=== Elimin8r is now known as Elimin8er [06:47] cpaelzer: I have a 2.9 snapshot in a PPA [06:47] but its failing some tests on arm64 atm [06:48] https://launchpad.net/~james-page/+archive/ubuntu/bionic/+packages [06:48] ok, thanks for the info [06:48] first thing I normally do with a new release is re-enable all of the skipped tests and try them out again === zyga_ is now known as zyga === fossfreedom_ is now known as fossfreedom [10:03] anyone around who has some familiarity with debhelper, dpkg, debug symbols and how that works in PPA's which generate and publish them? [10:05] I'm endeavouring to backport debhelper >= 11 to xenial for the Ubuntu Cloud Archive, but I'm missing something in the dance between debhelper, pkgbinarymangler and pkg-create-dbgsym [10:12] I'm aware there are alot of dpkg changes between xenial and bionic including the one to support Package-Type overrides which debhelper uses at bionic but hoping to avoiding including dpkg in the backport... === lool- is now known as lool [11:30] cjwatson: man-db ftbfs on ppc64el [11:36] meh no worries - figured it out and have enough of a compat shim for the backport [11:38] jamespage: Erm. [11:38] jamespage: Backporting debhelper should revert the ddeb support. [11:38] jamespage: Any other option will end in tears. [11:39] infinity: yeah I'd got that - the piece I'd missed was the change to dh_strip to stop debhelper building the debug packages [11:39] "+$dh{ENABLE_DBGSYM} = 0 if not $ENV{'DH_BUILD_DDEBS'};" [11:39] was my friend [11:40] infinity: I think this - http://paste.ubuntu.com/26524058/ does the trick [11:40] jamespage: https://launchpad.net/ubuntu/+source/debhelper/10.2.5ubuntu2 was where I swapped methods. [11:40] mwhudson: golang-1.10 ftbfs on armhf [11:41] jamespage: Yep, that looks like it should do it. [11:41] infinity: I'd generally got a bit confused as to which parts pkg-create-dbgsym, debhelper and pkgbinarymangler did [11:41] them diverts ... [11:42] anyway I now know (at least I think I do) [11:42] jamespage: The pkgbinarymangler dpkg-deb divert is not one of my prouder moments. Don't look at that postinst if you value your sanity. [11:42] lol [11:43] * jamespage shuts his eyes and just holds on for a bit [11:43] jamespage: (The part where there's a brief period on every single buildd build where dpkg-deb doesn't exist, in the middle of unpacking, is... Fun) [11:43] infinity: I also enjoyed the bit where pkgbinarymangler copies in part of debhelper as part of its build [11:45] coreycb: I think I have it nailed now - just working the updates into queens-staging to unbreak all the broken builds from friday... [12:05] doko: already dealt with in Debian; it'll auto-sync [12:39] jamespage: great thanks. so will the ppa still be able to generate ddebs? [13:14] coreycb: yes [13:15] jamespage: ok cool [14:01] chrisccoulson: Have you seen that adobe-flashplugin is stuck in -proposed? [14:03] https://community.ubuntu.com/t/flash-version-in-xenial-partner-archive-canonical-com/3865 [15:54] cjwatson, hi, I have a refreshed patch for py-macaroon-bakery 1.1.0: https://pastebin.canonical.com/209182/ I can build the updated package with it. would it be possible for you to update it? [15:58] cyphermox, doko: do you have any status update on bug 1636666? can you please reply to my comment there? [15:58] bug 1636666 in pcre2 (Ubuntu) "[MIR] pcre2" [Undecided,New] https://launchpad.net/bugs/1636666 [15:58] cjwatson, or, https://paste.ubuntu.com/26525107/ [15:58] ackk: Yeah, I guess I can just go with that, just a minute [15:59] mitya57: can it be built using pcre3? [15:59] No :( [15:59] and how much in main can be built using pcre2? [16:00] ackk: (I had half of that but was missing the other half, it seems, not sure why) [16:00] http://doc.qt.io/qt-5/whatsnew59.html — “QRegularExpression now requires the PCRE2 library version 10.20, or later. Support for the PCRE1 library was dropped.” [16:00] so how much can we rebuild? http://people.canonical.com/~ubuntu-archive/transitions/html/pcre2.html [16:01] There is some estimate in https://bugs.launchpad.net/ubuntu/+source/pcre2/+bug/1636666/comments/20 [16:01] Launchpad bug 1636666 in pcre2 (Ubuntu) "[MIR] pcre2" [Undecided,New] [16:01] That is at least half of the packages from that tracker [16:02] who is working on that to build packages with pcre2 if they can? [16:03] I can try to help with that, maybe jbicha and/or LocutusO- would be able to help too. [16:04] apache2 may be a problem though: https://bz.apache.org/bugzilla/show_bug.cgi?id=57471 [16:04] bz.apache.org bug 57471 in Build "Apache 2.4.10 Compilation fails when compiling against pcre2-10.00" [Enhancement,New] [16:04] ackk: Yep, thanks, that works, will upload [16:05] cjwatson, thank you [16:05] done [16:06] doko: I meant at least half of packages from *this* tracker: http://people.canonical.com/~ubuntu-archive/transitions/html/pcre2-main.html (your link includes also universe) [16:07] ahh, I see [16:07] cjwatson, awesome, thanks [16:07] exim4 may be a problem too: https://bugs.exim.org/show_bug.cgi?id=1878 [16:07] bugs.exim.org bug 1878 in Unfiled "add support for pcre2" [Wishlist,New] [16:08] I guess I should collect some bug links and add them to the description [16:08] no one is working on getting glib to not use pcre3 [16:10] on the other hand, tilix and gnome-builder have missing features because Ubuntu's vte2.91 isn't allowed to use pcre2 yet [16:10] php7.2 is missing in this tracker [16:10] silly gnome [16:11] the main tracker is manually compiled so you're welcome to add php7.2 if you like :) [16:11] So we have two solutions here: 1) Have both pcre3 and pcre2 in main, with a goal of removing pcre3 in later release, 2) Have only pcre3, Qt using its own copy and GNOME missing some features [16:12] GunnarHj, ah, crap. Thanks for that. sil2100, can you take care of that please? (I think you approved the upload originally) [16:13] sil2100, if you can't see the scrollback: [16:13] chrisccoulson: Have you seen that adobe-flashplugin is stuck in -proposed? [16:13] https://community.ubuntu.com/t/flash-version-in-xenial-partner-archive-canonical-com/3865 [16:13] Looks like php7.2 supports only pcre3 and php7.3 only supports pcre2 [16:13] Qt is in main because of qtubuntu (kiosk project) and mozc (Japanese input method) [16:14] And php7.3 is not going to be released any time soon… [16:15] chrisccoulson: hey, for this I think you'll need slangasek [16:18] jbicha, mitya57: I would like to avoid a situation where we have both in main, and everybody is comfortable with it [16:19] I understand, but the goal porting everything to pcre2 does not seem achievable in one release cycle [16:19] the current vte2.91 situation isn't comfortable :/ [16:20] Even if we keep pcre2 in universe for this cycle, next time php7.3 will be released and we'll get in the same situation [16:20] so do the promotion after the bionic release? [16:21] I won't mind that, but we will technically have pcre2 in main as part of Qt anyway [16:21] or get some buy-in from the tech board and the security team [16:41] doko: I just added lists of good/bad packages to the description of bug 1636666 [16:41] bug 1636666 in pcre2 (Ubuntu) "[MIR] pcre2" [Undecided,New] https://launchpad.net/bugs/1636666 [16:42] (It is worse than I thought, most of packages don't even have upstream bugs filed for pcre2 move) [16:42] sil2100, you should be able to do it (https://wiki.ubuntu.com/ArchiveAdministration#Handling_updates_to_partner) [17:25] Why does rmadison still show zesty information? [17:36] bdmurray: needs manual editing of snakefruit:~ubuntu-archive/.madison-lite/config. Fixed now [17:36] (should that be added to the EOL checklist, if not already there?) [17:37] cjwatson: thanks [17:38] chrisccoulson: "stuck in proposed" - so you're saying that the adobe-flashplugin versions in *-proposed are tested and ready to copy to release? [17:38] slangasek, yeah [17:39] mitya57: thanks! so I'm still not sure if it's better to delay that promotion until after bionic [17:39] nacc: Its not there https://wiki.ubuntu.com/EndOfLifeProcess [17:40] bdmurray: ack; i feel like that's probably the release team's page to edit :) [17:40] nacc: whatever, I'm gonna edit it [17:40] chrisccoulson: done [17:40] slangasek, thanks [17:41] bdmurray: :) [17:59] bdmurray: process> good call, thanks [19:41] hi, approx. on Fri 26. January I asked on #ubuntu-devel, what I can do, to put the package ganeti on a list for 'big' packages for autopkgtest(-VMs). The reason is http://autopkgtest.ubuntu.com/packages/g/ganeti/bionic/amd64. I've verified, that autopkgtest passed with 4GB of RAM. The default of 1536MB is to small. [19:41] because I'm no dev (just a user), someone on this channel said, that he has done it for me. Now (2018-02-05 15:13:30 UTC) the test still fails, with the error message: Not enough memory on node node1 for creating instance instance1: needed 1024 MiB, available 810 MiB [19:41] so it seems, ganeti is still with default size autopkgtest-VM === zyga is now known as potato === potato is now known as elvis === elvis is now known as zyga [20:59] nacc: I removed kopanocore, so zeroc-ice is the only one not building for php7.2. but mumble depends on that one [21:10] nice. lintian in bionic proposed is uninstallable [21:11] https://paste.ubuntu.com/26526608/ [21:13] ssleay [21:13] oy [21:14] doko: removed as in from the archive? [21:15] doko: or in which sense? I have fixes for it pending (which I will send to Debian in the bug you filled), but it's segfaulting in my testing (not sure if it's related to PHP yet [21:16] doko: i'm also workig onn zeroc-ice, which needs to be updated to 3.7.0 at lelast