[00:27] <slangasek> doko: you working on any other transitions currently? we just collided on uploads for poco :)
[00:36] <slangasek> doko: if you want to do some MIRs for python-pysnmp4 deps, they're all yours ;-)
[00:40] <slangasek> doko: (filed as LP: #1748572 and assigned to openstack team)
[03:58] <slangasek> ok, poco isn't going anywhere fast, I changed it to use mariadb so that it was buildable, now parts of the ros stack are uninstallable because they depend on both poco (mariadb-dev) and gdal + vtk6 (mysql-dev).
[05:15] <rbasak> slangasek: I'd rather not end up with a frankenstein archive like that :-/
[05:17] <rbasak> slangasek: packages should be compatible with both. This might be the first package that isn't. Can it be fixed instead?
[05:18] <rbasak> slangasek: MySQL upstream are generally looking after this kind of failure. We should be giving them an opportunity to fix it first at least. That'll make everyone's lives easier.
[05:20] <slangasek> rbasak: I'm looking now at fixing it
[05:21] <slangasek> but it's cmake, so I'm pouring myself a drink first
[07:34] <slangasek> doko: haha, so the nbdkit maintainer did "document" why the dependency on linux-image build-dependency is needed; and if I replace it with the Ubuntu ones, the build still fails because /boot/vmlinuz* is not readable in Ubuntu
[08:01] <doko> slangasek, infinity: do you want to override the glibc test failure on armhf (misc/tst-ttyname) now fails in autopkg test, but succeeded during the build
[09:45] <doko> http://autopkgtest.ubuntu.com/packages/w/wcc/bionic/amd64
[09:55] <doko> nm
[10:34] <doko> badpkg: Test dependencies are unsatisfiable. A common reason is that your testbed is out of date with respect to the archive, and you need to use a current testbed or run apt-get update or use -U.
[10:34] <doko> autopkgtest [10:05:27]: ERROR: erroneous package: Test dependencies are unsatisfiable. A common reason is that your testbed is out of date with respect to the archive, and you need to use a current testbed or run apt-get update or use -U.
[10:36] <doko> slangasek: force-badtest ubuntu-image/1.3+18.04ubuntu1 -> ubuntu2 update needed
[10:37] <doko> but I still don't understand the wcc failure
[10:41] <doko> ahh, mktemp is gone :-/
[10:55] <doko> why did grub2 land in the unapproved queue?
[11:10] <cjwatson> it always does
[11:10] <cjwatson> and it's because it ships objects for signing
[11:10] <doko> slangasek: https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic/bionic/amd64/c/crmsh/20180210_071110_3730d@/log.gz  you changed the test package from less to vim ...
[11:12] -queuebot:#ubuntu-release- Unapproved: accepted grub2 [amd64] (bionic-proposed) [2.02-2ubuntu7]
[11:12] -queuebot:#ubuntu-release- Unapproved: accepted grub2 [arm64] (bionic-proposed) [2.02-2ubuntu7]
[11:22] <ginggs> doko: why would python-defaults make autodocktools, mgltools-* and raccoon not-installable?
[11:39] <doko> ginggs: it needs python2.7 to migrate. see the crmsh question above
[11:40] <ginggs> doko: ah, thanks
[11:41] <doko> and the openvswitch to succeed on i386, triggered by sphinx
[11:46] <jbicha> doko: can you promote gtkspell3? it was in main since forever
[11:49] <jbicha> https://launchpad.net/ubuntu/+source/gtkspell/+publishinghistory
[11:50] <jbicha> it looks like it dropped out because of a mistake in gnome-software's meson conversion
[12:00] <doko> jbicha: done
[12:25] <ginggs> slangasek: please bump openvswitch hint 'openvswitch/2.9.0~git20180205.5a39582ca-0ubuntu1/i386'
[14:07] <LocutusOfBorg> doko, your gdbm delta makes not much sense to me
[14:07] <LocutusOfBorg> the dietlibc is not pulled into main, because the gdbm dietlibc version is a static library
[14:07] <LocutusOfBorg> can I drop it?
[14:07] <doko> no, it's pulled into main
[14:08] <LocutusOfBorg> how
[14:08] <doko> and build exactly *one* package with a different libc doesn't make any sense
[14:08] <LocutusOfBorg> Built-Using: dietlibc (= 0.34~cvs20160606-7)
[14:08] <LocutusOfBorg> this?
[14:09] <LocutusOfBorg> there is no dependency, just a built using I see
[14:09] <LocutusOfBorg> https://launchpadlibrarian.net/355362029/buildlog_ubuntu-bionic-amd64.gdbm_1.14.1-2_BUILDING.txt.gz
[14:10] <doko> yes, and that's causing the component mismatch
[14:10] <LocutusOfBorg> I'm preparing an upload that disables dietlibc in Ubuntu, and upload in Debian, but I need to understand *exactly* where is the issue
[14:10] <LocutusOfBorg> ok, so the built using is the problem
[14:10] <LocutusOfBorg> fine, lets do it then
[14:10] <doko> you should disable that crap in Debian too
[14:10] <LocutusOfBorg> I'm not the maintainer :)
[14:10] <doko> do a NMU
[14:11] <LocutusOfBorg> somebody requested it a while ago, and they agreed it was worth doing it
[14:11] <doko> not for Ubuntu, we don't want to maintain a second libc
[14:12] <doko> so maybe proposed by the dietlibc maintainers ...
[14:14] <LocutusOfBorg> this is why I'm disabling it conditionally in ubuntu only, but doing a "if derives-from Ubuntu"
[14:14] <LocutusOfBorg> not by maintaining a delta
[14:17] <doko> you still have it in the b-d's
[14:22] <LocutusOfBorg> who cares?
[15:19] <jbicha> doko: uh, sorry for the confusion, we need gtkspell3 in main, not gtkspell
[19:44] <doko> ginggs: libgpuarray autopkg test failure triggered by pocl
[19:47] <doko> jbicha: fixed
[23:32] -queuebot:#ubuntu-release- New binary: node-cli-spinners [amd64] (bionic-proposed/universe) [1.1.0-1] (no packageset)
[23:32] -queuebot:#ubuntu-release- New binary: node-is-generator-fn [amd64] (bionic-proposed/universe) [1.0.0-2] (no packageset)
[23:35] -queuebot:#ubuntu-release- New binary: glyphslib [amd64] (bionic-proposed/universe) [2.2.1-1] (no packageset)
[23:48] -queuebot:#ubuntu-release- New binary: colmap [amd64] (bionic-proposed/universe) [3.4-1] (no packageset)
[23:50] -queuebot:#ubuntu-release- New binary: colmap [i386] (bionic-proposed/universe) [3.4-1] (no packageset)