[00:20] xnox: s390x is being very unhelpful wrt gnucash [00:36] xnox: n/m, it's just guile being segfaulty [01:42] do you guys need any help with some migration or dep8 test? [01:44] man, so many tests in progress === ahasenack is now known as Guest4879 === kirkland is now known as Guest54596 === tyhicks is now known as Guest8831 [04:54] -queuebot:#ubuntu-release- Unapproved: rejected ubiquity [source] (bionic-proposed) [18.04.14.8] [07:43] -queuebot:#ubuntu-release- Unapproved: accepted console-setup [source] (bionic-proposed) [1.178ubuntu2.7] === agateau_ is now known as agateau [10:27] So, uh, shouldn't we be in Debian import freeze now, per the release schedule? [10:27] Or has that changed? [10:29] tsimonq2 did mention in passing he might ask for FF to be pushed back a week, but I have no clue if anything came of that === Guest4879 is now known as ahasenack [10:54] per #topic, we are not in FF yet [11:03] Yeah, I know that, but the release schedule on the wiki says otherwise so something has to give :) [11:04] (also, FF isn't necessarily the same as DIF ...) [11:11] I think there are not enough builders to cope with the huge queue, it will be a while [11:11] just guessing [11:22] That normally isn't a factor, since it's OK for things to land provided they were at least uploaded before the deadline [11:22] I'm counting on that actually [11:22] (You possibly mean autopkgtest runners. The Launchpad build farm is mostly idle right now.) [11:23] true [13:10] -queuebot:#ubuntu-release- New sync: devscripts-el (cosmic-proposed/primary) [40.1] [13:11] -queuebot:#ubuntu-release- New sync: dpkg-dev-el (cosmic-proposed/primary) [37.4] [13:12] slangasek, who cares about guile, when emacs is busted =) please accept above two NEW packages, which got split out of emacs-goodies-el (which i forced synced now) [13:13] once above is in, i will readd ubuntu suite names to the dpkg-dev-el one [13:44] -queuebot:#ubuntu-release- New source: spirv-tools (cosmic-proposed/primary) [2018.2-0ubuntu1] [15:17] did something happen to FeatureFreeze? [15:17] (proxy-requesting because dpb1 posted it in #ubuntu-server) [15:31] slangasek can you approve nodejs upload to bionic please [16:42] -queuebot:#ubuntu-release- New: accepted devscripts-el [sync] (cosmic-proposed) [40.1] [16:42] -queuebot:#ubuntu-release- New: accepted dpkg-dev-el [sync] (cosmic-proposed) [37.4] [16:44] cjwatson: yeah, we should be in DIF now, switch flipped [16:44] -queuebot:#ubuntu-release- New binary: devscripts-el [amd64] (cosmic-proposed/universe) [40.1] (no packageset) [16:45] -queuebot:#ubuntu-release- New binary: dpkg-dev-el [amd64] (cosmic-proposed/universe) [37.4] (no packageset) [16:49] thanks === slangasek changed the topic of #ubuntu-release to: Released: Xenial 16.04.5, Bionic 18.04.1 | Archive: Feature Freeze | Cosmic Release Coordination | Please don't upload things during freezes where you shouldn't, or be prepared to apologise to the release team | We accept payment in cash, check or beer | melius malum quod cognoscis [16:52] xnox: your new emacs-goodies-el has a pile of component-mismatches too [16:53] slangasek, which i solved by dropping emacs-goodies-el from the seed. [16:53] ok [16:53] slangasek, it's the other ones from new that we will want to keep in main, i.e. the debian/control|changelog modes which are now split out [16:55] slangasek: could you promote guile-2.2 to main and once aisleriot migrates, we can demote guile-2.0? [16:57] -queuebot:#ubuntu-release- New: accepted devscripts-el [amd64] (cosmic-proposed) [40.1] [16:57] -queuebot:#ubuntu-release- New: accepted dpkg-dev-el [amd64] (cosmic-proposed) [37.4] [17:04] doko, apw - sigh [17:04] /usr/include/x86_64-linux-gnu/bits/statx.h:25:8: error: redefinition of ‘struct statx_timestamp’ [17:05] /usr/include/linux/stat.h:56:8: note: originally defined here [17:13] ah, infinity ^ [17:14] i think there is misscompat somehow between headers or something [17:20] infinity, slangasek: Ping, I'd like an opinion on bug 1788904 when someone has the chance. [17:20] bug 1788904 in lubuntu-meta (Ubuntu) "Blanket Feature Freeze Exception: Lubuntu's LXQt Transition" [Critical,New] https://launchpad.net/bugs/1788904 [17:21] Basically, I'm asking a policy to be formalized which has seemingly already been in effect for the Ubuntu Desktop Team among others. [17:26] xnox: I can haz larger example? [17:26] infinity, sbuild of systemd fails now =) [17:26] xnox: Oh, that's no big loss. [17:26] * infinity ducks. [17:26] infinity, so $ sbuild -A -d cosmic systemd_239-7ubuntu4 [17:27] infinity, well you asked for a larger example, afterall =) [17:27] xnox: Hahaha. A slightly smaller one? :P [17:28] xnox: But yeah, looks like statx is new to glibc (and the kernel, probably), and needs all the usual guards. [17:28] Hate this dance. [17:28] xnox: Of course, you might also be able to fix it by not including linux/stat.h [17:36] infinity, #include #include -> is that broken to include both now? [17:36] and #include [17:36] xnox: It's almost always wrong for userspace to include anything in linux/ [17:37] Not always always, but almost always. [17:37] there is a mega comment saying that we want to use statx() cause it's exposed from v4.11+ and that it might not be there and fallback to something else. [17:37] i guess if statx is available now in userspace there is no need for linux/stat.h anymore..... [17:38] infinity, i'm just curious if this is a bug in systemd, or headers, or both. [17:38] It's certainly possible for glibc and the kernel to guard against this. [17:38] infinity, i mean i will fix systemd obviously, but not sure if the headers can be "better" (stupid-compatible?!) somehow. [17:39] On the other hand, Florian was the committer, and he's usually smarter than that. [17:39] So unless there appears to be a valid reason to want both, the conflict might be intentional. [17:39] 4/* Missing glibc definitions to access certain kernel APIs */ [17:39] is the heading used for these includes in systemd tree [17:39] And I don't see any reason to want the kernel version, at a glance. [17:39] so possibly obsolete now. [17:40] Unfortunate that an upstream fix will have to have a conditional include. [17:40] But not world-ending either. [17:41] include sys/stat, ifndef foo, include linux/statx [17:41] xnox: I'm assuming Fedora already ran into this, since they probably shipped glibc 2 weeks ago. [17:42] xnox: Unless they've not rebuilt systemd since. [17:42] Now, if I could remember how to find fedora sources.. [17:43] i'll propose conditional upstream i think. [17:43] (systemd) [17:44] Yeah, I'm going to look at what Fedora's done here for kicks. [17:46] Ahh, and that's that answer. systemd in Fedora hasn't been built since mid-July. [17:47] Oh, wait. [17:48] xnox: https://src.fedoraproject.org/cgit/rpms/systemd.git/commit/?h=f29&id=5306894742943e5d46ac705a3f2d5db1a48a0a67 [17:49] xnox: Looks like a Googler has already proposed a patch (and F29 is carrying it). [17:49] xnox: So, cheers. [17:50] infinity, lovely, cause propose to fedora and not to upstream? [17:50] * xnox checks master again [17:51] Gah, GitHub is all wonky. [17:51] WHY. [17:52] xnox: https://github.com/systemd/systemd/commit/75720bff62a84896e9a0654afc7cf9408cf89a38 [17:52] xnox: Looks upstreamish to me. [17:52] infinity: because GH is owned by Microsoft now? *shot* (no but seriously...) [17:53] teward: I suspect it doesn't particularly relate (except maybe Github higher-ups trying to look "busy" and "useful" to not get "fired" and made "homeless" by their new overlords) [17:53] Either way, this dashboard landing page madness is... Busy. [17:53] And useless. [17:53] infinity, it got replaced by https://github.com/systemd/systemd/commit/9c869d08d82c73f62ab3527567858ce4b0cf1257#diff-969b60ad3d206fd45c208e266ccfed38 [17:54] infinity: indeed. (It's why I've been shifting to other providers, or running my own GitLab instead for my projects, but I digress) [17:54] (I was going to ask "what happened to feature freeze" but I just saw the emails about it, so don't mind me :P ) [17:54] xnox: Fancy. Definitely a step in the not-silly direction. Maybe. [17:54] we have thousands of things in autopkgtest queue backlog still.... [17:54] hence to me FF is a bit meaningless [17:55] FF isn't about migration, it's about uploads. [17:55] ^ that [17:55] Certainly not meaningless to say "okay, stop uploading MORE feature changes". [17:56] indeed. which reminds me, roaksoax has a proposed change to Cosmic that'd alter the package structure, that'd need an FFe review right? [17:56] to Cosmic nginx* [17:56] (at this point it's not yet 'tested' or 'accepted' as working so it didn't get included in the latest uploads I did for it) [17:56] xnox: Anyhow, based on those two commits, I'd say the systemd community is accepting that it's their bug, and not a bug in the headers. And Lennart would argue six ways to Tuesday if he even 5% thought it was our bug, so I'm going to call that definitive. :P [17:58] infinity, he may have been on holiday since july and has not seen this commit yet [18:00] xnox: He committed the first one. [18:01] filbranden authored and poettering committed on Jul 15 [18:09] Oh look, and my build failed on two arches, despite testing it earlier. [18:09] * infinity looks. [18:21] -queuebot:#ubuntu-release- Unapproved: rax-nova-agent (bionic-proposed/main) [2.1.15-0ubuntu1~18.04.0 => 2.1.15-0ubuntu1~18.04.1] (no packageset) [18:47] doko: My updated assessment on the KDE autopkgtest situation is that there's enough library interdependencies that a build needs to be done for autopkgtest accuracy. [18:48] doko: Alternatively we're exploring just really tightening the autopkgtest deps on the packages, but Santa was working on that and he's been in and out (likely on vac) lately. [19:47] -queuebot:#ubuntu-release- New binary: linux-signed [ppc64el] (cosmic-proposed/main) [4.17.0-9.10] (core, kernel) [19:48] -queuebot:#ubuntu-release- New binary: linux-signed [amd64] (cosmic-proposed/main) [4.17.0-9.10] (core, kernel) [20:12] -queuebot:#ubuntu-release- New binary: murano-agent [amd64] (cosmic-proposed/universe) [1:3.5.1-0ubuntu2] (no packageset) [20:31] -queuebot:#ubuntu-release- New: accepted linux-signed [amd64] (cosmic-proposed) [4.17.0-9.10] [20:31] -queuebot:#ubuntu-release- New: accepted linux-signed [ppc64el] (cosmic-proposed) [4.17.0-9.10] [20:53] jbicha: guile-2.2 promoted [20:53] jbicha: did you see my question about whether you've received britney nagmails about zc.lockfile? [21:01] anyone care yet about node + rollup? or should I just remove the lot from -proposed? [21:06] slangasek, I care about node+rollup [21:07] at least after all the work I did [21:07] let me try to find a way [21:51] slangasek, is it possible to remove automake-1.15 from archive? useless now [21:51] NBS [21:52] LocutusOfBorg: 'NBS' - it doesn't show up on http://people.canonical.com/~ubuntu-archive/nbs.html , what do you mean? [21:53] the "automake" binary is now part of src:automake-1.16, migrated in cosmic hours ago [21:53] so, the src:automake-1.15 is producing nothing, I don't know what is the best wording :) [21:56] LocutusOfBorg: sounds like a source package whose removal we would pick up via Debian [21:56] [2018-08-10] automake-1.15 REMOVED from testing (Britney) [21:56] oh indeed, still part of unstable, meh [21:57] ok https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=906928 [21:57] Debian bug 906928 in ftp.debian.org "RM: automake-1.15 -- RoQA; binary package now built by automake-1.16" [Normal,Open] [21:57] yah, not worth putting extra effort into those kinds of removals, they'll eventually get reaped [21:57] I run process-removals several times between FF and release [22:02] nice [22:31] -queuebot:#ubuntu-release- New sync: debian-el (cosmic-proposed/primary) [37.6] [22:31] slangasek, one more NEW package needed for the emacs-goodies-el split.... sorry that i missed that one. ^ [22:34] xnox: done [22:34] -queuebot:#ubuntu-release- New: accepted debian-el [sync] (cosmic-proposed) [37.6] [22:38] -queuebot:#ubuntu-release- New binary: debian-el [amd64] (cosmic-proposed/none) [37.6] (no packageset) [23:38] jbicha: and guile-2.0 demoted [23:52] LocutusOfBorg, acheronuk: I see you both retried the kdeconnect/s390x autopkgtest failure; it looks entirely reproducible and looks to me like there's a good chance it's a big-endian bug. How should we proceed with this? [23:53] slangasek: I didn't see your question the first time but I'll take a look at zc.lockfile [23:54] jbicha: ok, thanks for looking - and did you previously receive emails? [23:54] because I'm trying to figure out if we have a bug generally in not notifying people of stuck force-syncs [23:56] I get lots of email from the Release Team :)