[00:13] <mwhudson> well i've found the binutils commit that's to blame
[00:13] <mwhudson> 32ec889602502348b704cfb16e65c83dc3eec095
[00:13] <mwhudson> "Tidy up readelf's use of boolean values."
[00:37] <Unit193> mwhudson: Ahaha!  Awesome timing, just now I get an email about it being stuck in proposed. :D
[00:38] <mwhudson> Unit193: heh
[00:39] <mwhudson> Unit193: gocryptfs?
[00:39] <Unit193> Yep.
[00:39] <mwhudson> the bot is a harsh mistress
[01:43] <blahdeblah> There is a brief outage on *.qa.ubuntu.com coming up shortly; apologies for any inconvenience.
[01:44] <blahdeblah> (also includes loco.ubuntu.com)
[01:45] <Unit193> That's just loco.
[01:46] <blahdeblah> Unit193: what do you mean?
[01:46] <Unit193> Bad joke, sorry.
[01:47] <blahdeblah> Some of us non-South Americans are a little slow
[01:47] <blahdeblah> (Or is that all Spanish speakers?  Shows how ignorant I am...)
[01:48] <Unit193> Should be a Spanish thing, yeah.  Also, Ohioian here, very north America.  Thanks for letting us know about the downtime.
[01:50] <blahdeblah> It's long since done, BTW
[03:34] <Logan> rbasak, smoser: would you mind taking a look at my merge request for scim-chewing when you get the chance? https://code.launchpad.net/~logan/ubuntu/+source/scim-chewing/+git/scim-chewing/+merge/327575
[03:35] <Logan> also, what's the process for becoming a member of the Ubuntu Server Dev import team?
[07:38] <LocutusOfBorg> mapreri, any idea for diffoscope autopkgtestsuite? it is blocking a lot of python stuff...
[07:38] <mapreri> yes
[07:38] <mapreri> LocutusOfBorg: it's due to this: https://sourceware.org/bugzilla/show_bug.cgi?id=21820
[07:38] <LocutusOfBorg> so, regression in binutils?
[07:39] <mapreri> Also reported as Debian #869532 for diffoscope
[07:40] <LocutusOfBorg> thanks :(
[07:42] <mapreri> LocutusOfBorg: do you have any idea about why binutils didn't trigger a diffoscope autopkgtest, and wasn't blocked in proposed?
[07:43] <mapreri> in general it feels like there are fewer tests than expected
[07:57] <ricotz> hi, there are still quite a bunch of rebuilds to be done -- http://people.canonical.com/~ubuntu-archive/transitions/html/python3.6.html
[08:13] <rbasak> Logan: it's on my team's list to take a look. Sorry it hasn't bubbled to the top yet.
[08:48] <LocutusOfBorg> ricotz, maybe people are waiting for autopkgtestsuite become green?
[08:48] <LocutusOfBorg> not sure who is handling that python transition
[08:51] <xnox> the new binutils is blocked in proposed and not migrating
[08:54] <ricotz> LocutusOfBorg, hopefully not, I guess it is mwhudson
[08:58] <rbasak> cyphermox: any news on bug 1624519 please? It unsnoozed on my board.
[09:01] <mwhudson> ricotz: huh i hadn't seen that that had been updated
[09:01] <mwhudson> mapreri: yeah that binutils thing is heaps of fun :(
[09:01] <LocutusOfBorg> mwhudson, revert that commit for binutils? :)
[09:02] <mwhudson> LocutusOfBorg: guess what the debdiff i have here does!
[09:02] <LocutusOfBorg> just to make things migrate, otherwise a lot of stuff including python is blocked
[09:02] <LocutusOfBorg> lol wonderful!
[09:02] <LocutusOfBorg> mwhudson, do you want some help in python? I have not much knowledge, but I can allocate some hours there probably
[09:03] <LocutusOfBorg> Its blocking my sane-backends gsoap and something more transitions :)
[09:03] <mwhudson> LocutusOfBorg: sure
[09:03] <LocutusOfBorg> and meh, there is a lot of perl fun that is waiting for it :)
[09:03] <mwhudson> there's not much stopping python3-defaults from migrating really
[09:04] <mwhudson> but knocking some things off the transition tracker would be nice
[09:04] <mwhudson> careful, some are kind of false hits though
[09:04] <mwhudson> e.g. boost i think
[09:05] <mwhudson> LocutusOfBorg: if you want to figure out what's going on with botch......
[09:05] <LocutusOfBorg> python-pex seems to be easily fixable
[09:05] <LocutusOfBorg> mwhudson, botch is out of my knowledge
[09:05] <mwhudson> yeah that one looks easy
[09:05] <LocutusOfBorg> I already tried
[09:05] <LocutusOfBorg> :)
[09:06] <mwhudson> i'm surprised the package builds, but it does
[09:06] <mwhudson> LocutusOfBorg: dammit
[09:06] <LocutusOfBorg> I read your mail to ubuntu-devel list, and looked at it already, but I'll do a deeper look
[09:06] <mwhudson> ubuntu-make is confusing me
[09:06] <LocutusOfBorg> that one too, meh, strange
[09:06] <LocutusOfBorg> only on some archs, it makes no sense
[09:07] <LocutusOfBorg> maybe a meld of the good vs bad arch can show some stuff with new version built only partially
[09:11] <LocutusOfBorg> pex uploaded, lets hope
[09:11]  * LocutusOfBorg leaves changing network
[09:14] <mapreri> mwhudson: the problem wrt diffoscope is that its behaviour when an external command fails is quite hardwired, and 1. changing it only for the elf comparator is just not possible without rewriting everything or doing very nasty hacks 2. changing it for all comparators is sure to have unintended effects
[09:14] <mapreri> I suppose we could figure something out, but…
[09:25] <mwhudson> mapreri: that was the impression i got
[09:25] <mwhudson> mapreri: a wrapper around readelf? :)
[09:25] <mapreri> mwhudson: That's an example of nasty hack :3
[09:26] <mapreri> besides, the elf comparator is so complex by itself
[09:26] <mwhudson> mapreri: people working on transitions get good at nasty hacks :)
[09:26] <mapreri> why are elf files so complex
[09:26] <mapreri> everything should be an interpreted language
[09:26] <Unit193> Sloooow.
[09:34] <mwhudson> zip should be good enough for everyone
[09:42] <ricotz> mwhudson, regarding python-transition: boost is not a false positive
[10:18] <LocutusOfBorg> mwhudson, python-pex is fixed, and I emailed the patch to Barry
[10:29] <doko> jamespage: #1698350, missing bug subscribers
[10:30]  * jamespage looks
[10:30] <jamespage> bug logfile not supported in this QEMU binary
[10:30] <jamespage> bug 1698350
[10:33] <jamespage> doko: done - commented on bug
[10:36] <LocutusOfBorg> mwhudson, for some reasons pex is not fixed
[10:42] <doko> jbicha: I see you changed the myspell-* seed stuff in supported. I'd like to demote dict-nr dict-ns dict-ns dict-st dict-tn dict-ts dict-xh (ftbfs and not updated for years). How would that be done?
[10:42] <doko> Laney: ^^^
[10:52] <LocutusOfBorg> mwhudson, fixed now, the testsuite was running correctly and returning 120, what a mess
[10:57] <mwhudson> ricotz: oh right i guess it needs a rebuild for that one extension module it rebuilds
[11:00] <mwhudson> mapreri, doko: here's the binutils patch that might fix the diffoscope build http://paste.ubuntu.com/25161894/
[11:00] <mwhudson> (testing diffoscope build with those packages now)
[11:00] <mapreri> that's basically a revert of the regressing upstream commit, right?
[11:02] <doko> mwhudson: is this just reverting an upstream commit?
[11:03] <doko> is this somewhere tracked?
[11:03] <mapreri> doko: https://sourceware.org/bugzilla/show_bug.cgi?id=21820
[11:04] <mwhudson> doko: yeah it's reverting the commit mentioned in the bugzilla
[11:04] <mwhudson> (with some fixups)
[11:05] <mwhudson> oh ffs doesn't fix the build??
[11:05] <mwhudson> ah no, holding it wrong
[11:06] <mwhudson> i love[0] the way sbuild --extra-package x does not fail if x does not exist [0] may be a lie
[11:06] <doko> ta, tracking the issue
[11:07] <ginggs> doko: hi, any objections to a tbb sync? i have test-built a couple of rdeps, and i'll keep a lookout for any new failures
[11:07] <mwhudson> i noticed that readelf exits with code 0 if you try to read a *missing* section
[11:07] <doko> mwhudson: can you attach the object file to the bug report, just in case ...
[11:07] <mwhudson> so exiting with code 1 for a present, empty, section must be a bug
[11:07] <doko> ginggs: from experimental?
[11:08] <ginggs> doko: yes, from experimental, including the multiarch
[11:08] <doko> ginggs: sounds fine. we need it for GCC 7 anyway
[11:08] <ginggs> doko: ta
[11:11] <mwhudson> doko: done
[11:15] <mwhudson> well that patch doesn't fix the build completely :(
[11:15] <mwhudson> mapreri: does http://paste.ubuntu.com/25161949/ mean anything to you?
[11:16] <mapreri> yes, that should be fixed with the last upload I think
[11:16] <mapreri> at least I remember noticing it, and doing something to fix it.
[11:16] <mapreri> ah, no, only in git still :(
[11:16] <mapreri> mwhudson: https://anonscm.debian.org/git/reproducible/diffoscope.git/commit/?id=c8927db98131a129c5d7354bfe8a6b0886ca66f8 + https://anonscm.debian.org/git/reproducible/diffoscope.git/commit/?id=e62306e17e8a9b981d3b7f427bc99f8582a0301a is what I did to fix it.
[11:17] <mapreri> I thought that was released already, but apparently I did it right after the last release was cut :S
[11:17] <mapreri> (+ https://anonscm.debian.org/git/reproducible/diffoscope.git/commit/?id=c12bee5fc718c918d6280102d4dbe688b0a553cf)
[11:27] <mwhudson> mapreri: time for an upload then? :)
[11:27]  * mwhudson goes to bed
[11:27] <mapreri> mwhudson: currently git doesn't pass the test as somebody did an incomplete addition... also would love to fix the autopkgtest also in debian :S
[11:50] <hjd> Looks like the jruby and ruby-psych packages have a circular dependency on artful https://launchpad.net/ubuntu/artful/amd64/+builds?build_text=&build_state=depwait
[11:51] <hjd> Anyone around to do a bootstrap or should I file a bug report on it? :)
[12:12] <ahasenack> hi, could someone please sponsor https://bugs.launchpad.net/ubuntu/+source/samba/+bug/1644428 ?
[12:12] <LocutusOfBorg> mwhudson, done
[12:12] <ahasenack> it's a 2nd time regression, and a patch/MP has been up for weeks
[12:13] <LocutusOfBorg> ahasenack, zesty-updates version + patch on comment #32?
[12:13] <ahasenack> LocutusOfBorg: yes (the MP has been up for longer, but that's an internal problem I think that no-one looks at it)
[12:14] <LocutusOfBorg> Hunk #1 FAILED at 1.
[12:14] <LocutusOfBorg> 1 out of 1 hunk FAILED -- saving rejects to file debian/changelog.rej
[12:14] <LocutusOfBorg> please rebase it
[12:14] <ahasenack> of course...
[12:14] <ahasenack> story of my life :)
[12:14] <ahasenack> will get to it
[12:15] <LocutusOfBorg> meh I can do it
[12:15] <LocutusOfBorg> nvm
[12:15] <LocutusOfBorg> debdiff is trivial
[12:17] <jbicha> doko: the 'supported' seed was updated to match language-selector's data/pkg_depends
[12:18] <ahasenack> LocutusOfBorg: thanks
[12:18] <LocutusOfBorg> done yw
[12:21] <LocutusOfBorg> mwhudson, isn't the botch failure some sort of missing sorting for the output?
[12:41] <rbasak> jbicha: o/
[12:42] <rbasak> jbicha: in bug 1698553, are you saying that the ppc related bit of the diff definitely isn't a regression? Or are you not commenting on that bit?
[12:44] <rbasak> I suppose if autoreconf has always been running then it shouldn't matter, but I'm not sure exactly how to confirm that.
[12:49] <jbicha> you can confirm autoreconf is running by checking the build logs, here's the artful build which should be pretty similar
[12:49] <jbicha> https://launchpad.net/ubuntu/+source/gjs/1.48.5-0ubuntu1
[12:49] <jbicha> I don't know anything about what autotools does to ppc except that 1.48.5 built on artful for all supported Ubuntu arches
[12:50] <jbicha> dh simple rules with debhelper compat 10 in zesty will run dh_autoreconf unless it's turned off (which isn't the case here)
[12:52] <rbasak> by "that", I meant the "shouldn't matter" part, not that autoconf has always been running. Sorry I wasn't clear.
[13:38] <smoser> Logan, i'll read... process... we dont really have a process at the moment.  the three members were just the people that kind of kick-started this. ultimately the goal is to have bot-only doing imports.
[13:39] <smoser> Logan, still there ?
[13:42] <ricotz> jbicha, hi, regarding mozjs52 I haven't spot something obvious yet
[13:43] <ricotz> jbicha, the "autoreconf" patch is done with a ubuntu/debian toolchain?
[13:45] <jbicha> ricotz: it came from https://github.com/ptomato/mozjs/commit/740bc5d9d66433370ba4137d03390d5ee21f26b4
[13:50] <smoser> i think i misunderstood...
[13:53] <ricotz> jbicha, so you don't know, is autoconf2.13 going away? otherwise I would suggest to keep the "real autoreconf"
[14:31] <xnox> mwhudson, all i can say is that even diffoscope master FTBFS with old binutils and python3.6
[14:31] <xnox> https://launchpadlibrarian.net/330406667/buildlog_ubuntu-artful-amd64.diffoscope_85~ubuntu1_BUILDING.txt.gz
[15:05] <ginggs> doko: nice, your blender rebuild exercised the new tbb
[17:08] <ahasenack> after verifying the fix for all affected releases, should I flip the global "verification-needed" tag to "verification-done"? Case in point: https://bugs.launchpad.net/ubuntu/+source/openvpn-auth-ldap/+bug/1602813
[17:08] <ahasenack> I did set verification-done-$release for all affected releases after testing
[17:49] <tsimonq2> Logan: ack
[18:32] <bdmurray> Why does autopkgtest end up using a lot of space on my / mount when I am using a qemu image on a different mount point?
[18:48] <xnox> i thought it does a cow snapshot, to keep the golden image unmodified, and stores the copy-on-write layer somewhere in /tmp no?
[18:48] <bdmurray> Could be, I haven't dug into it.
[20:29] <tsimonq2> Is there a Core Developer available that would be able to help me with a Qt transition? I need a couple more rebuilds in the PPA then someone will need to land it. Details here: https://bileto.ubuntu.com/#/ticket/2819
[20:31] <tsimonq2> Before that's landed though, I think we need to get the rest of the Unity 8 removals done so that there's minimal impact on the transition.
[20:31] <tsimonq2> mitya57 went on vacation, so I'll be on my own for this one.
[20:31] <tsimonq2> He told me before he left that slangasek had more information... what's new?
[21:48]  * mwhudson is lost in a maze of britney output :(
[21:48] <mwhudson> why is hunspell not migrating?
[21:51] <jbicha> mwhudson: because of libreoffice which depends on python3-defaults
[21:51] <mwhudson> ah
[21:51] <mwhudson> same old story
[21:53] <mwhudson> i think the link-grammar autopkgtests are now failing because the build the package with some of proposed enabled and then switch to running it with all of propsed
[21:53] <mwhudson> or something along those lines
[21:57] <mwhudson> need an aa to apply some force to this migration i think
[22:51] <mwhudson> ok lets see if i can fix ubuntu-make today
[23:04] <jbicha> LO i386's autopkgtest would need to be ignored too