[02:10] -queuebot:#ubuntu-release- New binary: cinder [amd64] (cosmic-proposed/main) [2:13.0.0~b3-0ubuntu1] (openstack, ubuntu-server) [03:42] -queuebot:#ubuntu-release- Unapproved: gnome-control-center (bionic-proposed/main) [1:3.28.2-0ubuntu0.18.04.1 => 1:3.28.2-0ubuntu0.18.04.2] (ubuntu-desktop) [05:11] -queuebot:#ubuntu-release- New binary: nsync [ppc64el] (cosmic-proposed/universe) [1.20.1-2] (no packageset) [05:11] -queuebot:#ubuntu-release- New binary: nsync [s390x] (cosmic-proposed/universe) [1.20.1-2] (no packageset) [05:13] -queuebot:#ubuntu-release- New binary: nsync [amd64] (cosmic-proposed/universe) [1.20.1-2] (no packageset) [05:13] -queuebot:#ubuntu-release- New binary: nsync [i386] (cosmic-proposed/universe) [1.20.1-2] (no packageset) [05:13] -queuebot:#ubuntu-release- New binary: nsync [arm64] (cosmic-proposed/universe) [1.20.1-2] (no packageset) [05:13] -queuebot:#ubuntu-release- New binary: nsync [armhf] (cosmic-proposed/universe) [1.20.1-2] (no packageset) [07:33] sil2100: Can we let python-apt go from proposed to release? apport fails a bit on i386, but it does most of the time (fix pending). I want to upload a new python-apt with f/e locking soon, going back to the 1.7 branch, so I'd like the revert to stable 1.6 to go in now so we don't break the py3.7 migration [07:34] (The new one might cause failures again, I don't know yet) [07:36] (Fix is a strong word for the apport stuff - it connects to real Launchpad which produces 502 and friends sometimes, hence I'm adding automatic retrying of LP API calls on those errors) [07:53] juliank: looking at the apport failure itself, strange thing on the failing case I see [07:54] sil2100: What's strange? It faisl to install packages from launchpad because launchpad times out [07:56] juliank: so test_install_packages_from_launchpad times out? There's no retries done right now, right? [07:56] sil2100: Yeah, you essentially see the same thing in https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-cosmic/cosmic/i386/a/apport/20180725_010756_63e4f@/log.gz [07:56] which was for qterminakl [07:56] I was just wondering since the printed out API urls aren't properly urlencoded, just hope they're not used literally as they're printed out [07:57] Well, it does not time out client side, it times out on the server side and LP gives you the 502 error [07:57] You must have seen those in the browser too [07:58] The printed out urls felt strange to me, but I guess they're probably are properly urlencoded before the actual API call happens [07:58] Anyway, yeah, let me hint it [07:58] yeah, they work fine in the other runs [08:36] apw, for some reasons php-net-ldap2 migrated again :/ [08:36] instead of kicking it again to proposed, what about ignoring the testsuite? autopkgtest for php-net-ldap2/2.2.0-3ubuntu1: amd64: Regression ♻ , arm64: Regression ♻ , armhf: Regression ♻ , i386: Regression ♻ , ppc64el: Regression ♻ , s390x: Regression ♻ [08:37] this way we can move forward with php, and not demoting it forever, since it might be just a testsuite bug [08:37] (the failure is in phpunit changes in testsuite, it doesn't seem to affect the normal usage, but I'm not a php savvy man anymore= [08:43] LocutusOfBorg, sigh, or it should have been blocked in -proposed via a bug [08:50] -queuebot:#ubuntu-release- Unapproved: ceph (bionic-proposed/main) [12.2.4-0ubuntu1.1 => 12.2.7-0ubuntu0.18.04.1] (desktop-core, ubuntu-server) [08:50] sil2100: hey, I'm getting spammed with emails about that new ubuntu-image PR because autopkgtest-web can't post its results back [08:50] maybe something broke with the recent change to require 2fa and it needs a new token or something? could you help please? [08:51] if you go to autopkgtest-web0 and run "DEBUG=1 ~/autopkgtest-cloud/webcontrol/update-github-jobs bionic-amd64-ubuntu-image-156-1350dac39be5aa92eadde8c1ebb558e730f4829f" you can see it's getting a 404 [08:51] which according to https://developer.github.com/v3/troubleshooting/ means authentication problem [09:19] apw, but why blocking a set of packages if just testsuite is broken? isn't an hint better in this case? [09:22] LocutusOfBorg, possibly so, but you seem to be just assuming the test suite is broken, rather than the package is broke [09:22] "just" [09:26] to be sure an ordering question [09:26] dovecot 2.3.2.1 is in c-proposed and there waiting on the new queue [09:27] I know that later on we will need to push a new dovecot-antispam no change rebuild against the new ABI [09:27] can I upload the latter already while the former is waiting in new queue? [09:28] or would the build not pick up the new ABI from dovecot in -proposed [09:28] I don't think the rebuild will see the package in the new queue (whether new source or binary). [09:28] TL;DR if in new queue, is it never the less available and will be pulled on builds [09:28] So I think you need to wait. [09:28] Oh [09:28] OK :) [09:28] the above was a question [09:28] the TL;DR [09:28] add a ? [09:28] Oh [09:29] Well in that case I think my answer stands. [09:29] great [09:29] I believe binaries in new are invisible to builds. [09:29] (since they are not published until accepted) [09:29] I understand, but the particular binary that the rebuild would need is not in NEW [09:30] only a few other new binaries that were added to the dovecot source [09:30] Is the new binary build dep visible and installable by apt with cosmic-proposed enabled? Eg. can chdist on cosmic-proposed see it? [09:30] That's I think the crucial question. [09:34] no, even though the LP page says "published" the new versions do not show up yet in a system with proposed enabled as of now [09:35] * cpaelzer continues to wait then [09:36] You could use a versioned build-dep if you wanted [09:36] to wait until it is available? [09:36] I guess that is too mcuh for the case here [09:37] * Laney shrugs [09:37] hehe [09:37] Just giving you a suggestion that will let it work automatically. If you want to be a human poll(), be my guest. [09:38] I don't want to change the package, currently is is a sync with a bunch of no change rebuilds [09:38] the versioned dep would break that [09:38] No. [09:38] apw, the new php changed the naming of *test* classes, and the ~15 packages I had to patch, were just lots of sed in the testsuite code [09:38] no? [09:38] please explain [09:38] Make it a buildN version and add a changelog comment saying why you did it [09:38] but I honestly don't care too much, the package is out of debian testing, so removing is fine too [09:38] I can open a bug if you hint it [09:39] autosync only looks at the version, so the diff will still be killed the next time [09:39] cpaelzer, right things in New do not yet exist, as we might reject them [09:39] in an archive pool sense [09:39] Laney: that is very interesting - so a ...build1 delta will still be sync killed [09:39] i have a shell script that runs curl and grep-dctrl to see when things are /really/ published... [09:39] didn't know that [09:39] Thanks ++ [09:41] 👍 [09:43] Laney: hey! Ugh [09:43] Laney: ok, didn't know anything about it, I'll look into it today [09:43] sil2100: thanks [09:44] There's some ubuntu-autopkg-bot thing that's probably somehow involved [09:44] Maybe it lost permissions or something [09:44] LocutusOfBorg: Is this really actually hard to fix? And did it need 5 retries per arch? [09:46] -queuebot:#ubuntu-release- Unapproved: protracker (bionic-proposed/universe) [2.3d.r92-1 => 2.3d.r92-1ubuntu0.1] (no packageset) [09:53] Laney, for me it is hard to fix, and retries were part of a bunch of fixes I did, I just didn't remove the package from the list [09:54] we got lots of new phpunits in the meanwhile [09:54] btw the package is unmaintained even upstream, so I don't know, how to fix [09:54] I fixed the renaming, but testsuite fails after that point [09:55] I honestly would remove unmaintained upstream packages, in case nobody here or in Debian wants and know how to fix them [09:58] LocutusOfBorg: I don't have a particular problem with kicking stuff out. I do have a problem with saying that "just" testsuite breakage is fine. [09:58] I don't think we know if the thing is busted or not given how early it breaks. [09:59] and I do request people to not submit excessive retries please [10:12] Laney, ack on all of them [10:13] Maybe that upload fixes it and you can argue about the removal under less pressure. :P [10:18] mmm I did that change already... [10:18] but maybe something else changed in the meanwhile [10:27] dunno, the error looked pretty clear to me [10:29] I even looked at the new prototype in phpunit to see differences, and I found that return issue [10:30] but maybe I crafted a wrong patch [10:32] oh, hold on, NULL is different from '', probably I changed only the return value and not the parameters passed in the function [10:37] :> [11:10] -queuebot:#ubuntu-release- New binary: r-cran-qap [ppc64el] (cosmic-proposed/universe) [0.1-1-1] (no packageset) [11:10] -queuebot:#ubuntu-release- New binary: r-cran-tsp [ppc64el] (cosmic-proposed/universe) [1.1-6-1] (no packageset) [11:10] -queuebot:#ubuntu-release- New binary: r-cran-qap [s390x] (cosmic-proposed/universe) [0.1-1-1] (no packageset) [11:10] -queuebot:#ubuntu-release- New binary: r-cran-tsp [s390x] (cosmic-proposed/universe) [1.1-6-1] (no packageset) [11:11] -queuebot:#ubuntu-release- New binary: golang-github-gdamore-tcell [amd64] (cosmic-proposed/universe) [1.1.0-1] (no packageset) [11:11] -queuebot:#ubuntu-release- New binary: r-cran-qap [amd64] (cosmic-proposed/universe) [0.1-1-1] (no packageset) [11:11] -queuebot:#ubuntu-release- New binary: r-cran-tsp [amd64] (cosmic-proposed/universe) [1.1-6-1] (no packageset) [11:11] -queuebot:#ubuntu-release- New binary: golang-github-juju-collections [amd64] (cosmic-proposed/universe) [0.0~git20180717.9be91dc-1] (no packageset) [11:11] -queuebot:#ubuntu-release- New binary: r-cran-tsp [i386] (cosmic-proposed/universe) [1.1-6-1] (no packageset) [11:11] -queuebot:#ubuntu-release- New binary: r-cran-qap [i386] (cosmic-proposed/universe) [0.1-1-1] (no packageset) [11:14] -queuebot:#ubuntu-release- New binary: python-ulmo [amd64] (cosmic-proposed/universe) [0.8.4+dfsg1-1] (no packageset) [11:14] -queuebot:#ubuntu-release- New binary: r-cran-qap [armhf] (cosmic-proposed/universe) [0.1-1-1] (no packageset) [11:15] -queuebot:#ubuntu-release- New binary: r-cran-qap [arm64] (cosmic-proposed/universe) [0.1-1-1] (no packageset) [11:15] -queuebot:#ubuntu-release- New binary: r-cran-tsp [armhf] (cosmic-proposed/universe) [1.1-6-1] (no packageset) [11:15] -queuebot:#ubuntu-release- New binary: r-cran-tsp [arm64] (cosmic-proposed/universe) [1.1-6-1] (no packageset) [11:29] Laney: I think the u-i issue should be fixed now, give me a sign if you still get the spam [11:29] sil2100: sweet! [11:29] check the PR to see if the results came in [11:31] Yep! [11:31] So all good [11:31] 🕺 [11:31] good work! [11:32] was the bot account losing its authorisation right? [11:32] Yeah, got kicked out of CanonicalLtd because of missing 2fa [11:32] that ol thing [11:32] So actually it was IS that did most of the work here ;) [11:33] TEAM! [11:58] -queuebot:#ubuntu-release- Unapproved: open-iscsi (xenial-proposed/main) [2.0.873+git0.3b4b4500-14ubuntu3.4 => 2.0.873+git0.3b4b4500-14ubuntu3.5] (ubuntu-desktop, ubuntu-server) [13:39] -queuebot:#ubuntu-release- New source: python-blazarclient (cosmic-proposed/primary) [2.0.0-0ubuntu1] [13:57] -queuebot:#ubuntu-release- Unapproved: open-iscsi (bionic-proposed/main) [2.0.874-5ubuntu2 => 2.0.874-5ubuntu2.1] (ubuntu-desktop, ubuntu-server) [14:28] hello, http://autopkgtest.ubuntu.com/packages/s/saods9/cosmic/arm64 can we please hint it? [14:28] it is regressed in release [15:31] -queuebot:#ubuntu-release- New: accepted gkl [amd64] (cosmic-proposed) [0.8.5+dfsg-1] [15:31] -queuebot:#ubuntu-release- New: accepted golang-github-gogo-googleapis [amd64] (cosmic-proposed) [1.0.0-1] [15:31] -queuebot:#ubuntu-release- New: accepted gotest.tools [amd64] (cosmic-proposed) [2.1.0-1] [15:31] -queuebot:#ubuntu-release- New: accepted photutils [arm64] (cosmic-proposed) [0.5-1] [15:31] -queuebot:#ubuntu-release- New: accepted photutils [i386] (cosmic-proposed) [0.5-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-uinput [amd64] (cosmic-proposed) [0.11.2-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-uinput [armhf] (cosmic-proposed) [0.11.2-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-uinput [ppc64el] (cosmic-proposed) [0.11.2-1] [15:31] -queuebot:#ubuntu-release- New: accepted golang-github-gdamore-tcell [amd64] (cosmic-proposed) [1.1.0-1] [15:31] -queuebot:#ubuntu-release- New: accepted photutils [amd64] (cosmic-proposed) [0.5-1] [15:31] -queuebot:#ubuntu-release- New: accepted photutils [ppc64el] (cosmic-proposed) [0.5-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-uinput [i386] (cosmic-proposed) [0.11.2-1] [15:31] -queuebot:#ubuntu-release- New: accepted golang-github-juju-collections [amd64] (cosmic-proposed) [0.0~git20180717.9be91dc-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-uinput [arm64] (cosmic-proposed) [0.11.2-1] [15:31] -queuebot:#ubuntu-release- New: accepted photutils [armhf] (cosmic-proposed) [0.5-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-uinput [s390x] (cosmic-proposed) [0.11.2-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-ulmo [amd64] (cosmic-proposed) [0.8.4+dfsg1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-flexmix [amd64] (cosmic-proposed) [2.3-14-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-prabclus [amd64] (cosmic-proposed) [2.2-6-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-qap [arm64] (cosmic-proposed) [0.1-1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-qap [i386] (cosmic-proposed) [0.1-1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-qap [s390x] (cosmic-proposed) [0.1-1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-tsp [arm64] (cosmic-proposed) [1.1-6-1] [15:31] -queuebot:#ubuntu-release- New: accepted python-ulmo [amd64] (cosmic-proposed) [0.8.4-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-qap [amd64] (cosmic-proposed) [0.1-1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-qap [ppc64el] (cosmic-proposed) [0.1-1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-gclus [amd64] (cosmic-proposed) [1.3.1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-trimcluster [amd64] (cosmic-proposed) [0.1-2.1-1] [15:31] -queuebot:#ubuntu-release- New: accepted r-cran-qap [armhf] (cosmic-proposed) [0.1-1-1] [15:32] -queuebot:#ubuntu-release- New: accepted r-cran-tsp [amd64] (cosmic-proposed) [1.1-6-1] [15:32] -queuebot:#ubuntu-release- New: accepted r-cran-tsp [i386] (cosmic-proposed) [1.1-6-1] [15:32] -queuebot:#ubuntu-release- New: accepted r-cran-tsp [s390x] (cosmic-proposed) [1.1-6-1] [15:32] -queuebot:#ubuntu-release- New: accepted r-cran-tsp [armhf] (cosmic-proposed) [1.1-6-1] [15:32] -queuebot:#ubuntu-release- New: accepted r-cran-tsp [ppc64el] (cosmic-proposed) [1.1-6-1] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [amd64] (cosmic-proposed) [1.20.1-1] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [ppc64el] (cosmic-proposed) [1.20.1-1] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [amd64] (cosmic-proposed) [1.20.1-2] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [armhf] (cosmic-proposed) [1.20.1-2] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [ppc64el] (cosmic-proposed) [1.20.1-2] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [arm64] (cosmic-proposed) [1.20.1-1] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [arm64] (cosmic-proposed) [1.20.1-2] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [s390x] (cosmic-proposed) [1.20.1-2] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [s390x] (cosmic-proposed) [1.20.1-1] [15:33] -queuebot:#ubuntu-release- New: accepted nsync [i386] (cosmic-proposed) [1.20.1-2] [16:04] -queuebot:#ubuntu-release- New binary: budgie-desktop [ppc64el] (cosmic-proposed/universe) [10.4+git20180806.01.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [16:04] -queuebot:#ubuntu-release- New binary: budgie-desktop [s390x] (cosmic-proposed/universe) [10.4+git20180806.01.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [16:06] -queuebot:#ubuntu-release- New binary: budgie-desktop [amd64] (cosmic-proposed/universe) [10.4+git20180806.01.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [16:08] -queuebot:#ubuntu-release- New binary: budgie-desktop [arm64] (cosmic-proposed/universe) [10.4+git20180806.01.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [16:08] -queuebot:#ubuntu-release- New binary: budgie-desktop [armhf] (cosmic-proposed/universe) [10.4+git20180806.01.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [16:09] -queuebot:#ubuntu-release- New binary: budgie-desktop [i386] (cosmic-proposed/universe) [10.4+git20180806.01.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [16:27] -queuebot:#ubuntu-release- Unapproved: fonts-liberation2 (bionic-proposed/main) [2.00.1-5 => 2.00.1-7ubuntu0.18.04.1] (kubuntu, personal-gunnarhj, ubuntu-desktop) [16:27] -queuebot:#ubuntu-release- Unapproved: fonts-liberation (bionic-proposed/main) [1:1.07.4-5 => 1:1.07.4-7ubuntu0.18.04.1] (desktop-core, personal-gunnarhj) [17:25] -queuebot:#ubuntu-release- New binary: networking-bagpipe [amd64] (cosmic-proposed/universe) [9.0.0~b3-0ubuntu1] (no packageset) [17:51] Several packages in cosmic-proposed are blocked on ffmpeg 7:4.0.2-1. Is anybody working on that? [19:27] hello, anybody please update mariadb hint to mariadb-10.1/1:10.1.34-1ubuntu2 [19:33] I can't do it, but hear you can send MPs now :) [19:33] (but I hear) [19:46] -queuebot:#ubuntu-release- New binary: python-thriftpy [ppc64el] (cosmic-proposed/universe) [0.3.9+ds1-1ubuntu2] (no packageset) [19:46] -queuebot:#ubuntu-release- New binary: python-thriftpy [s390x] (cosmic-proposed/universe) [0.3.9+ds1-1ubuntu2] (no packageset) [19:46] -queuebot:#ubuntu-release- New binary: python-thriftpy [i386] (cosmic-proposed/universe) [0.3.9+ds1-1ubuntu2] (no packageset) [19:48] -queuebot:#ubuntu-release- New binary: python-thriftpy [amd64] (cosmic-proposed/universe) [0.3.9+ds1-1ubuntu2] (no packageset) [19:48] -queuebot:#ubuntu-release- New binary: python-thriftpy [armhf] (cosmic-proposed/universe) [0.3.9+ds1-1ubuntu2] (no packageset) [19:49] -queuebot:#ubuntu-release- New binary: python-thriftpy [arm64] (cosmic-proposed/universe) [0.3.9+ds1-1ubuntu2] (no packageset) [20:22] GunnarHj: Yeah, it's a tangled mess at the moment. [20:25] I think it would be appropriate to reject budgie-desktop from cosmic NEW because of https://lists.ubuntu.com/archives/ubuntu-desktop/2018-August/005408.html [20:28] jbicha: I would agree that it's not good that the uploader did that, but I disagree that it is grounds for rejection. [20:28] If that's the only problem with the package, it's relatively simple to fix. [20:29] well anyway, I filed bug 1786107 instead [20:29] bug 1786107 in budgie-desktop (Ubuntu) "budgie file conflict with caffeine" [High,New] https://launchpad.net/bugs/1786107 [20:30] I think it'd be more painful to fix when users have had those files taken over by this other new package [20:31] as long as it doesn't leave -proposed, it won't cause much pain. I was just thinking rejecting from NEW keeps it out of -proposed too [20:31] It's also a nonsensical relationship. [20:31] It's using a made-up version of caffeine. [20:32] A versioned breaks/replaces like that is only sane if the target package is removing those files at that version. [20:34] Oooh, while you're here, I have one for you. [20:34] mysql-server-core-8.0 breaks/replaces mysql-server-core-5.7. [20:34] Normally unversioned breaks/replaces are a red flag, AIUI. [20:35] But in this case, we seem be artifically inserting one for no real reason. Can this breaks/replaces remain unversioned? [20:35] Or perhaps we should move towards mysql-server-core instead? [20:35] We once shipped two MySQL versions in one release (Trusty IIRC), but I never want to do that again anyway. [20:36] (because though it is technically unversioned it is effectively versioned by the version being in the package name) [20:37] infinity: ^ [20:37] -queuebot:#ubuntu-release- New: rejected budgie-desktop [amd64] (cosmic-proposed) [10.4+git20180806.01.933f78fc03d-0ubuntu1] [20:37] -queuebot:#ubuntu-release- New: rejected budgie-desktop [armhf] (cosmic-proposed) [10.4+git20180806.01.933f78fc03d-0ubuntu1] [20:37] -queuebot:#ubuntu-release- New: rejected budgie-desktop [ppc64el] (cosmic-proposed) [10.4+git20180806.01.933f78fc03d-0ubuntu1] [20:37] -queuebot:#ubuntu-release- New: rejected budgie-desktop [arm64] (cosmic-proposed) [10.4+git20180806.01.933f78fc03d-0ubuntu1] [20:37] -queuebot:#ubuntu-release- New: rejected budgie-desktop [s390x] (cosmic-proposed) [10.4+git20180806.01.933f78fc03d-0ubuntu1] [20:37] -queuebot:#ubuntu-release- New: rejected budgie-desktop [i386] (cosmic-proposed) [10.4+git20180806.01.933f78fc03d-0ubuntu1] [20:38] rbasak: An unversioned Breaks/Replaces is a Conflicts, for all intents and purposes, and should be. [20:39] I was under the impression that it was mildly preferable to allow the avoiding of deconfiguring before replacement [20:39] (this matters more for mysql-server-5.7 vs -8.0) [20:39] rbasak: (At the dpkg level, it's subtly differnt, and *wrong* for it to be a Breaks/Replaces, as Replaces lets you arbitrarily overwrite files without removing the original package, and not having a version constraint on that can lead to unforseen consequences of file disappearance) [20:39] rbasak: People using Breaks/Replaces for Conflicting packages are just wrong and got bad advice from somewhere. [20:40] OK [20:40] rbasak: Breaks = I'm compatible with versions outside the given constraint, Replaces = I overwrite files in versions that match the constraint, Conflicts = Fuck this package, I hate it and can't coexist. [20:41] rbasak: Basically, if your Breaks/Replaces can't be sanely versioned, it's a Conflict. If your Conflict CAN be versions, it's a Breaks (and maybe B/R if file overlaps). [20:48] infinity, jbicha quick clarification if poss on the caffeine situation. If caffeine is split into two packages one for icons, one for the main body, how is that going to help? On upgrade, both packages will still be installed and will clash with what upstream budgie has done [20:49] fossfreedom: I'm not sure splitting is the answer here. It looks like budgie-desktop includes a full version of caffeine. Correct me if I'm wrong... [20:50] fossfreedom: either they're the same files and you should both depend on a package that provides them, or they're entirely different semantaically and you can ideally rename to avoid conflict to permit concurrent installation of the same packages, or declare a Conflicts. [20:50] Oh [20:50] fossfreedom: Which means that budgie-desktop shouldn't be arbitrarily overwriting files from caffeine, it should either just straigh tup conflict with it, or the embedded copy should be namespaced to not conflict. [20:50] But also, unembed? [20:50] fossfreedom: Or, yes, budgie-desktop shouldn't ship any of that, and should just depend on caffeine instead. [20:51] it has its own caffeine budgie applet now infinity - built in the same source package [20:51] fossfreedom: Yes. I see that. Is it somehow better/different? [20:52] The difference is that it is native to budgie and as one extra capability not in the appindicator - you can add a timer. Other than that - similar, using the same icons [20:52] has [20:53] fossfreedom: If it's indeed better/different, it should either be namespaced to not conflict, or you should just conflict with caffeine (and stop including it in your desktop seed :P) [20:54] fossfreedom: The Breaks/Replaces is entirely wrong here, as you don't want to just be overwriting files from other packages, you want to not coexist with them. [20:54] not really understanding the terminology - "namespaced" ? [20:54] fossfreedom: Namespaced, as in you could rename the budgie variant "budgie-caffeine" so it doesn't have any file overlaps. [20:54] fossfreedom: But if they're basically the same thing, I don't see why you'd want both installed at all. [20:54] fossfreedom: So a Conflict makes more sense. [20:55] rbasak, I did it once or twice, but I think it is not worth the double effort... [20:55] hmm - I would have to devise a patch to build a binary - its not something upstream would be interested in doing - they insist all applets to be installed in one go (in this case "budgie-core") [20:56] fossfreedom: Erm, I'm not suggesting building another binary. [20:57] Just replacing the breaks/replace with a provides/conflicts/replace ? [20:57] fossfreedom: I'm suggesting either the caffeine stuff in budgie should be renamed to make it clear it's a budgie fork of caffeine, or budgie-desktop should just "Conflict: caffeine" [20:57] fossfreedom: Not provides/conflicts/replaces, no. Just a Conflict. [20:58] k - we can do that [20:58] https://code.launchpad.net/~costamagnagianfranco/britney/hints-ubuntu/+merge/3527837 [20:58] rbasak, ^^ do you like it? [20:58] fossfreedom: provide/conflicts/replaces is a magic trio for interchangeable metapackages (like mail-transport-agent). Unless you're suggesting budgie-desktop would be an approrpriate thing to get when a GNOME user asked to install caffeine, that would not be appropriate. [20:58] It doesn't seem right to me to have the conflict. [20:58] They should be concurrently installable. There's no fundamental reason they can't be. [20:59] One user might want the caffeine applet and another one the budgie one. [20:59] rbasak: Making them coinstallable is maybe the better option, but also required either selling upstream on it, or carrying a large local patch. [20:59] s/required/requires/ [20:59] IMHO, not being hostile to other packages is part of the packagers responsibility. [21:00] rbasak: I don't disagree. Though it seems mostly academic here, as caffeine has no rdeps other than (ironically) ubuntu-budgie-desktop [21:00] People use caffeine directly [21:01] People also like having multiple desktop environments concurrently installed. [21:01] (eg. for a two-user home computer) [21:02] fossfreedom: Does the budgie version provide all the same functionality as the non-budgie version, just with a few extra bits? [21:02] LocutusOfBorg: 404? [21:02] infinity, all the same functionality + the extra timer capability [21:02] What's with the extra 7? :) [21:03] infinity: hi, do you have time to review the x driver removals sometime soon? [21:03] fossfreedom: Is there any intent to push the extra functionality upstream, or will it stay forked forever? [21:03] tjaalton: Sometime, but I'm in the middle of 3 discussions right now. [21:03] sure [21:03] but like this week maybe :) [21:04] LocutusOfBorg: the MP looks fine. I don't know if the actual intent is sane, but I can't review anyway, remember? I'm not on the release team. [21:04] (for the development release) [21:04] infinity, it isnt a fork - written from scratch by Solus (or a contributor). [21:04] fossfreedom: then I think it should take care to install to paths that don't collide. [21:04] fossfreedom: Err, really? They rewrote it from scratch and used the same name? [21:05] fossfreedom: or the packaging should ensure this is the case, and encourage upstream to take that. [21:05] I think the only file names in conflict are the icon names, and there are reasonable reasons to keep the same icon names (to make things slightly easier for icon theme devs) [21:06] They used the hicolor icon names because that is what some/most iconsets use to provide their caffeine icons. The hicolor icons is a fallback if the icontheme in use doesnt have a caffeine icon [21:06] If you want theme devs to be able to replace the icons, then the icons should be in their own package. Or depend on caffeine to get them. [21:07] (or, at the least, they should be provided by _one_ package) [21:07] Either they're treated as the same and should be shipped by exactly one package, or they're treated as different and should be shipped at different paths. [21:09] Eh, they're not even the same icons. [21:09] icon theme devs are unlikely to provide budgie caffeine icons and "caffeine" icons [21:10] infinity, yeah - Solus designed their own for the fallback icons. [21:11] fossfreedom: If it's your intent to keep shipping 'caffeine' as well, for the CLI tools, I'd just stop shipping the budgie icons and make budgie-desktop depend on caffeine. [21:12] Now that I realise they're really not the same thing. [21:12] so hold a local patch then? [21:13] Choosing not to ship two files isn't really a "local patch". [21:14] Your sense of "hold a local patch" is exactly what package maintenance is. Your job is to adjust everything to make it comply with distribution expectations. Ideally this is done by sending patches upstream such that both distribution and upstream needs are met by the same code base. [21:14] Other options include dpkg-divert (please don't), update-alternatives (meh)... [21:14] They are in the Solus source package - so would need to change the meson.build stuff to not install those icon files [21:14] fossfreedom: Or change the packaging to not ship them. [21:14] fossfreedom: You already have a massive override_dh_auto_install, adding two "rm -f" to it wouldn't be hard. :P [21:16] umm - wait though - a depends on the caffeine package would install the appindicator - so budgie users will see both the native and appindicator at the same time [21:16] fossfreedom: caffeine is in your desktop seed currently, so that's likely already true? [21:17] I removed it today from the seeds for cosmic - it is in the seed for bionic - so 18.04 to 18.10 users will travel with the caffeine package installed [21:17] fossfreedom: Also, caffeine-indicator.desktop has: [21:17] OnlyShowIn=GNOME;KDE;LXDE;LXQt;MATE;Razor;ROX;TDE;Unity;XFCE;EDE;Cinnamon;Pantheon; [21:17] Are you one of those? [21:18] GNOME [21:18] That seems poorly thought-out. :P [21:18] I'll send Ikey Doherty around to argue with you :P [21:18] I've heard that's super fun. [21:19] And I already know what his reasons would be. [21:19] Doesn't make 'em right. [21:19] Oh well. [21:19] * rbasak senses some extensive distro patching coming on [21:19] It sounds like the least worse solution is a Conflict on caffeine [21:20] The least bad solution is probably renaming the icons and fixing the two references in the source. [21:20] Indeed, that means you wouldn't get caffeine icon themes "for free", but... It's not caffeine. [21:20] And you ship your own icons anyway, so meh? [21:21] The Conflicts would unreasonably inflict your insanity on others, which is why I think it's just about the worst solution here. [21:21] reluctant - users like theming ... so budgie just using a fallback may look odd [21:21] So fix your broken desktop then [21:22] ... no sway with upstream I'm afraid [21:22] Then you get to patch. [21:23] "Upstream won't like it" is not a valid reason for shipping a bad/hostile distribution package. [21:24] The point of packaging is to produce an integrated distribution for users. If upstream is hostile to that integration, then it sucks to be the package maintainer. [21:24] fossfreedom: Here, while we were talking: http://paste.ubuntu.com/p/gcCvvyjNts/ [21:25] Oh, that's one patched file too many. ;) [21:27] is your other nickname speedy gonzales ?! [21:29] going to have a close look - thanks for the patch :) [21:29] that's not the name i usually give him :/ [21:29] (j/k) [21:31] fossfreedom: http://paste.ubuntu.com/p/q9m6FGVJ3s/ <-- That might even work. [21:31] wxl: Meanie. [21:32] infinity: yep, that's closer ;) [21:33] cheers. Much appreciated. [21:39] stgraber: do you realise some of the unrelated patches in bug 1785499 are changed, including some functional changes? [21:39] bug 1785499 in squashfs-tools (Ubuntu Xenial) "Make squashfs-tools in Xenial in sync with Bionic and Cosmic" [High,Triaged] https://launchpad.net/bugs/1785499 [21:39] stgraber: http://paste.ubuntu.com/p/t37ngVTSTP/ is what I see. [21:40] stgraber: eg. line 236/237 and 252/254 [21:40] stgraber: eg. line 236/237 and 252/253 [21:41] I hate reading diffs of diffs, one sec [21:42] rbasak: Just looks rebased to me. [21:42] rbasak: The patches themselves didn't change. [21:43] rbasak: I think it's diff being utterly confusing here, it suggests that write_xattr was in the stickybit patch which it never was [21:43] (check indentation of +/-) [21:43] Oh [21:43] I'm sorry. [21:43] looking at the individual patches in debian/patches I see the content I'd expect for those patches [21:43] Long day. Perhaps I should stop for the day. [21:43] I did mention on IRC that the tricky part of reviewing this one is making sense of the diff :) [21:43] Usually I'm fine with reviewing a diff of a diff. [21:44] I do it all the time. But I've lost it today. [21:44] I'll look again with fresh eyes tomorrow. [21:44] well, the issue here is that it's diff of diff when some files got re-numbered in debian/patches [21:44] and with multiple of those patches touching the same lines [21:44] so doesn't make for a pleasant output [21:45] stgraber: The renumbering is actually tracked here (I assume he used git diff), but the rebasing is unpleasant to read for some, I imagine. :) [21:46] Yeah I got git to handle the renumbering for me. [21:46] git ubuntu queue sync etc [21:46] That paste is three easy commands away :) [21:46] rbasak: Anyhow, the rebasing looks correct to me, if you want to just review 007 and 008 as the new patches. [21:46] I no longer trust my eyes for today. [21:46] But accept if you're happy? [21:51] stgraber: Did you quilt refresh (or similar) while you were in here? [21:52] stgraber: Some line number changes and an odd context line change are curious. [21:52] The context line change being most curious, cause how was that not previously fuzzy> [21:53] Oh, hahaha, that's not in context, it's in the patch header. [21:53] infinity: I re-synced the patch content with what's in bionic/cosmic [21:54] which look like someone ran a quilt refresh on at some point (possibly the Debian maintainer did) [21:55] stgraber: So, looks like the only change not accounted for as a refresh or rebase is a 1-liner in 0005-add-fstime.patch [21:55] -+time_t forced_time = NULL; [21:55] ++time_t forced_time = (time_t)0; [21:56] yeah, apparently that was changed in a more recent version of that patch... not sure why though the two should be identical in this case. [21:57] Changed twice, even. [21:57] * debian/patches/0005-add-fstime.patch: Fix -Wint-conversion warning by [21:57] initializing the time_t variable with (time_t)-1 instead of NULL [21:57] * debian/patches/0005-add-fstime.patch: initializing the time_t variable [21:57] with (time_t)0 instead of (time_t)-1 to avoid creating all filesystems [21:57] on "Wed Dec 31 23:59:59 1969" [21:57] stgraber: I don't object to backporting that fix too, but it should be mentioned in the changelog, IMO. [21:57] haha, yeah I can see how -1 would be a bad idea :) [21:58] sure, let me update the changelog quickly [21:59] stgraber: Ta. [21:59] Is squashfs upstream dead, or do we just never update? [22:02] The time_t one was the first one I saw. Then I started interpreting every line as a real change without noticing it wasn't a double-barreled diff change prefix [22:02] Anyway, thank you for taking over. [22:02] * rbasak goes to bed [22:02] kinda looked dead, people seem to e-mail LKML for it instead which is a bit odd [22:02] Hrm. Last upstream release 2014, last upstream commit 2017. [22:02] So, not dead, but very slow. [22:03] That's unfortunate. [22:03] Also, upstream git has zstd support (which is the last commit). [22:03] That sounds fun. [22:03] would be neat if they felt like releaseing every once in a while :) [22:03] especially given that they haven't released since a CVE fix... [22:03] Yeah, 4.4 is long overdue, I'd say. :P [22:04] And all our local patches should probably go upstream. [22:04] But I'm very not going to do that. [22:04] -queuebot:#ubuntu-release- Unapproved: squashfs-tools (xenial-proposed/main) [1:4.3-3ubuntu2.16.04.2 => 1:4.3-3ubuntu2.16.04.3] (core) [22:04] But with so much of our infra (livefs, some virt/container images, snaps) relying on squashfs, it'd be nice if we could help revitalize upstream. [22:05] both of the new patches are upstream at least [22:06] -queuebot:#ubuntu-release- Unapproved: rejected squashfs-tools [source] (xenial-proposed) [1:4.3-3ubuntu2.16.04.3] [22:06] stgraber: I didn't really care about the two non-functional changes, but thanks. :) [22:07] well, figured if I was going to be listing changes due to the resync with bionic :) [22:08] I can easily find about half of our patches upstream, maybe more are there and I just can't match them in the git history [22:08] kfreebsd appears missing (not super surprising) and I'm not seeing the lzma magic one either [22:09] ah and no mention of stickybit [22:09] Kay, so maybe it's not that dire, and someone just needs to poke upstream to release more than once every 4 years. [22:09] the rest appears to all be upstream in one way or another at least [22:10] -queuebot:#ubuntu-release- Unapproved: accepted squashfs-tools [source] (xenial-proposed) [1:4.3-3ubuntu2.16.04.3] [22:11] I'm always slightly terrfied when I run into a project that's still hosted on SourceForge. [22:11] Although, SF looks much more pleasant than when it went through that awful "let's try to monetize with nasty ads" phase. [22:13] stgraber: Any plans to do the same thing to trusty, or no carefactor there? [22:14] infinity: wasn't planning on it, mostly because the trusty kernel isn't going to get unpriv file caps so caps will just remain busted with LXD on that one... [22:15] stgraber: Kay. [22:15] just looked and it's on 4.2, so would need to check what else changed between that and that 4.3 we have everywhere else now [22:15] stgraber: Yeah, I wans't suggesting you had to, or even should, as just asking if I should look out for an upload. [22:16] Generally, I despise the "backport all the things" mentality that makes people want to make trusty and xenial be exact copies of bionic. :P [22:16] (THough, for utils like this that almost never change, I get it) [22:17] yeah, I only really cared about the fscaps fix but the different numbering in patches bugged me so I just decided to check what else was missing and include that too, at which point, making debdiff cross-series be clean by reshuffling things a bit usually makes sense [22:17] so trusty -> xenial looks like: [22:17] 132 files changed, 28145 insertions(+), 21015 deletions(-) [22:17] that doesn't look very pleasant, looks like upstream decided to reshuffle things a bit between 4.2 and 4.3 [22:18] Yeah, pass on that, then. [22:18] Honestly, I'm happy for enough functionality lacking in old LTSes that it's easier for you guys stuck on the front lines of Mark-facing projects to just say "No" when suggested that you should backport all the things. [22:19] I'd much rather our users have incentive to upgrade long before an LTS goes EOL. [22:20] well, in this case, the LXD snap is available on 14.04 and will include the new unsquashfs after the SRU has gone through, so that takes care of shiny new LXD on old trusty [22:58] -queuebot:#ubuntu-release- New binary: budgie-desktop [s390x] (cosmic-proposed/universe) [10.4+git20180806.02.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [22:59] -queuebot:#ubuntu-release- New binary: budgie-desktop [amd64] (cosmic-proposed/universe) [10.4+git20180806.02.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [22:59] -queuebot:#ubuntu-release- New binary: budgie-desktop [ppc64el] (cosmic-proposed/universe) [10.4+git20180806.02.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [22:59] -queuebot:#ubuntu-release- New binary: budgie-desktop [i386] (cosmic-proposed/universe) [10.4+git20180806.02.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [23:01] -queuebot:#ubuntu-release- New binary: budgie-desktop [arm64] (cosmic-proposed/universe) [10.4+git20180806.02.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) [23:01] -queuebot:#ubuntu-release- New binary: budgie-desktop [armhf] (cosmic-proposed/universe) [10.4+git20180806.02.933f78fc03d-0ubuntu1] (personal-fossfreedom, ubuntu-budgie) === adam_g_ is now known as adam_g [23:08] -queuebot:#ubuntu-release- Unapproved: console-setup (bionic-proposed/main) [1.178ubuntu2.4 => 1.178ubuntu2.5] (core) [23:14] -queuebot:#ubuntu-release- Unapproved: accepted console-setup [source] (bionic-proposed) [1.178ubuntu2.5] [23:18] exit [23:18] oooops [23:37] -queuebot:#ubuntu-release- Unapproved: console-setup (bionic-proposed/main) [1.178ubuntu2.5 => 1.178ubuntu2.6] (core) [23:44] -queuebot:#ubuntu-release- Unapproved: accepted console-setup [source] (bionic-proposed) [1.178ubuntu2.6]