[00:05] RAOF: upstreams that don't include tests in their tarballs on pypi though.... [00:05] :) [00:18] -queuebot:#ubuntu-release- New: accepted show-in-file-manager [source] (jammy-proposed) [1.1.4-0ubuntu1] [00:21] -queuebot:#ubuntu-release- New binary: show-in-file-manager [amd64] (jammy-proposed/none) [1.1.4-0ubuntu1] (no packageset) [00:43] Eickmeyer: The gmail issue was definitely separate, and may be fixed now if I'm reading my RT email correctly [00:44] cjwatson: Oh so it wasn't due to failed bounces just coinciding with the ISO build failure notice? [00:44] Or, was that timing merely happenstance? [01:11] -queuebot:#ubuntu-release- New: accepted show-in-file-manager [amd64] (jammy-proposed) [1.1.4-0ubuntu1] [01:25] Eickmeyer: Extra messages may have tipped over some kind of threshold maybe, but other than that I think it was coincidence. [01:25] Gmail has started enforcing some more DKIM-like things recently. [01:25] That's kindof what I was thinking and told the team. Thanks, cjwatson. :) [01:26] And lists.u.c apparently now does outbound DKIM signing. [01:27] cjwatson: Gmail are fans of changning things without notice, unfortunately. [04:11] -queuebot:#ubuntu-release- Unapproved: accepted gstreamer1.0 [source] (focal-proposed) [1.16.3-0ubuntu1] [04:22] -queuebot:#ubuntu-release- New binary: tbb [i386] (jammy-proposed/universe) [2020.3-1ubuntu3] (i386-whitelist, kubuntu) [04:24] -queuebot:#ubuntu-release- New: accepted tbb [i386] (jammy-proposed) [2020.3-1ubuntu3] [04:25] -queuebot:#ubuntu-release- New binary: tbb [amd64] (jammy-proposed/universe) [2020.3-1ubuntu3] (i386-whitelist, kubuntu) [04:25] -queuebot:#ubuntu-release- New binary: tbb [s390x] (jammy-proposed/universe) [2020.3-1ubuntu3] (i386-whitelist, kubuntu) [04:30] -queuebot:#ubuntu-release- New binary: tbb [ppc64el] (jammy-proposed/universe) [2020.3-1ubuntu3] (i386-whitelist, kubuntu) [04:30] -queuebot:#ubuntu-release- New binary: tbb [riscv64] (jammy-proposed/universe) [2020.3-1ubuntu3] (i386-whitelist, kubuntu) [04:37] -queuebot:#ubuntu-release- New: accepted tbb [amd64] (jammy-proposed) [2020.3-1ubuntu3] [04:37] -queuebot:#ubuntu-release- New: accepted tbb [riscv64] (jammy-proposed) [2020.3-1ubuntu3] [04:37] -queuebot:#ubuntu-release- New: accepted tbb [ppc64el] (jammy-proposed) [2020.3-1ubuntu3] [04:37] -queuebot:#ubuntu-release- New: accepted tbb [s390x] (jammy-proposed) [2020.3-1ubuntu3] [04:41] -queuebot:#ubuntu-release- New binary: tbb [arm64] (jammy-proposed/universe) [2020.3-1ubuntu3] (i386-whitelist, kubuntu) [04:58] -queuebot:#ubuntu-release- Unapproved: accepted gst-plugins-good1.0 [source] (focal-proposed) [1.16.3-0ubuntu1] [05:04] -queuebot:#ubuntu-release- New: accepted tbb [arm64] (jammy-proposed) [2020.3-1ubuntu3] [05:13] -queuebot:#ubuntu-release- Unapproved: accepted gst-plugins-base1.0 [source] (focal-proposed) [1.16.3-0ubuntu1] [07:44] LocutusOfBorg: I see you made some changes to pybind11 regarding python3.10 compatibility; I'm trying to get the gemmi package to build (never before built in Ubuntu) and running into problems with the pybind11 cmake rules installing to /usr/local/lib/python3.10 instead of /usr/lib/python3, any ideas? [07:49] vorlon: update excuses list a ppc64el regression for open3d, does it need some hints to get in? Anything I can do? [07:50] jochensp: thanks for flagging, we can force a retest that should let it go through [07:50] thx :) [08:02] -queuebot:#ubuntu-release- New sync: numba (jammy-proposed/primary) [0.55.1-0ubuntu1] [08:02] -queuebot:#ubuntu-release- New: accepted numba [sync] (jammy-proposed) [0.55.1-0ubuntu1] [08:23] Hello everyone! We're trying something experimental for ubuntu-archive work - we're trying to ressurect archive admin shifts but in a more manageable way [08:23] https://wiki.ubuntu.com/ArchiveAdministration#Archive_administration_shifts [08:24] We want to allocate strict timelines to be available to do ubuntu-archive stuff [08:24] (we have a google calendar for those as well but still need to figure out how to make it public-public [08:24] ) [08:25] Anyway, since I have an errand today possibly after 11:00 UTC, I'm available already today === sil2100 changed the topic of #ubuntu-release to: Released: 21.10 Impish Indri | Archive: DIF, FF | Highlight ubuntu-archive or sil2100 for archive admin help | Jammy Release Coordination | We accept payment in cash, cheque or gin | melius malum quod cognoscis | infinity, you will be missed [08:49] vorlon, I'll have a look [08:53] In file included from /<>/tests/main.cpp:3: [08:53] 4041 | static char altStackMem[4 * SIGSTKSZ]; [08:54] looks not a pybind issue? [08:58] sil2100, hi! I noticed that there are issues with the Desktop image build process for Jammy, e.g. https://launchpadlibrarian.net/589763997/buildlog_ubuntu_jammy_amd64_ubuntu_BUILDING.txt.gz [08:58] sil2100, is that on your radar already? [08:59] paride, it's a known infra issue, if you read the backlog from the channel from yesterday evening there are some details [08:59] paride: hey! Yeah, this was mentioned by seb128 at the start of the week, I think the Launchpad team is aware of it as it's an infra issue [08:59] seb128, sil2100 oh good, thanks [08:59] paride, sil2100, https://portal.admin.canonical.com/C136237 [09:05] -queuebot:#ubuntu-release- Unapproved: accepted linux-firmware [source] (focal-proposed) [1.187.28] [10:08] vorlon: stuff should install into like usr/lib/python3.10 and then later dh_python would move stuff about. The /usr/local looks odd === sil2100 changed the topic of #ubuntu-release to: Released: 21.10 Impish Indri | Archive: DIF, FF | Highlight ubuntu-archive for archive admin help | Jammy Release Coordination | We accept payment in cash, cheque or gin | melius malum quod cognoscis | infinity, you will be missed [11:49] paride,seb128: The metadata service issue that was affecting livefs builds is supposed to be fixed now. Can you let us know if you see it again? [11:50] cjwatson, will do, thanks! [11:50] could someone kick a build for https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/jammy/ubuntu on amd64 ? [11:51] sil2100, ^ [12:01] thanks Colin, once new pending images are published I'll check that they pass the smoke test and get promoted to current [12:07] -queuebot:#ubuntu-release- Packageset: Removed onetbb from i386-whitelist in jammy [12:07] -queuebot:#ubuntu-release- Packageset: Added libsoup3 to i386-whitelist in jammy [12:14] On it o/ [12:16] ah, crap, thought seb wanted a canary rebuild [12:30] Mm, canary won't work yet, separate issue (though fix in progress) [12:39] cjwatson: just failed https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/jammy/kubuntu/+build/331796 [12:43] -queuebot:#ubuntu-release- New binary: opm-simulators [amd64] (jammy-proposed/universe) [2021.10-4ubuntu1] (no packageset) [12:45] RikMills: hmph. thanks [13:04] RikMills: reported [13:09] -queuebot:#ubuntu-release- New binary: opm-simulators [ppc64el] (jammy-proposed/universe) [2021.10-4ubuntu1] (no packageset) [14:41] juliank, what's the bug tracker to use for autopkgtest infra issues? [14:42] juliank, somehow https://git.launchpad.net/autopkgtest-cloud/commit/?id=8faf684741 isn't working, we had a recent case of retry trigger with a trigger on migration-reference/0 and all-proposed=1 and it ended up enabling proposed and consider the output as a release reference, considering it as a non regression and letting the package migrate [14:42] Commit 8faf684 in autopkgtest-cloud "Reject additional triggers, PPAs, all-proposed for migration-reference/0 autopkgtest-web-64" [14:45] juliank, unsure if making https://git.launchpad.net/autopkgtest-cloud/tree/charms/focal/autopkgtest-cloud-worker/autopkgtest-cloud/worker/worker#n651 conditional to not have proposed set if migration-reference is used would be right? [15:07] vorlon, set(PYTHON_INSTALL_DIR ${PYTHON_SITE_PACKAGES}) [15:07] this is the culprit [15:09] and yes, pybind11 needs patch [15:27] -queuebot:#ubuntu-release- New binary: opm-simulators [arm64] (jammy-proposed/universe) [2021.10-4ubuntu1] (no packageset) [15:41] LocutusOfBorg: right sorry, the "trying to get the package to build" on my part was "comment out the tests which are not worth porting for a completely new package introduced after glibc 2.34 landed". Thanks for probing this further; is the pybind11 patch something you would follow up on? [15:41] I'm pulling this thread btw because of macromoleculebuilder being stuck and depending on an NBS lib [15:41] vorlon, I just finished the patch [15:42] and it works [15:42] haha nice [15:42] but I can't commit [15:42] and I also reverted automake patch [15:42] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1006784 [15:42] Debian bug 1006784 in src:automake-1.16 "automake: fixes for python3.10 distutils changes" [Important, Open] [15:42] looks like dh_python according to stefanor should be smart enough to understand usr/local and to move into usr/lib [15:43] but AFAIK it doesn't do this (also other python libs are suffering from this issue) [15:43] but yes, people using automake locally shouldn't install by default into /usr/lib [15:43] so my automake patch is wrong and I reverted it [15:43] but I think (I asked for clarification in https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1006615) that the right fix is not yet in dh-python [15:43] Debian bug 1006615 in python3-pybind11 "python3-pybind11: Handle Python 3.10's default sysconfig paths" [Normal, Open] [15:44] ah [15:46] LocutusOfBorg: hrm, your email to bug 1006615 also points at an issue in the pybind11 patch, it shouldn't be installing headers to /usr/linclude by default, either [15:46] Bug 1005604 in Files "Gschema error if Marlin is installed too" [Low, Fix Released] [duplicate: 1006615] https://launchpad.net/bugs/1005604 [15:46] (debian bug 1006615) [15:46] Debian bug 1006615 in python3-pybind11 "python3-pybind11: Handle Python 3.10's default sysconfig paths" [Normal, Open] https://bugs.debian.org/1006615 [15:47] indeed please can you add your comment? [15:47] also, looks like gemmi and firewalld are using cmake/automake not dh_python [15:47] so they both needs fixes (automake and cmake helpers) [15:48] they don't use dh_python? [15:48] surely they call dh_python3 [15:49] they don't call dh_python3 for install target=? [15:49] the patch in dh_python is in install function [15:49] https://salsa.debian.org/python-team/tools/dh-python/-/commit/6137db4dc870672615c31c9d1c9535dafe5b0d2a [15:49] Commit 6137db4 in python-team/tools/dh-python "pybuild plugin: Explicitly unpack wheels using the 'deb_system' sysconfig scheme, allowing python3.10 to change the default sysconfig scheme to 'posix_local'. (LP: #1960608)" [15:49] that's not what dh_python3 does [15:49] I don't remember how it works tbh :D [15:50] is it called after dh_install? [15:50] I think you're confusing pybuild's distutils plugin with dh_python3 [15:50] they both come from the same package, dh-python, but they're different things [15:50] yes [15:50] so the package debian/filename.install should refer /usr/local? [15:50] this looks... sad? [15:51] yeah, probably [15:51] rbasak: Hi, if you have a chance we have a small fix to manila in the impish unapproved queue that we'd like to get a review on. [15:51] or an option/env var should be passed to its build system to explicitly specify an install location [15:52] can't I just override dh_install to call dh_python3 before? [15:52] or call whatever is the name of the tool [15:53] dh_python3 won't rearrange the contents of debian/tmp, which is what you're looking for there [15:54] it might do it :D [16:01] coreycb: I'm reviewing manila [16:02] bdmurray: thank you :) [16:04] coreycb: How will people know that policy.json was moved / disabled? [16:06] bdmurray: they aren't notified but if there's a good way to do that I can look into it. if it exists it will be renamed and left in the same directory so that it's not lost. [17:27] RikMills: OK, in theory the metadata service thing is fixed harder now ... [17:30] cjwatson: thank you. can't test it right now as the iso.qa.ubuntu.com still sees that failed build as (building) so won't ley me have another go until the next one from the cron runs in the morning [17:30] that is a separate annoyance I keep meaning to report! [17:35] I'll actually look at livefs build failures for a while rather than ignoring and deleting them [18:13] Is there any chance we can hint lsp-plugins? It's stuck in proposed for way too long because it's misssing armhf. This is reported upstream to Debian with no action. [19:40] Eickmeyer: "hint" - no. The out-of-date binaries need to be resolved by removal or fixing the build failure [19:41] vorlon: Ok, upstream is "meh" about armhf, so I guess we can remove the out-of-date binaries. [19:41] and it's not the sort of bug that we would normally remove architecture support for [19:41] this isn't some fundamental portability issue, this is just bad build system [19:42] ... including bad build system that doesn't even print out the commandline [20:27] -queuebot:#ubuntu-release- New source: qwe (jammy-proposed/primary) [0.1-0ubuntu1~ppa1] [20:27] ubuntu-archive: Please reject qwe, it was supposed to be a PPA upload (sorry!). [20:30] -queuebot:#ubuntu-release- New: rejected qwe [source] (jammy-proposed) [0.1-0ubuntu1~ppa1] [20:30] Eickmeyer, ^ [20:37] apw: Thanks! [20:43] gemmi fixed vorlon [20:44] LocutusOfBorg: cheers! [20:44] g++ -o /tmp/lsp-plugins-1.1.31/.build/src/core/3d/Allocator3D.o -c 3d/Allocator3D.cpp -fPIC -Wdate-time -D_FORTIFY_SOURCE=2 -g -O2 -ffile-prefix-map=/tmp/lsp-plugins-1.1.31=. -fstack-protector-strong -Wformat -Werror=format-security -march=armv7-a -marm -std=c++98 -fno-exceptions -fno-rtti -fdata-sections -ffunction-sections -fno-asynchronous-unwind-tables -fvisibility=hidden -pipe -Wall [20:44] -DLSP_MAIN_VERSION=\"1.1.31\" -DLSP_INSTALL_PREFIX=\"/usr\" -O2 -DLSP_NO_EXPERIMENTAL -lpthread -I"/tmp/lsp-plugins-1.1.31/include" -I/usr/include/opus [20:44] so there's lsp-plugins with VERBOSE=1 [20:45] now to hunt down the wrong -m options :P [20:46] FWIW I did report it to upstream, but they're already moving on to 1.2, and have stated that armhf is experimental, so I doubt we'll see any movement from them. [20:48] -queuebot:#ubuntu-release- Unapproved: command-not-found (impish-proposed/main) [21.10.0 => 21.10.1] (core) [20:50] Eickmeyer: yes, this is a trivial build system patch to get them to not use the wrong target name [20:50] -queuebot:#ubuntu-release- Unapproved: rejected command-not-found [source] (impish-proposed) [21.10.1] [20:51] vorlon: At the very least this will help them and countless others :) [20:57] vorlon, did you already fix it then? [20:57] because I have an armhf patch too if you want [20:57] en route [20:58] nice! [21:05] \o/ [21:07] firewalld and wine-development are fixed [21:57] -queuebot:#ubuntu-release- Unapproved: command-not-found (focal-proposed/main) [20.04.5 => 20.04.6] (core) [22:01] -queuebot:#ubuntu-release- Unapproved: rejected command-not-found [source] (focal-proposed) [20.04.6] [22:10] -queuebot:#ubuntu-release- Unapproved: command-not-found (bionic-proposed/main) [18.04.5 => 18.04.6] (core) [22:12] -queuebot:#ubuntu-release- Unapproved: rejected command-not-found [source] (bionic-proposed) [18.04.6] [22:47] -queuebot:#ubuntu-release- New binary: macromoleculebuilder [amd64] (jammy-proposed/universe) [3.5+dfsg-4ubuntu1] (no packageset) [22:48] -queuebot:#ubuntu-release- New binary: macromoleculebuilder [ppc64el] (jammy-proposed/universe) [3.5+dfsg-4ubuntu1] (no packageset) [22:49] -queuebot:#ubuntu-release- New: accepted macromoleculebuilder [amd64] (jammy-proposed) [3.5+dfsg-4ubuntu1] [22:49] -queuebot:#ubuntu-release- New: accepted macromoleculebuilder [ppc64el] (jammy-proposed) [3.5+dfsg-4ubuntu1] [22:56] -queuebot:#ubuntu-release- New binary: macromoleculebuilder [arm64] (jammy-proposed/universe) [3.5+dfsg-4ubuntu1] (no packageset)