <lamont-away> but those give you success logs...
[08:30] <lamont-away> I wish expect would either switch to tcl8.3, or they would fix the threads in 8.4
[08:32] <lamont-away> elmo: fwiw, it'd be really cool if the .changes files were publicly available somewhere.
[08:33] <lamont-away> (for both debian and ubuntu)
[08:33] <elmo> they are, on merkel
[08:33] <elmo> for debian, for ubuntu - yeah
[08:33] <lamont-away> oh, even better
[08:33] <elmo> I don't want to bloat the archive with them tho
[08:33] <doko> elmo: that was me. I tried with expect first, which failed, then expect-tcl8.3. you should have a mail in your mailbox
[08:33] <lamont-away> no sense to mirror them, that's for sure
[08:33] <lamont-away> elmo: see, I'm innocent
[08:34] <elmo> ok, doko and lamont are not allowed to go to conferences anymore
[08:34] <elmo> clearly doko has been hanging around lamont too long ;-P
[08:34] <lamont-away> elmo - I always leave tracks.
[08:34] <elmo> doko: dude, the correct fix for that is a bug, not a random upload of a package that can't be built on the buildds
[08:42] <doko> elmo: submitted.
[08:56] <lamont-away> elmo: fwiw, doko is just one more in the long line of people who "fix" hppa build failures that I've left failed by uploading something they built themselves.  go debian
[08:58] <jbailey> lamont-away: There is possibly a reasonable argument that the buildd is broken if it fails to build something that can be reasonably built in a clean chroot on a standard Debian setup.
[08:58] <elmo> jbailey: dude, this is NOT the buildd being broken
[08:58] <elmo> jbailey: and even if it was, that's not an even remotely sane argument
[08:58] <elmo> this same buildd has to be used for security builds
[08:59] <jbailey> Right, but if the buildd is configured in a different way than the standard OS (different kernel, different kernel settings), then that machine is misconfigured in an important way
[08:59] <elmo> and should be fixed
[08:59] <elmo> not worked around
[08:59] <jbailey> I'm not saying the right soluton is to build it yourself and upload it.
[09:00] <jbailey> I'm arguing that having alignment checking turned on by default on the buildd machines for hppa and not on the standard hppa kernel is probably a bug.
[09:00] <jbailey> Among other tweaks that might be done to various buildds.
[09:00] <jbailey> rather alignment checking turned on for standard builds, and not turned on for the buildd.
[09:00] <elmo> yeah, fine I can agree with that
[09:00] <jbailey> But you know what I mean.
[09:01] <elmo> but one day, I really am going to block non-buildd uploads in debian too :-P
[09:01] <lamont-away> jbailey: yeah...
[09:01] <lamont-away> but I can't really file a bug against camm for having broken thinking in how he does packaging.
[09:01] <lamont-away> although i suppose I could file a bug against the b0rked packages - not that it would change much
[09:02] <jbailey> elmo: And I will phear i386 on Debian so much less when you do. =)
[09:02] <jbailey> lamont-away: Hmm, bug queues on people.
[09:02] <jbailey> lamont-away: I wonder if there should be some ritual when people show up at Debconf where the people with the most bugs on them get thrown out on the assumption that they must be mouldy and rotten?
[09:03] <jbailey> Or perhaps we can be gentler and write "corpse" on their forehead with indelible marker? =)
=== lamont-away had a group of friends that had George Takei sign someone's forehead once...
[09:05] <lamont-away> with a sharpie.  the day before a job interview.  upside down.
[09:07] <elmo> what's a sharpie?
[09:07] <lamont-away> permanent marker
[09:07] <lamont-away> "sharpie" is a brand name
[09:08] <lamont-away> the person was up for several hours turning his autographed forehead into a red not-quite-bloody forehead.
[09:09] <elmo> oh
[09:09] <elmo> I assume it was like those compasses you had at school, the one with the sharp metal ends
[09:09] <elmo> which would have been, like, harsh
[09:11] <lamont-away> yeah, well, they would have had a hard time talking George into it, too... as it was, they just bodily carried jack over and said "Mr Takei, would you please sign this?  thx"
[09:13] <lamont-away> hrm.. /me realizes that signing foreheads has nothing to do with toolchains anywhere, let alone ubuntu
=== doko_ [n=doko@dslb-084-059-086-237.pools.arcor-ip.net]  has joined #ubuntu-toolchain