/srv/irclogs.ubuntu.com/2020/02/26/#ubuntu-devel.txt

cpaelzerrafaeldtinoco: FYI for kronosnet (which somewhat belongs to corosync) and is failing on i386 => https://code.launchpad.net/~paelzer/britney/hints-ubuntu-focal-kronosnet-i386/+merge/37987709:52
cpaelzerthe commit message hopefully explains all the details09:52
cpaelzerso far I've seen you and vorlon retrying this test since the i386 switch09:53
cpaelzerat http://autopkgtest.ubuntu.com/packages/k/kronosnet/focal/i386 and since then it was part of the whitelist (but versioned to the former version)09:54
cpaelzerthanks LocutusOfBorg for the massive set of test triggers around llvm-8 - that will unblock a bunch of stuff10:08
LocutusOfBorgcpaelzer, I hope it will make something migrate :D10:31
LocutusOfBorge.g. having ocaml candidate was my primary goal10:31
rbalinthi, i'd like to perform a mini transition to kodi 18 later today11:17
rbalintall of the kodi-reverse deps need sourceful uploads and the main ones are built in ppa:ci-train-ppa-service/3204 , and i'm in the process of updating all addons on Salsa11:18
rbalintthe reason of going ahead of Debian is libkodiplatform so bump being in NEW11:20
rafaeldtinococpaelzer: thanks! reviewing your merge12:12
ahasenackdoko: hi, do you know about this glibc issue?12:17
ahasenackI just got this ping:12:17
ahasenack"""12:17
ahasenackcan we get this fixed in Ubuntu releases? https://sourceware.org/bugzilla/show_bug.cgi?id=23844#c1412:17
ahasenackIt causes BIND 9.16 to deadlock under load.12:17
ahasenack"""12:17
ubottusourceware.org bug 23844 in nptl "pthread_rwlock_trywrlock results in hang" [Normal,Resolved: fixed]12:17
ahasenackcomment #23 in that bug also says it's being seen in 18.04 in openvswitch12:18
dokoahasenack: no. infinity ^^^12:26
tjaaltonahasenack: btw, bind-dyndb-ldap upstream said it'll take a couple of weeks to get it ported to bind 9.1613:43
ahasenackcool, then we might be able to keep it13:43
tjaaltonand get freeipa back13:44
ahasenackthat needs pkcs11, no?13:46
tjaaltonwhat does?14:15
tjaaltonyou mean opendnssec?14:16
tjaaltonor if you mean the native-pkcs11 patch in the current bind, aiui that can go14:19
ahasenacktjaalton: I meant that native pkcs11 hack, yeah14:31
tjaaltonI guess it works with openssl just the same14:32
ahasenacksounds like a better way forward even if it doesn't work on the first try14:34
LocutusOfBorgricotz, you sure wrt meson upload? Explicit depends on rustc and valac, so autopkgtests can pick it up this should be useless14:52
rafaeldtinocokanashiro: https://bugs.launchpad.net/ubuntu/+source/ruby-defaults/1:2.5.2ubuntu1/+build/1859630414:53
ubottuLaunchpad bug 1 in Ubuntu Malaysia LoCo Team "Microsoft has a majority market share" [Critical,In progress]14:53
LocutusOfBorgconsidering that the new code has this: " @skipIfNoExecutable('rustc')"14:53
kanashirorafaeldtinoco, any issue with the sync?14:54
rafaeldtinocokanashiro: nope, it was me14:54
rafaeldtinocoi had an issue with my container14:54
rafaeldtinocolet me sync it14:54
rafaeldtinocokanashiro: syncpackage: Source ruby-defaults -> focal/Proposed: current version 1:2.5.2ubuntu1, new version 1:2.5.7.114:56
kanashirorafaeldtinoco, we want the version 1:2.7~0 from experimental which drops Ruby 2.5 support14:57
rafaeldtinocoyep, I think it has to go to unstable14:57
rafaeldtinocofor the sync14:57
rafaeldtinocoahasenack: ^14:57
ahasenackhm?14:57
rafaeldtinococan we "sync" from experimental ?14:57
rafaeldtinocoput a pkg on "sync" with experimental ?14:57
ahasenackin terms of tooling? yes14:57
kanashirorafaeldtinoco, can't you use "-d experimental"?14:58
rafaeldtinocokanashiro: yep, i wasnt sure if it would be ok14:58
kanashiroah ok14:58
rafaeldtinocoalright. done.14:59
kanashirobecause we don't have enough time to wait for Debian14:59
rafaeldtinocoyep14:59
rafaeldtinocolet me check my email14:59
rafaeldtinocoif its good ill close the bug14:59
seb128LocutusOfBorg, the changes are not new from this upload, also the depends are there so an upload to those packages trigger the autopkgtests which they wouldn't do otherwise, not to handle the missing compiler15:00
rafaeldtinocokanashiro: alright. done15:00
rafaeldtinoco=) have fun15:00
kanashirorafaeldtinoco, I am not quite sure how to proceed but I have a list of packages that needs to be rebuilt against ruby2.7, should I ask on #ubuntu-release for binNMUs or a core dev like you should help me on that?15:03
tkamppeterLocutusOfBorg, it seems that sane-backends is hanging on the gscan2pdf autopkgtest for arm64, and gscan2pdf fails generally on arm64. gscan2pdf is a Universe package where no one here at Ubuntu cares, simply auto-synced. Can we do an exception in the autopkg tests here?15:05
rafaeldtinocoI'm able to help. Let's get it published first and then we can start doing build only uploads15:05
ahasenacktkamppeter: I have a bug for gscan2pdf getting stuck on arm6415:05
ahasenacktkamppeter: upstream tried to help, but this wasn't reproduceable outside of canonistack instances15:05
kanashirorafaeldtinoco, great, thanks for helping me on this :)15:05
rafaeldtinocoanytime!15:06
* rafaeldtinoco goes for lunch now before meeting ;)15:06
ahasenacktkamppeter: https://bugs.launchpad.net/ubuntu/+source/gscan2pdf/+bug/1860592 you might want to badtest it using this bug as the reason15:06
ubottuLaunchpad bug 1860592 in gscan2pdf (Ubuntu) "Test t/06190 hangs on arm64" [Undecided,New]15:06
tkamppeterahasenack, thanks, how do I "badtest" this?15:17
LocutusOfBorgseb128, context?15:19
seb128LocutusOfBorg, <LocutusOfBorg> ricotz, you sure wrt meson upload? Explicit depends on rustc and valac, so autopkgtests can pick it up this should be useless15:19
LocutusOfBorgtkamppeter, you have to ask release chanel...15:19
LocutusOfBorgseb128, yes, but better convince debian to also add them...15:19
LocutusOfBorgin any case, removing is not making the situation worse, because now the code handles that case15:20
seb128LocutusOfBorg, that's a different statement than 'this should be useless'15:20
LocutusOfBorgyes, true :D15:20
seb128LocutusOfBorg, it does make a different as I just wrote15:20
seb128LocutusOfBorg, without that the meson tests wouldn't trigger on a vala uplaod15:21
LocutusOfBorgI mean, this is now an useless delta, because it might trigger test failures that debian isn't experiencing, and we should convince them that they are real issues15:21
seb128but yes, ideally Debian would take those changes15:21
LocutusOfBorgopening RC bugs in Debian is trivial now that they have a ci15:21
seb128LocutusOfBorg, it's not useless delta, it prevents us from regressions15:21
LocutusOfBorgbut we have to convince them to extend testsuites15:21
seb128it makes meson be tested on vala updates15:21
seb128which wouldn't otherwise15:21
LocutusOfBorgwill anybody open a bug against meson in debian with a patch? :)15:22
seb128but agreed, better to thave that in Debian15:22
seb128LocutusOfBorg, I said to Rico that we should forward the change when he added it yes15:22
seb128LocutusOfBorg, but feel free to it if you have free cycles since he didn't seem to have done it (yet)15:23
rbasakHow long has vim defaulted to restoring the cursor to its previous position on a re-edit of a file?15:40
rbasakIt keeps throwing me editing debian/changelog :-/15:40
* rbasak isn't sure if he likes this or not15:40
rbasakAlso on git commit messages15:40
tkamppeterahasenack, LocutusOfBorg, did you already badtest something? On #ubuntu-release they say one needs to do a merge request on hints-ubuntu, and there seems to be no documentation about how this works.15:42
rbasakSo what's the convention in our team for package merge MPs currently wrt. test builds, PPAs, and dep8?15:43
rbasakI never used to do anything but a local test build on amd64 only, but I think you've upped the bar nowadays?15:43
rbasak(in our team -> server team)15:43
tkamppeterahasenak, seems that you have added "force-badtest gscan2pdf/2.6.3-1/arm64" could you update it or better expand it to all/arbitrary versions?15:44
tkamppeterahasenack, it even contains a reference to your bug report. I found it via "blame" in the ubuntu-release file.15:46
rbasakahasenack: oh look. You've done the PASSWORD -> SHA1 thing before! https://merges.ubuntu.com/p/pam-mysql/pam-mysql_0.8.0-1ubuntu3.patch15:47
seb128vorlon, libvisual autopkgtest fail on i386 because it depends on g++ ... IIRC you said to use build-essential instead? is it tehnically wrong to use g++ (I'm just trying to figure out what to write if I forward that request to Debian)15:52
ahasenacktkamppeter: it's the release tem who accepts or not merges against that branch15:59
ahasenacktkamppeter: what you should do next I think is propose a bump in the version of that hint15:59
ahasenacktkamppeter: this was the merge proposal I had for that, and that was merged: https://code.launchpad.net/~ahasenack/britney/gscan2pdf-badtest/+merge/37795516:00
ahasenackyou can use it as a template16:00
ahasenackjust add the new ubuntu version next in that line16:00
ahasenackrbasak: heh, good times16:03
LocutusOfBorgtkamppeter, thanks for doing it!16:04
LocutusOfBorgseb128, I'll do it16:05
seb128LocutusOfBorg, thanks!16:05
tkamppeterahasenack, how do I clone the original bzr branch to my LP account so that I get a "Propose for merge" entry?16:06
ahasenacktkamppeter:      bzr branch lp:~ubuntu-release/britney/hints-ubuntu16:06
tkamppeterahasenack, this way I could download and edit it. But how do I "bzr push" it into my LP account?16:07
ahasenackyou have to branch it, then hack on it, and then bzr push lp:~<yourlp>/britney/<name-of-your-branch>16:08
tkamppeterWhy "britney"?16:08
ahasenackit's the name of the project16:08
tkamppeterahasenack, now it works, thanks.16:10
LocutusOfBorgseb128, I sent the email, but you seem to have a failure on testsuite16:10
ahasenacktkamppeter: cool16:10
LocutusOfBorgexhaustive           FAIL non-zero exit status 116:10
LocutusOfBorgcrossbuild           FAIL non-zero exit status 116:10
LocutusOfBorgthat exhaustive is failing ubuntu-only and not debian...16:10
LocutusOfBorgprobably that valac /rustc addition?16:10
seb128LocutusOfBorg, that wouldn't make any sense, those are pulled in in Debian as well by @build-depends@, it's not that the autopkgtest service doesn't know to use @build-depends@ to trigger other tests16:12
seb128LocutusOfBorg, also where was 0.53.2 tested in Debian? it's not on https://ci.debian.net/packages/m/meson/unstable/amd64/ yet?16:12
seb128LocutusOfBorg, it's the same failure than https://ci.debian.net/data/autopkgtest/unstable/amd64/m/meson/4340268/log.gz though16:13
seb128so maybe .2 didn't resolve the problem as Jussi though it would16:14
tkamppeterahasenack, LocutusOfBorg: Here is my merge request: https://code.launchpad.net/~till-kamppeter/britney/hints-ubuntu/+merge/379910 I hope it is correct.16:16
ahasenacktkamppeter: make it against britney/hints-ubuntu16:17
ahasenacknot just britney16:17
ahasenackit's confusing, I also make that mistake a few times16:17
ahasenacktkamppeter: I mean the proposal, your branch is fine16:17
ahasenackclick "resubmit" on the top right, and append "/hints-ubuntu" to the target16:18
ahasenackin the end, "merge into" in the mp page, should look like "lp:~ubuntu-release/britney/hints-ubuntu"16:18
ahasenackand then your diff will show up correctly16:18
kanashirorafaeldtinoco, re ruby 2.7 transition: we need to start with the packages in level 1 here: https://people.canonical.com/~ubuntu-archive/transitions/html/html/ruby2.7-add.html16:20
rafaeldtinocowh0t ?116:21
rafaeldtinocolol16:21
tkamppeterHave done it. Now it reads: lp:~ubuntu-release/britney/hints-ubuntu16:21
tkamppeterahasenack, why does this not work in the first place. Or why do we not switch to GitHub or GitLab?16:22
ahasenacktkamppeter: I'm not the one to answer these questions, sorry :)16:22
ahasenackI'm just a client, just like you :)16:22
ahasenacktkamppeter: is it still https://code.launchpad.net/~till-kamppeter/britney/hints-ubuntu/+merge/379910 ?16:25
ahasenackah, wait16:25
LocutusOfBorgseb128, https://ci.debian.net/data/autopkgtest/testing/amd64/m/meson/4381867/log.gz16:25
ahasenackI see the superseeded bit16:25
ahasenacktkamppeter: yeah, diff looks correct now16:25
LocutusOfBorghttps://code.launchpad.net/~till-kamppeter/britney/hints-ubuntu/+merge/37991116:26
LocutusOfBorgthis one is nice16:26
seb128LocutusOfBorg, that one is red though?16:28
LocutusOfBorgyes, but exhaustive           PASS16:29
LocutusOfBorgyou can see the run here https://tracker.debian.org/pkg/meson16:29
seb128LocutusOfBorg, and not, the other fail is the other patch tumbleweed added in https://launchpad.net/ubuntu/+source/meson/0.53.0-1ubuntu216:29
LocutusOfBorgor here https://ci.debian.net/packages/m/meson/unstable/amd64/16:29
seb128LocutusOfBorg, see https://github.com/mesonbuild/meson/pull/665616:29
seb128it trigger the same issue in upstream CI16:29
seb128tumbleweed, ^16:29
LocutusOfBorgseb128, so shouldn't we just revert the patch? :)16:30
seb128LocutusOfBorg, you could but then it would fail on python missing16:31
seb128which Debian doesn't have as an issue atm16:31
tumbleweedthat patch did the right thing on Ubuntu, but wasn't appropriate for Debian at the time16:31
LocutusOfBorgsure16:31
tumbleweedfrom my PoV, a quick hack, rather than trying to do something generic16:31
LocutusOfBorgwhy not move to python3?16:31
seb128tumbleweed, the tests and upstream CI fail with 'meson.build:20:2: ERROR: Assert failed: Didn't load python from config file'16:31
seb128tumbleweed, so seems it's buggy16:32
seb128in ubuntu16:32
LocutusOfBorgmaybe xenial or whatever upstream has didn't have python2?16:32
LocutusOfBorgI mean, the symlink16:32
tumbleweedseb128: It got it to build on debian, where ubuntu1 didn't16:32
tumbleweeds/debian/ubuntu/16:33
seb128tumbleweed, it built but autopkgtests are failing due to it16:33
seb128so it didn't help much16:33
tumbleweedit must have passed, because it migrated16:34
seb128tumbleweed, anyway, I'm not interested to argue, was just pointing out that meson is still blocked because that patch created a regression16:34
LocutusOfBorgmmm16:34
LocutusOfBorgpython2 not installed?16:34
LocutusOfBorgit is not listed in debian/tests/control lol16:34
seb128LocutusOfBorg, /usr/bin/python got removed from focal16:34
tumbleweedright, so I should have updated debian/tests/control too16:34
LocutusOfBorgseb128, I'm not saying that16:34
LocutusOfBorgseb128, python2 is not getting installed16:35
seb128LocutusOfBorg, if python2 is needed it's easy to add to the control16:35
LocutusOfBorgin debian it is16:35
tumbleweedDebian hasn't removed /usr/bin/python, yet16:35
seb128still we need to fix the other issue (which is also in Debian)16:35
LocutusOfBorgmmm it is not listed but it gets installed due to something else16:36
LocutusOfBorg        self._simple_test('python2', 'python2')16:40
LocutusOfBorgthat would fix it16:40
LocutusOfBorgseb128, can you please update the pull request?16:40
LocutusOfBorgso we get easy testing16:40
seb128LocutusOfBorg, sure, will do that in a bit16:40
LocutusOfBorg<316:40
vorlonseb128: build-essential gets handled specially by the patches to autopkgtest; so it's not "technically wrong", just incompatible with the current implementation for cross16:43
seb128vorlon, what's the right thing to state in a Debian bug report? 'please use build-essential instead of g++, should work the same for you and it fixes cross build in Ubuntu'?17:04
vorlonseb128: that's what I would say yes17:04
seb128vorlon, thanks17:04
vorlonseb128: if you want to steal some of my boilerplate language, you could use e.g. Debian bug #95030317:05
ubottuDebian bug 950303 in libgpg-error "libgpg-error: Please make autopkgtests cross-test-friendly" [Minor,Fixed] http://bugs.debian.org/95030317:05
seb128vorlon, thanks17:06
seb128LocutusOfBorg, https://github.com/mesonbuild/meson/pull/6656 still not happy, so I don't find the issue in the log17:35
seb128LocutusOfBorg, ah, I see it now17:36
seb128'meson.build:1:0: ERROR: Value "python2" for combo option is not one of the choices. Possible choices are: "find_program", "config_dep", "python3", "python".'17:36
ricotzseb128, for my defence, I wrote jpakkane about it in #mesonbuild18:01
LocutusOfBorgseb128, its self._simple_test('python', 'python2')18:03
LocutusOfBorgprobably18:03
LocutusOfBorgseb128, it is good, you probably need this vi "./test cases/unit/47 native file binary/meson_options.txt"18:08
LocutusOfBorgand then change python to python218:08
LocutusOfBorgcan you please try?18:08
tumbleweedOn Python 2.7 doko: FYI: https://bugs.launchpad.net/ubuntu/+source/python-pip/+bug/1864766 (there's a link at the bottom to a pypa discussion about this)19:10
ubottuLaunchpad bug 1864766 in python-pip (Ubuntu Xenial) "[SRU] pip in xenial is installing packages incompatible with Python 2.7 (and those are becoming common)" [Medium,In progress]19:10
gbit86Does anyone know if it is possible to configure xkb to cancel the Shift key if BOTH shift keys are held down?20:06
=== JanC_ is now known as JanC
ahasenackinfinity: hi, do you remember why this cross-compile patch wasn't forwarded to debian at the time? It has an explicit "Forwarded: not needed". Was it because it's from upstream, and it was thought it would be in a new upstream release by now?21:32
ahasenackinfinity: https://pastebin.ubuntu.com/p/vctTmQFSKn/21:32
infinityahasenack: Exactly that, yes.  It was yeeeears ago, I kinda figured upstream would have released something including that by now.21:37
LocutusOfBorgseb128, https://github.com/mesonbuild/meson/pull/670322:14
LocutusOfBorgthis might be the solution22:14
vorlonseb128: libsoup2.4 fixed btw, new php had needed adding to the whitelist which is done now22:24
seb128LocutusOfBorg, good, one less thing for me to refresh22:24
seb128vorlon, ah ok, that was still on my todo but I'm struggling a bit due to ff coming ... thanks for the status update and for fixing!22:25
seb128vorlon, I need also to read again L_aney's hints from yesterday about how the whitelist and refreshes are being handled22:25
seb128vorlon, while you are around, what's the word with icu? I'm still holding to upload some of the GNOME stuff but ff is tomorrow and I would prefer to no have to end up having to do ffe parperwork ... :)22:26
vorlonseb128: I intend to finish it today before I EOD; how much I have to break in anger to accomplish that is another question22:27
vorlon(I just had to roll back an upload of a ruby package which depends on icu, which would've entangled icu + ruby-defaults)22:27
vorlonseb128: is the stuff you're holding back currently in -proposed?22:27
* sarnold hands vorlon a larger hammer22:27
seb128we don't have a a 'force in hammer' for those cases?22:28
vorlonwe do have one22:28
vorlonbut it's also a moving target because the force hammer only works if the versions of the packages in -proposed /stay the same/ between the time i write the hint and the time britney runs22:28
seb128vorlon, yes, like I want to upload webkit2gtk 2.27, but https://launchpad.net/ubuntu/+source/webkit2gtk/2.26.4-1ubuntu1 is on the 'rebuilt with the new icu' list22:28
vorlonack22:28
vorlonyeah please hold off just a little longer22:29
seb128also poppler and gnome-desktop transitions are started22:29
vorlonI think 1-2 britney runs and we should be there22:29
seb128but they can't be completed without rebuilding libreoffice and other things22:29
seb128k, hopefully things migrate by the time I start my day tomorrow :)22:29
vorlonif they haven't, then I am having a very bad day22:30
seb128let's hope you don't then :)22:30
LocutusOfBorgdpkg: error processing archive /tmp/apt-dpkg-install-wNmXFW/264-libx11-dev_2%3a1.6.9-1_s390x.deb (--unpack):22:45
LocutusOfBorg trying to overwrite '/usr/include/X11/extensions/XKBgeom.h', which is also in package x11proto-dev 2018.4-422:45
seb128LocutusOfBorg, https://packages.qa.debian.org/libx/libx11/news/20200226T170800Z.html22:47
seb128which didn't autosync yet (might be worth doing a manual sync to avoid waiting?)22:48
LocutusOfBorgI did it :D22:48
LocutusOfBorgtjaalton, ^^ I'm trying to fix therion22:48
seb128thx22:48
LocutusOfBorgalso xorgproto syncd22:48
* LocutusOfBorg goes to sleep22:49
seb128calling it a day on that note since the syncs/uploads I just did failed due to that bug22:49
seb128same here22:49
LocutusOfBorgg night22:49
seb128'night!22:49

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!