| xnox | infinity, i believe desktop team maintains that package. I am not going to touch that anymore, as my upload was reverted by them. | 00:24 |
|---|---|---|
| xnox | (upload for an indicator) | 00:24 |
| slangasek | LocutusOfBorg: hey, so you've merged remmina, which has changed build-deps from libfreerdp-dev to freerdp2-dev and wants to pull in several dozen packages into main including ffmpeg and that's not happening. How should this be handled? | 04:07 |
| Ukikie | Didn't Ubuntu GNOME drop remmina? | 04:11 |
| slangasek | I don't know what Ubuntu GNOME did; it's still listed as a recommends of ubuntu-desktop | 04:20 |
| jbicha | slangasek: freerdp2 is buildable without ffmpeg so I guess I'll upload that | 04:29 |
| slangasek | ok | 04:29 |
| jbicha | "Skipping recommended feature FFmpeg for multimedia (multimedia redirection, audio and video playback)" | 04:38 |
| jbicha | but it turns out we already do that in our freerdp packaging | 04:39 |
| slangasek | jbicha: remmina-plugin-rdp -> libfreerdp2-2 -> libavcodec57; doesn't look built without ffmpeg | 04:57 |
| slangasek | and gr-iio -> libiio -> mono FTBFS on ppc64el+s390x, entangled in icu transition | 04:58 |
| jbicha | slangasek: I meant "turns out we already do that in our freerdp1 packaging", I just uploaded fixed freerdp2 now | 05:05 |
| slangasek | ah ok | 05:06 |
| Ukikie | slangasek: Ah sorry, I was referring to the GNOME version of Ubuntu, thus Ubuntu proper/Mainbuntu. | 05:20 |
| jbicha | Ukikie: as of 17.10 (and current 18.04), remmina is still included in Ubuntu Desktop by default | 05:21 |
| Ukikie | jbicha: OK, it was proposed to remove it, IIRC? | 05:21 |
| jbicha | Ukikie: yes I proposed it 😉 | 05:23 |
| jbicha | slangasek: are you ok with letting libgsm back in main (it's needed by freerdp2)? | 05:23 |
| slangasek | jbicha: hmm, it seems there was never previously an MIR for it; looking closer | 05:25 |
| slangasek | "Compression and decompression run faster than realtime on most SPARCstations" be still my heart | 05:27 |
| slangasek | jbicha: if it's really needed, yes. I'm unclear what speech compression has to do with RDP | 05:28 |
| jbicha | slangasek: never mind, I found the knob to turn libgsm off | 05:34 |
| jbicha | apologies to sparc fans | 05:34 |
| slangasek | I think there might be a few other audio codecs that decompress in realtime on current CPUs | 05:35 |
| slangasek | doko: kido ftbfs on ppc64el shows a stacktrace pointing at c++ w/ std::basic_string in liburdfdom-world; the segfault is corrected with a no-change rebuild of urdfdom; do you know if this is safe to do as a no-change rebuild, or does this point to ABI breakage of some sort? | 05:46 |
| slangasek | infinity: mono ftbfs on ppc64el points to 'struct ucontext' not being exposed via #include <ucontext.h>. Mono bug, or glibc bug for asm/ucontext.h not being transitively included? | 05:54 |
| slangasek | infinity: hmm possibly more subtle than that; but still, wondering whose bug it is | 06:00 |
| LocutusOfBorg | slangasek, thanks for caring and fixing, I briefly looked at it yesterday, but I did something else... | 06:35 |
| LocutusOfBorg | should we have both freerdp and freerdp2 in main then? | 06:35 |
| LocutusOfBorg | I would say only remmina is in main, so freerdp can go in universe... | 06:38 |
| doko | slangasek: mono bug, was removed in glibc by intent | 09:13 |
| doko | slangasek: kido/urdfdom, not aware of changes in std::basic_string between GCC 6 and 7 | 09:18 |
| -queuebot:#ubuntu-release- Unapproved: accepted indicator-sound-gtk2 [source] (xenial-proposed) [12.10.0.1-0ubuntu5.16.04.1] | 09:30 | |
| infinity | doko, slangasek: Worked around the mono ftbfs by just removing it (it was pretty much a no-op upload anyway). z3, libiio, and uwsgi now building. | 09:33 |
| infinity | (We'll need to double back on the mono thing, of course) | 09:33 |
| doko | infinity: do you know the current status of entangled transitions? | 09:40 |
| infinity | doko: Trying to determine that right now. | 09:40 |
| doko | what I see is: icu libcdio lincdio-paranoia netcdf opencv mrpt gnuradio | 09:42 |
| infinity | Yeah, and opencv/mrpt seems to have some FTBFS sadness. | 09:42 |
| infinity | So glad that one got started and entangled. :/ | 09:42 |
| infinity | doko: If mrpt is tied up anyway, should we try a sync? :/ | 09:45 |
| infinity | 1:1.5.3-1 is in testing. | 09:45 |
| doko | sure, why not. | 09:46 |
| infinity | Or, it has no rdeps? Or is reverse-depends lying to me? | 09:46 |
| * infinity crosses his fingers. | 09:46 | |
| doko | sil2100: what's status about fixing these unity related ftbfsses | 09:46 |
| doko | opencv is sadness ... | 09:47 |
| sil2100 | I'm looking at biometryd now... | 09:53 |
| === maclin1 is now known as maclin | ||
| doko | xnox: to get libcolumbus disentangled with boost I could build-depend on the old versioned boost b-d's | 10:14 |
| xnox | doko, let me check if it is a CMake module problem, and anything that uses boost-python via cmake is broken. | 10:15 |
| xnox | doko, hm.... that does not help. | 10:20 |
| -queuebot:#ubuntu-release- New binary: mrpt [ppc64el] (bionic-proposed/universe) [1:1.5.3-1] (no packageset) | 10:23 | |
| ginggs | would someone bump the freecad hint in a-p-w's hints please? | 10:25 |
| doko | tyhicks: fgrep -ri copyright shows more than one copyright holder ... | 10:25 |
| xnox | doko, seems like PyInit got.... namespaced? | 10:26 |
| xnox | 37: 000000000021d180 8 OBJECT LOCAL DEFAULT 25 _ZGVZ15PyInit_columbusE9m | 10:26 |
| xnox | 38: 000000000021d0c0 104 OBJECT LOCAL DEFAULT 24 _ZZ15PyInit_columbusE9mod | 10:26 |
| xnox | 39: 0000000000013b20 40 OBJECT LOCAL DEFAULT 14 _ZZ15PyInit_columbusE14in | 10:26 |
| xnox | 43: 000000000021d1a0 32 OBJECT LOCAL DEFAULT 25 _ZZ15PyInit_columbusE15in | 10:26 |
| xnox | ginggs, make a bzr merge proposal? | 10:26 |
| apw | ginggs, looking | 10:27 |
| apw | ginggs, done | 10:28 |
| ginggs | apw: thanks! | 10:28 |
| sil2100 | huh, looks like my change helped, the build is moving forward | 10:29 |
| -queuebot:#ubuntu-release- New binary: mrpt [s390x] (bionic-proposed/universe) [1:1.5.3-1] (no packageset) | 10:29 | |
| sil2100 | I wonder why it suddenly stopped working in bionic, this wasn't related to boost or anything else obvious | 10:30 |
| doko | xnox: rather boosted. That function has to be a C function | 10:30 |
| xnox | yeah | 10:30 |
| sil2100 | doko, xnox: ok, so biometryd is moving forward now as cmake is not failing anymore, but there's some sign-compare C++ errors now coming from gtest, I'll be digging further | 10:33 |
| ginggs | apw: you missed build3/ppc64el | 10:37 |
| -queuebot:#ubuntu-release- New binary: mrpt [i386] (bionic-proposed/universe) [1:1.5.3-1] (no packageset) | 10:38 | |
| LocutusOfBorg | does anybody know why plplot segfaults on pppc64el? this is probably blocking the transition too | 10:39 |
| -queuebot:#ubuntu-release- Unapproved: openvswitch (xenial-proposed/main) [2.5.2-0ubuntu0.16.04.3 => 2.5.4-0ubuntu0.16.04.1] (ubuntu-server) | 10:40 | |
| apw | ginggs, re-fixed | 10:41 |
| ginggs | apw: re-thanks | 10:41 |
| apw | ginggs, also reformed so i don't do that again | 10:41 |
| -queuebot:#ubuntu-release- New binary: mrpt [amd64] (bionic-proposed/universe) [1:1.5.3-1] (no packageset) | 10:42 | |
| LocutusOfBorg | doko, feeling perl mergy? | 10:53 |
| LocutusOfBorg | after icu and so on I guess? | 10:54 |
| doko | LocutusOfBorg: I really would prefer if you would finish all the transitions you are starting ... | 10:54 |
| LocutusOfBorg | casablanca is over, and the others are entangled with other transitions | 10:54 |
| LocutusOfBorg | my transitions are all over I guess | 10:54 |
| xnox | horum... pre-processed code looks ok http://paste.ubuntu.com/25850991/ | 10:55 |
| xnox | building without -fvisibility=hidden -fvisibility-inlines-hidden | 10:57 |
| infinity | LocutusOfBorg: Define "over". They're only over once they've migrated, otherwise you just keep entangling new transitions with current ones. | 10:58 |
| slashd | sil2100, good day ;) could you please release lshw for both : T & X | (LP:# 1471983) | 10:58 |
| sil2100 | slashd: o/ Let me take a lookie | 10:58 |
| LocutusOfBorg | my subtransitions are all ready to migrate, waiting for icu and opencv and so on | 10:58 |
| xnox | doko, in libcolumbus -fvisibility=hidden -fvisibility-inlines-hidden -> seems to be what is breaking it. Pre-processed code looks ok. | 10:58 |
| LocutusOfBorg | I had to sync e.g. casablanca because the current one was boost-foo bad building | 10:59 |
| LocutusOfBorg | I did start only the one that were fixing boost failures | 10:59 |
| LocutusOfBorg | and I finished them, even if not migrating, they are not blocking migration | 10:59 |
| doko | boost is non-blocking | 10:59 |
| LocutusOfBorg | so what is blocking now? I lost the picture... mrpt seems mostly good | 11:00 |
| doko | opencv gnuradio | 11:03 |
| infinity | openimageio's new FTBFS needs love. blender's sudden hatred of itself as well. | 11:04 |
| LocutusOfBorg | opencv is worked out by mapreri I think | 11:05 |
| LocutusOfBorg | I stopped merging/syncing gnuradio some releases ago, just because... it sucks | 11:05 |
| doko | ginggs: https://launchpadlibrarian.net/343506262/buildlog_ubuntu-bionic-amd64.caffe-contrib_1.0.0-4build1_BUILDING.txt.gz | 11:06 |
| infinity | LocutusOfBorg: The point being made here is that if you start a transition that gets entangled in another one, you've just made yourself partially responsible for all of them. :P | 11:06 |
| infinity | You don't get to say "oh, well, my part's done, can I start another that makes things worse?" | 11:06 |
| infinity | And the tracker hasn't updated for 2 hours. Hrm. | 11:07 |
| doko | did I break the transition tracker with syntax errors? | 11:07 |
| infinity | W: parse error in file "config/monitor/ongoing/opencv.ben", line 5, character 0 | 11:07 |
| infinity | Uncaught exception: ben-specific error: parse error in file "config/monitor/ongoing/opencv.ben", line 5, character 0 | 11:07 |
| infinity | doko: Yup. | 11:07 |
| doko | how can I see these errors? | 11:08 |
| -queuebot:#ubuntu-release- New binary: gr-radar [ppc64el] (bionic-proposed/universe) [0.0.0.20161018-3] (no packageset) | 11:08 | |
| infinity | doko: That's in ubuntu-archive@snakefruit's mailbox. | 11:08 |
| -queuebot:#ubuntu-release- New binary: gr-radar [amd64] (bionic-proposed/universe) [0.0.0.20161018-3] (no packageset) | 11:09 | |
| doko | so maybe have a look at the gnuradio tracker as well | 11:09 |
| LocutusOfBorg | infinity, hey, I didn't say "I'm washing my hands", I lost a whole sunday trying to sort out ocaml | 11:09 |
| -queuebot:#ubuntu-release- New binary: gr-radar [i386] (bionic-proposed/universe) [0.0.0.20161018-3] (no packageset) | 11:10 | |
| -queuebot:#ubuntu-release- New binary: gr-radar [s390x] (bionic-proposed/universe) [0.0.0.20161018-3] (no packageset) | 11:10 | |
| LocutusOfBorg | I'm just pointing out that when I start something, I finish it, even if subpartially, and I'm trying to sort out that icu stuff since the begin, with some little help where I can | 11:10 |
| xnox | doko, i do think this is a boost bug. due to changes in boost config, it no longer sets visibility to default. | 11:10 |
| LocutusOfBorg | because I want the autosync open :D | 11:10 |
| infinity | doko: Looks like it's just missing a trailing semicolon on is_bad. Hence the syntax error being char 0 of the next line. | 11:14 |
| infinity | doko: Fix commited. | 11:16 |
| ginggs | doko: ack caffe-contrib - will look soon | 11:17 |
| LocutusOfBorg | ginggs, I was looking at it | 11:17 |
| LocutusOfBorg | libkf5kface --> there is a version in debian experimental, maybe it is already fixeed, we can cherry-pick | 11:18 |
| LocutusOfBorg | openimageio needs a merge? | 11:18 |
| LocutusOfBorg | sigh | 11:18 |
| ginggs | LocutusOfBorg: don't stop :) | 11:18 |
| -queuebot:#ubuntu-release- New binary: gr-radar [arm64] (bionic-proposed/universe) [0.0.0.20161018-3] (no packageset) | 11:19 | |
| ginggs | ^looking at caffe-contrib that is | 11:19 |
| -queuebot:#ubuntu-release- New binary: gr-radar [armhf] (bionic-proposed/universe) [0.0.0.20161018-3] (no packageset) | 11:21 | |
| infinity | LocutusOfBorg: Testing the experimental kf5kface here. | 11:21 |
| infinity | LocutusOfBorg: If it builds okay in bionic, I'll just sync it -- it has no rdeps. | 11:21 |
| infinity | (If it doesn't, I'll demote it, same reason) | 11:21 |
| LocutusOfBorg | infinity, nah | 11:25 |
| LocutusOfBorg | debdiff is showing 0 changes | 11:25 |
| LocutusOfBorg | who in the world can do this? | 11:25 |
| LocutusOfBorg | #if OPENCV_TEST_VERSION(3,1,0) | 11:25 |
| LocutusOfBorg | so, with opencv 3.2.0 the #else branch is taken | 11:25 |
| LocutusOfBorg | booooh kf5kface | 11:25 |
| * LocutusOfBorg https://launchpad.net/~costamagnagianfranco/+archive/ubuntu/locutusofborg-ppa/+build/13644881 | 11:26 | |
| LocutusOfBorg | that patch is a good start, but not enough | 11:29 |
| -queuebot:#ubuntu-release- New binary: mrpt [arm64] (bionic-proposed/universe) [1:1.5.3-1] (no packageset) | 11:30 | |
| -queuebot:#ubuntu-release- Unapproved: openvswitch (artful-proposed/main) [2.8.0-0ubuntu2 => 2.8.1-0ubuntu0.17.10.1] (ubuntu-server) | 11:33 | |
| -queuebot:#ubuntu-release- New: accepted gr-radar [amd64] (bionic-proposed) [0.0.0.20161018-3] | 11:33 | |
| -queuebot:#ubuntu-release- New: accepted gr-radar [armhf] (bionic-proposed) [0.0.0.20161018-3] | 11:33 | |
| -queuebot:#ubuntu-release- New: accepted gr-radar [ppc64el] (bionic-proposed) [0.0.0.20161018-3] | 11:33 | |
| -queuebot:#ubuntu-release- New: accepted gr-radar [arm64] (bionic-proposed) [0.0.0.20161018-3] | 11:33 | |
| -queuebot:#ubuntu-release- New: accepted gr-radar [s390x] (bionic-proposed) [0.0.0.20161018-3] | 11:33 | |
| -queuebot:#ubuntu-release- New: accepted gr-radar [i386] (bionic-proposed) [0.0.0.20161018-3] | 11:33 | |
| LocutusOfBorg | infinity, what about merging openimageio? it seems to have the opencv fixes, but they seems to be changing API | 11:38 |
| LocutusOfBorg | so, doing that transition will fix the current one and blender, but meh, I prefer another look | 11:38 |
| * LocutusOfBorg tries in a ppa | 11:38 | |
| sil2100 | doko: I have a full-fix for biometryd against bionic and the new boost, will be batching it up with others that hopefully have the same problems | 11:53 |
| sil2100 | Will be released after I get back from lunch | 11:53 |
| jbicha | if an SRU is rejected from the unapproved queue, can it be pushed back to the unapproved queue or can it only be pushed to accepted? | 13:31 |
| infinity | jbicha: The latter. | 13:32 |
| LocutusOfBorg | infinity, permission to upload openimageio merge? | 13:32 |
| infinity | LocutusOfBorg: NACK, I already have it staging. | 13:32 |
| infinity | LocutusOfBorg: (And opencolorio) | 13:32 |
| LocutusOfBorg | I didn't parse your sentence, but in case you are interested... https://launchpad.net/~costamagnagianfranco/+archive/ubuntu/locutusofborg-ppa/+sourcepub/8432927/+listing-archive-extra | 13:34 |
| LocutusOfBorg | error: namespace "std" has no member "signbit" | 13:40 |
| LocutusOfBorg | no idea for caffe-contrib | 13:40 |
| infinity | Wish I could figure out why 'blender --help' now hangs on 3/6 arches... | 13:41 |
| infinity | xnox: ^-- I want to blame boost. Can I? | 13:42 |
| xnox | infinity, sure =) it does use boost-python, which may have a small regression (failing to import modules) | 13:45 |
| infinity | xnox: Failing to import modules doesn't seem like it would cause a futex wait hang. | 13:46 |
| infinity | But maybe? :P | 13:47 |
| xnox | infinity, there are loads of UI things that are actually boost-python plugins in blender.... if i recall correctly | 13:48 |
| xnox | which do generate / inject help options.... | 13:48 |
| infinity | xnox: Well, --help on s390x between the old and new binaries seem identical, save that one hangs and the other doesn't. :/ | 13:48 |
| xnox | lovely | 13:48 |
| infinity | Oh, but this is pretty suspect... | 13:49 |
| infinity | http://paste.ubuntu.com/25851814/ | 13:50 |
| infinity | xnox: ^-- I'm going to guess that linking against two versions is Bad? | 13:50 |
| xnox | infinity, yes, that is bad very bad | 13:50 |
| infinity | That might be the problem. | 13:50 |
| infinity | But also WTF. | 13:50 |
| xnox | i guess it was built too early / has transitive deps | 13:50 |
| infinity | xnox: It literally just built. | 13:50 |
| infinity | So dunno about too early. | 13:50 |
| infinity | Ahh, openimageio. | 13:52 |
| infinity | Okay, this might self-solve shortly. | 13:52 |
| === NCommander is now known as KD2JRT | ||
| xnox | infinity, and it builds and links because we do bump the abi on all of boost, every point release =/ because boost. | 13:55 |
| infinity | xnox: I guess the more interesting question that I don't expect an answer to is if the double-linked boost-* turns out to be the problem, why is it only a problem on 3/6 arches? | 13:56 |
| infinity | But meh. | 13:57 |
| infinity | Life's too short to investigate that, if this solves it. | 13:57 |
| infinity | But I can definitely see how linking to two things called "libboost_thread" could lead to a futex going AWOL. | 13:59 |
| xnox | the three are kvm, the other two are containers, and arm64 is not done testing yet, and boost adt tests need full vm for X? | 13:59 |
| infinity | xnox: Nothing to do with testing, this is during the build. | 13:59 |
| infinity | xnox: All of which are kvm, except for s390x. | 13:59 |
| xnox | armhf is containers | 13:59 |
| xnox | no? | 14:00 |
| infinity | No. | 14:00 |
| xnox | infinity, is this built-time tests, rather than adt? | 14:00 |
| infinity | I mean, all of them are now lxd-on-kvm, for other weird reasons. But they're all the same. | 14:00 |
| infinity | xnox: This is running "blender --help" directly after the build (in the build chroot). | 14:01 |
| xnox | fun | 14:01 |
| infinity | Amd amd64, armhf, and i386 succeed. arm64, ppc64el, and s390x fail. | 14:01 |
| infinity | s/fail/hang/ | 14:01 |
| infinity | But I'm going to stop wasting brain cycles on it until after I've tested a non-double-linked build. | 14:02 |
| xnox | right. yeah, let's do that. | 14:02 |
| infinity | Which need opencolorio built against the new openimageio. Grind, grind... | 14:03 |
| LocutusOfBorg | caffe-contrib is fixed | 14:06 |
| infinity | LocutusOfBorg: \o/ | 14:06 |
| LocutusOfBorg | sigh ginggs already did it | 14:07 |
| LocutusOfBorg | ginggs, did you use -D_GLIBCXX_USE_C99_MATH=1? | 14:07 |
| ginggs | LocutusOfBorg: no i did horrible hacky things - feel free to overwrite once it migrates :) | 14:10 |
| ginggs | i diff'd cmath from gcc 5 and 7 and found weird if __GLIBCXX_HAVE_OBSOLETE_ISNAN and __GLIBCXX_HAVE_OBSOLETE_ISINF things | 14:11 |
| LocutusOfBorg | I think using gcc-5 makes the C99_MATH somewhat needed | 14:20 |
| LocutusOfBorg | I would override it right now, because it won't migrate | 14:20 |
| -queuebot:#ubuntu-release- New binary: fonttools [amd64] (bionic-proposed/universe) [3.16.0-2] (no packageset) | 14:22 | |
| LocutusOfBorg | libkf5kface demoted, so maybe better wait | 14:26 |
| infinity | xnox: Yay, the non-double-linked build worked. | 14:33 |
| infinity | Not sure I have the brainpower to hunt the SIGBUS in mrpt/armhf this morning, though. | 14:34 |
| LocutusOfBorg | infinity, should I upload caffe-contrib *now* or wait^ | 14:36 |
| LocutusOfBorg | it is still not published the one right now in -proposed | 14:36 |
| infinity | LocutusOfBorg: Did you look at ginggs' diff yet to determine if yours is objectively "better"? | 14:36 |
| LocutusOfBorg | yes I did | 14:36 |
| infinity | Then go for it. | 14:36 |
| LocutusOfBorg | +- y[i] = static_cast<bool>((std::signbit)(x[i]))) | 14:36 |
| LocutusOfBorg | ++ y[i] = static_cast<bool>((::__builtin_signbit)(x[i]))) | 14:36 |
| LocutusOfBorg | this is what worries me more of his change, while mine seems "better" (even if I don't understand c++, or at least I like to avoid it) | 14:37 |
| tyhicks | doko: thanks for the explanation - I missed the different copyright info in pam/constants.go | 14:37 |
| LocutusOfBorg | ginggs, ^^ uploaded | 14:38 |
| ginggs | LocutusOfBorg: thanks! | 14:38 |
| LocutusOfBorg | yw! | 14:39 |
| -queuebot:#ubuntu-release- Unapproved: ubuntu-drivers-common (zesty-proposed/main) [1:0.4.22.1 => 1:0.4.22.2] (desktop-core, ubuntu-server) | 14:43 | |
| LocutusOfBorg | can't we move freerdp to universe now? I don't know how to check | 14:44 |
| jbicha | LocutusOfBorg: remmina needs to migrate to bionic first, which appears to require remmina-plugin-spice being demoted to universe (it is a new binary package) | 14:45 |
| jbicha | there's a few ways to check, you can look at http://people.canonical.com/~ubuntu-archive/component-mismatches.html | 14:46 |
| LocutusOfBorg | $ ./demote-to-proposed debhelper --dry-run -m "foo" | 14:46 |
| LocutusOfBorg | I'm trying to play with such commands | 14:46 |
| jbicha | also comes in a .svg version that's a bit different and there are component-mismatches-proposed versions too | 14:46 |
| LocutusOfBorg | I know that page, but I prefer the bash tool, even if I'm not able to use it yet | 14:47 |
| -queuebot:#ubuntu-release- Unapproved: ubuntu-drivers-common (xenial-proposed/main) [1:0.4.17.3 => 1:0.4.17.4] (desktop-core, ubuntu-server) | 14:50 | |
| LocutusOfBorg | slow publisher? | 14:50 |
| infinity | LocutusOfBorg: Uhh, demote-to-proposed has nothing to do with components. | 14:50 |
| LocutusOfBorg | lol you right! ahahahah bad monday here :p | 14:51 |
| infinity | LocutusOfBorg: As for, "how do we know when to demote", the report *is* how we know. There's no commandline tool. | 14:51 |
| LocutusOfBorg | oh... ok! I wanted to look for a single package output | 14:52 |
| infinity | LocutusOfBorg: Grep the report? | 14:53 |
| infinity | LocutusOfBorg: There's no way to examine a single package and determine which component it belongs in, as it's a function of the archive as a whole combined with the seeds. Hence the report. | 14:53 |
| LocutusOfBorg | btw, stuff is "accepted" since 45 minutes... do we have some public log? | 14:54 |
| infinity | Nope. | 14:54 |
| LocutusOfBorg | does publisher need britney to finish? | 14:54 |
| infinity | Which "stuff" are you after? | 14:54 |
| apw | and 45m is not utterly unreasonable, that is a publisher-run sized delay | 14:54 |
| LocutusOfBorg | caffe-contrib, I uploaded the ubuntu2 (showing as "accepted") and the ubuntu1 is "accepted" too | 14:55 |
| apw | if there is a lot to copy it can take much longer than it often does | 14:55 |
| LocutusOfBorg | ack, lets wait some more time | 14:55 |
| * LocutusOfBorg brings coffee to britney | 14:55 | |
| * LocutusOfBorg please forward to publisher if you need it :) | 14:55 | |
| infinity | LocutusOfBorg: You don't really need to babysit it. :P | 14:55 |
| apw | ubuntu1 is showing as published in the history, and ubuntu2 is pending waiting on the next publisher | 14:56 |
| infinity | It's built, it'll publish shortly. | 14:56 |
| LocutusOfBorg | apw, I mean binaries, not source "Bionic: [FULLYBUILT] amd64 (Accepted) [FULLYBUILT] ppc64el (Accepted)" | 14:56 |
| LocutusOfBorg | anyhow, lets wait, and hope everything goes in release | 14:56 |
| apw | must have missed the publisher indeed ... such things are sent to try us | 14:56 |
| infinity | It won't right away. | 14:56 |
| infinity | Still a few more bits. | 14:56 |
| LocutusOfBorg | ack, will you sort them out? | 14:57 |
| infinity | I think I'll demote and block mrpt, pending more time to find the sigbus. | 14:57 |
| infinity | And blender needs to finish building in my PPA so I can copy it over. | 14:57 |
| infinity | Which will be a couple of hours. | 14:57 |
| LocutusOfBorg | ok thanks | 14:57 |
| LocutusOfBorg | I have a perl merge ready, if I can upload after the big run | 14:58 |
| infinity | Sonofa. And a sigbus in blender on armhf too. | 14:58 |
| infinity | WHY. | 14:58 |
| -queuebot:#ubuntu-release- New source: fscrypt (bionic-proposed/primary) [0.2.2-0ubuntu2] | 15:03 | |
| tyhicks | doko: ^ I've fixed fscrypt's debian/copyright in ubuntu2 - thanks again! | 15:05 |
| cyphermox | can someone please murder the nplan test on xenial that has been running for > 96 hours? | 15:27 |
| -queuebot:#ubuntu-release- Unapproved: edk2 (artful-proposed/universe) [0~20170911.5dfba97c-1 => 0~20170911.5dfba97c-1ubuntu0.1] (no packageset) | 15:28 | |
| smoser | hi. is 'verification-failed' sufficient to stop release of sru ? | 15:49 |
| smoser | https://bugs.launchpad.net/ubuntu/+source/cloud-init/+bug/1721847 specifically want to stop that. | 15:49 |
| ubot5 | Ubuntu bug 1721847 in cloud-init (Ubuntu Zesty) "sru cloud-init 2017-10-06 (17.1-18-gd4f70470-0ubuntu1)" [Medium,Fix committed] | 15:49 |
| smoser | blackboxsw: ^ rharper ^ | 15:49 |
| slangasek | doko: urdfdom, do you think I should just do a no-change rebuild then? actually kido didn't ftbfs on ppc64el in the gcc7 rebuilds so it's something besides just straight gcc7 abi change | 16:11 |
| apw | smoser, yes verification-failed will prevent it being a candidate for release | 16:18 |
| apw | smoser, will you be replacing it shortly ? | 16:18 |
| -queuebot:#ubuntu-release- Unapproved: google-cloud-sdk (artful-release/partner) [176.0.0-0ubuntu1 => 176.0.0-0ubuntu1] (no packageset) (sync) | 16:19 | |
| jbicha | apw: please reject mutter and gtk+3.0 from artful unapproved, we'll be doing point releases this week instead of cherry-picking patches | 16:21 |
| -queuebot:#ubuntu-release- Unapproved: google-cloud-sdk (zesty-release/partner) [176.0.0-0ubuntu1~17.04.0 => 176.0.0-0ubuntu1~17.04.0] (no packageset) (sync) | 16:22 | |
| -queuebot:#ubuntu-release- Unapproved: ubuntu-drivers-common (artful-proposed/main) [1:0.4.23 => 1:0.4.23.1] (desktop-core, ubuntu-server) | 16:23 | |
| -queuebot:#ubuntu-release- Unapproved: google-cloud-sdk (xenial-release/partner) [176.0.0-0ubuntu1~16.04.0 => 176.0.0-0ubuntu1~16.04.0] (no packageset) (sync) | 16:23 | |
| -queuebot:#ubuntu-release- Unapproved: google-cloud-sdk (trusty-release/partner) [176.0.0-0ubuntu1~14.04.0 => 176.0.0-0ubuntu1~14.04.0] (no packageset) (sync) | 16:24 | |
| sil2100 | jbicha: ACK | 16:30 |
| sil2100 | jbicha: done | 16:32 |
| -queuebot:#ubuntu-release- Unapproved: rejected gtk+3.0 [source] (artful-proposed) [3.22.24-0ubuntu2.1] | 16:32 | |
| -queuebot:#ubuntu-release- Unapproved: rejected mutter [source] (artful-proposed) [3.26.1-2ubuntu2.1] | 16:32 | |
| smoser | apw: we're working on it. | 16:38 |
| apw | sil2100, ahh thats why i can't find them : | 16:40 |
| apw | :_ | 16:40 |
| apw | ;_ | 16:40 |
| apw | :) | 16:40 |
| slashd | sil2100, thanks | 16:49 |
| sil2100 | slashd: yw! Sorry it took so long, got pulled into something else so my SRU shift has shifter | 16:54 |
| sil2100 | *shifted | 16:54 |
| slashd | sil2100, no problem at all | 16:56 |
| slashd | sil2100, just by curiosity, did you see andreas comment about ubuntu-advantage-tools ? | 16:56 |
| slashd | on #ubuntu-devel | 16:57 |
| sil2100 | Yeah, didn't look at the bug yet | 16:57 |
| slashd | sil2100, ack | 16:57 |
| -queuebot:#ubuntu-release- Unapproved: nvidia-graphics-drivers-384 (artful-proposed/restricted) [384.90-0ubuntu3 => 384.90-0ubuntu3.17.10.1] (no packageset) | 17:02 | |
| -queuebot:#ubuntu-release- Unapproved: nvidia-graphics-drivers-384 (xenial-proposed/universe) [384.90-0ubuntu0.16.04.1 => 384.90-0ubuntu0.16.04.2] (no packageset) | 17:03 | |
| -queuebot:#ubuntu-release- New: accepted mrpt [amd64] (bionic-proposed) [1:1.5.3-1] | 17:04 | |
| -queuebot:#ubuntu-release- New: accepted mrpt [i386] (bionic-proposed) [1:1.5.3-1] | 17:04 | |
| -queuebot:#ubuntu-release- New: accepted mrpt [s390x] (bionic-proposed) [1:1.5.3-1] | 17:04 | |
| -queuebot:#ubuntu-release- New: accepted mrpt [arm64] (bionic-proposed) [1:1.5.3-1] | 17:04 | |
| -queuebot:#ubuntu-release- New: accepted mrpt [ppc64el] (bionic-proposed) [1:1.5.3-1] | 17:04 | |
| -queuebot:#ubuntu-release- New: accepted fonttools [amd64] (bionic-proposed) [3.16.0-2] | 17:04 | |
| infinity | xnox: So, want more boost fun? Linked against 1.62, blender on armhf works. Double-linked, it works. Linked only against 1.65, it SIGBUS. | 17:16 |
| infinity | xnox: Methinks 1.65 may have some latent alignment issues. | 17:16 |
| infinity | (This could also explain the mrpt SIGBUS, but maybe they're unrelated) | 17:16 |
| infinity | slangasek, doko: ^-- I think, if mrpt and blender are either fixed or removed, we might be about ready to transition the ICU mess. Maybe. I need sleep, so handing off. | 17:17 |
| slangasek | infinity: update_output looks like the uninstallable count has regressed; I guess I'll dig to understand? | 17:20 |
| slangasek | e.g. a bunch of evolution stuff that wasn't there last night | 17:20 |
| infinity | slangasek: That would be because boost isn't valid right now cause someone repromoted your -doc hacks to main. I've re-demoted again. | 17:21 |
| infinity | slangasek: A publisher cycle should make it look less bad. | 17:21 |
| xnox | boost-doc should not be in main | 17:21 |
| infinity | slangasek: I've re-demoted and updated the seeds, so component-mismatches stops lying. | 17:21 |
| xnox | (fyi) | 17:21 |
| infinity | xnox: It always has been. Anyhow, I've bounced it to universe the correct way now. | 17:22 |
| xnox | infinity, long story short, it ships no docs for a while now =) | 17:22 |
| slangasek | infinity: k. | 17:22 |
| xnox | trying to fix that for realz | 17:22 |
| slangasek | infinity: right; that was a temporary demotion on my part before to make it migratable until we had a decision on whether it was correct to demote, I wasn't going to modify the seeds until decided. Didn't count on people being mismatch-happy | 17:24 |
| slangasek | xnox: it has no docs, but it wants to pull more javascript stuff into main - check. | 17:24 |
| infinity | Err, yeah, if it ships no docs, dropping that dep would seem like a no-brainer. :P | 17:25 |
| infinity | Erm, and another publisher run to pick up the libgnuradio-radar3.7.10 NBS removal I just had to do from -proposed. | 17:32 |
| acheronuk | infinity: I reported https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=880166 | 17:36 |
| ubot5 | Debian bug 880166 in dpkg "regression in dpkg-gensymbols output on symbols changes" [Normal,Open] | 17:36 |
| acheronuk | not sure is that is impact enough to cherry pick before debian fix it | 17:36 |
| acheronuk | though it did confuse the log parsing of the Kubuntu CI, hence my report | 17:36 |
| infinity | acheronuk: Do you have an example log? | 17:37 |
| infinity | acheronuk: Oh, I see, it's just the header of the output that's wrong. It still presents the diff and errors out. | 17:38 |
| acheronuk | infinity: https://launchpadlibrarian.net/343625396/buildlog_ubuntu-bionic-amd64.libjpeg-turbo_1.5.2-0ubuntu6~18.04~ppa1_BUILDING.txt.gz | 17:39 |
| infinity | acheronuk: So, no, the header being incorrect doesn't bug me enough to warrant an emergency upload. | 17:39 |
| acheronuk | where I deliberately removed a symbol to test | 17:39 |
| acheronuk | infinity: that is fine | 17:40 |
| acheronuk | only impact I found so far was our CI, and have done a temp fix there | 17:42 |
| acheronuk | so all is calm :) | 17:42 |
| === alan_g is now known as alan_g|EOD | ||
| LocutusOfBorg | jbicha, so now some better possibility to move freerdp to universe? | 19:10 |
| jbicha | sure | 19:12 |
| slangasek | it's already on the component-mismatches list and will find itself demoted once icu migrates | 19:13 |
| nacc | slangasek: iiuc, php7.1 in b-p is also blocked by icu (based upon update_output.txt) -- reading it, though, i'm a bit confused why php7.1-intl failed to install on armhf, it is from src:php7.1. Will this just 'resolve itself' once icu migrates? | 20:02 |
| slangasek | nacc: that only shows that php7.1 cannot migrate solo because doing so makes php7.1-intl uninstallable (precisely because of icu). | 20:06 |
| slangasek | nacc: the interesting output is what comes after the autohinter | 20:07 |
| nacc | slangasek: ah right, thank you | 20:08 |
| tyhicks | xnox: have you heard anything about the libseccomp xenial SRU that you have in the unapproved queue? I suspect that the artful release and bionic opening starved it of attention but I wanted to make sure there wasn't something else | 20:09 |
| xnox | tyhicks, i have not had any feedback about it, no. | 20:09 |
| tyhicks | alright | 20:10 |
| tyhicks | thanks! | 20:10 |
| -queuebot:#ubuntu-release- Unapproved: gtk+3.0 (artful-proposed/main) [3.22.24-0ubuntu2 => 3.22.25-0ubuntu0.1] (ubuntu-desktop) | 20:16 | |
| nacc | slangasek: if i'm reading the second autohint correctly, (with php7.1 in the "Trying ..."), both mpd and poedit have been rebuilt in b-p with the new icu, but come up as being uninstallable? | 20:18 |
| slangasek | nacc: the first autohint is the more complete one. if the big transition doesn't go, the autohinter sometimes checks to see if there's a smaller subset that might be more manageable to push through | 20:25 |
| nacc | slangasek: ah i see, that makes sense | 20:26 |
| -queuebot:#ubuntu-release- Unapproved: lshw (xenial-proposed/main) [02.17-1.1ubuntu3.3 => 02.17-1.1ubuntu3.4] (core) | 20:53 | |
| -queuebot:#ubuntu-release- Unapproved: livecd-rootfs (trusty-proposed/main) [2.208.14 => 2.208.15] (desktop-core) | 21:37 | |
| * tsimonq2 scratches head | 21:53 | |
| tsimonq2 | https://launchpad.net/ubuntu/+source/ktexteditor | 21:53 |
| tsimonq2 | So when that was released to artful-updates it wasn't removed from artful-proposed O_o | 21:54 |
| slangasek | correct; that's modus operandi | 21:56 |
| slangasek | post-promotion removals happen as a separate step | 21:57 |
| tsimonq2 | But usually those removals are done pretty quick, no? That was 4 days ago. ;) | 21:57 |
| slangasek | not that quick. | 21:57 |
| tsimonq2 | Huh, ok, always thought it was within hours... anyways, carry on, thanks slangasek :) | 21:58 |
| slangasek | definitely not, it's a manual operation | 21:58 |
| xnox | infinity, is autosync on? | 22:19 |
| xnox | infinity, or is postgresql-10 blacklisted? | 22:19 |
| tsimonq2 | xnox: (afair autosync was waiting for icu + boost + etc. to migrate) | 22:20 |
| xnox | icu is good | 22:23 |
| xnox | boost is not | 22:23 |
| tsimonq2 | xnox: Are the icu and boost transitions tangled? | 22:25 |
| xnox | yes, but icu is done rebuilding things, but entagled to migrate things | 22:31 |
| tsimonq2 | Ok | 22:32 |
| jbicha | xnox: see https://irclogs.ubuntu.com/2017/10/29/%23ubuntu-release.html | 22:33 |
| slangasek | xnox: it's not on because until icu/boost migrates, turning on autosync invites people to do rebuilds against other migrations in -proposed and grow the entanglement | 23:19 |
| slangasek | xnox: infinity's last analysis was that only blender and mprt were still blocking but I haven't really had a chance to look | 23:20 |
| xnox | at first i thought icu.... but then i had gag reflex when i saw -std=c++03 | 23:58 |
| xnox | https://launchpadlibrarian.net/343166764/buildlog_ubuntu-bionic-amd64.cegui-mk2_0.8.7-1.3build3_BUILDING.txt.gz | 23:58 |
| xnox | welcome to the 10s? | 23:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!