=== dax is now known as gender === gender is now known as dax [09:12] RAOF, hi :), could you take a look at https://launchpadlibrarian.net/447970389/vala_0.40.17-0ubuntu1_source.changes -- https://bugs.launchpad.net/ubuntu/+source/vala/+bug/1803136 [09:12] Launchpad bug 1803136 in vala (Ubuntu Bionic) "[SRU] Update to vala 0.40.17 in bionic" [Low,Confirmed] [09:35] kirkland, your bikeshed package depends on bzr-fastimport, but fastimport has to disappear due to bzr->brz transition... [09:35] can you please do some update of the packaging? [09:39] hey all, what is the best way to convert a git tree (with debian directory) to a deb package - without any source.tgz? [09:40] enaut[m]: create a source.tgz without debian directory [09:40] enaut[m]: if there's a dsc link dget that [09:40] enaut[m]: which software is that? [09:41] the software is ltsp-manager kind of my own ;) however I think it is so complicated to build the deb-file locally for testing purposes. [09:43] isn't there a switch for just building the whole directory with dh_make - I mean everything is there to build it just does not build automagically? [09:44] LocutusOfBorg: brz provides bzr-fastimport though [10:36] cjwatson, yes true, we should probably make that fastimport a transitional package? meh, don't care. [10:39] LocutusOfBorg: I mean ... it is [10:39] bzr-fastimport | 0.13.0+bzr361+brz1 | focal-proposed/universe | source, all [10:39] LocutusOfBorg: Maybe you're looking in focal rather than focal-proposed [10:55] cjwatson, I just syncd it [10:57] i confused with bzr-webdav, that one probably needs to be kicked out === fenris is now known as Guest55271 === Wryhder is now known as Lucas_Gray === ricab is now known as ricab|lunch [13:39] marcustomlinson: any update with the libreoffice autopkg test? [13:40] doko: will be fixed tomorrow or thursday with the new version [13:42] marcustomlinson: how do you know? https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=943401 [13:42] Debian bug 943401 in libstdc++6 "libreoffice C++ Unit tests failing since 9.2.1-12 ((Failure instantiating exceptionprotector)" [Serious,Open] [13:42] doko: I will disable the test as Debian has done [13:44] could I just ask for a day so I can correlate the fix with the new release [13:54] marcustomlinson: seriously disabling without checking for the issue? [13:55] doko: libreoffice is unfortunately not the only pain in my ass right now [13:55] ;-P [13:56] marcustomlinson: is this test run during the build as well, and if not, how can it be run? [13:57] doko: afaiu it does run during build yes [13:58] though, it looks like Debian have disabled it during build too. hmm, might have to as well [13:59] how can it be run > make check [14:01] marcustomlinson: please could you enable building with the internal cppunit, and see if the problem persists? a stack trace would be good to have === ricab|lunch is now known as ricab [16:16] ahasenack: I have synced postgresql-common 208 [16:16] it was ready now and has the changes we need [16:17] I'll let it run tests over night as they auto-trigger, then we might need to go to p-migraton and check which older tests need custom triggers to properly work [16:25] cpaelzer__: A psql transition in the middle of unwinding the initial autosync? You're brave. [16:26] infinity: it already started by exactly that thundering-herd autosync [16:26] doko: made me aware that several things were failing [16:27] Oh, as in stuff had been fixed to work with 12, but somehow not backward compatible with 11? [16:27] turned out libpq-dev was already pointing to 12 and postgresql-common to 11 still - and things being auto-synced broke on that [16:27] Oh, or that. I thought psql-common was supposed to own libpq-dev for that very reason. [16:27] Maybe I gave the design too much credit. [16:28] infinity: I didn't say I wanted to transition now, but the syncs that already happened and the current state in debian made going back as painful as going forward [16:28] It very clearly comes from psql itself. Special. That seems entirely wrong. [16:28] cpaelzer__: Yup, agreed. I missed the puzzle piece that libpsql-dev is coming from (IMO) the wrong source package. :P [16:29] yeah, agreed - there should be no two packages defining what is currently "the version" [16:29] but for now things are as they are [16:29] Makes the whole "default" selection in psql-common seem sort of pointless. [16:35] Hey juliank, regarding https://bugs.launchpad.net/ubuntu/+source/urdfdom-headers/+bug/1817595, there doesn't seem to be anything in -proposed yet. What is the next step there? [16:35] Launchpad bug 1817595 in urdfdom-headers (Ubuntu Bionic) "[SRU] urdfdom-headers and urdfdom should not use locale dependent parsing for floating point numbers" [Undecided,Fix committed] [16:44] kyrofa: It's in the queue https://launchpad.net/ubuntu/bionic/+queue?queue_state=1&queue_text= - bug some sru people (sru vanguard) about apporivng theme [16:47] Ah yes, thank you! [18:05] hi, I need some apt pinning help. I want to have trusty-proposed main in general downgraded (400, for example), but for one package that comes from it: ubuntu-advantage-tools [18:05] it's not working: https://pastebin.ubuntu.com/p/sVB7TCvTXZ/ [18:06] i.e., I don't want dist-upgrade to suddenly upgrade to anything in proposed, but for that one package [18:08] ahasenack: iirc proposed is handled funny, you have to use something like apt install package=version to use proposed [18:08] hm, wait a sec, it's saying the pin prio is 400, but dist-upgrade is pulling it in [18:10] sarnold: can't do that, this is part of a dist-upgrade from precise [18:10] ahasenack: define "not working" please? [18:10] I want the dist-upgrade to precise to consider just u-a-t from trusty-proposed, not the rest [18:10] rbasak: I was expecting to see a "500" pin prio next to the u-a-t package from trusty-proposed, instead I see 400 [18:10] Your pastebin shows 500? [18:11] no [18:11] 19.6~ubuntu14.04.2 500 [18:11] 400 http://br.archive.ubuntu.com/ubuntu/ trusty-proposed/main amd64 Packages [18:11] that's 400 [18:11] Ah [18:11] oh, hm, what is the 500 next to the version? [18:11] is that the actual package-specific pin? [18:12] and the 400 below for the trusty-proposed line is the overall trusty-proposed pin prio? [18:12] * ahasenack uses better numbers [18:12] ok, yeah [18:13] https://pastebin.ubuntu.com/p/FFN7QKDwhh/ ok, I guess it is working [18:15] I'm not sure what those two numbers mean [18:16] "The APT preferences do not affect the choice of instance, only the choice of version." [18:16] So it's the score by the version string that's important. [18:16] I don't know what the score by the URL means [18:17] it's for the * package I guess [18:17] but yeah, I hadn't noticed the score next to the version ever before [18:18] * ahasenack tries another lengthy dist-upgade [18:19] can we clear our trusty-proposed except for the actively maintained bits (ua client)? https://people.canonical.com/~ubuntu-archive/pending-sru.html [18:23] actually verification was done on https://bugs.launchpad.net/ubuntu/trusty/+source/gnutls26/+bug/1444656 before close of trusty... [18:23] Launchpad bug 1444656 in gnutls26 (Ubuntu Trusty) "GnuTLS TLS 1.2 handshake failure" [High,Fix committed] [18:25] I think the bcmwl update should just be marked won't fix for Trusty and will proceed with doing that in a few... [18:40] Skuggen: hi! [18:42] Skuggen: I'm preparing a mysql 8.0.18 update for eoan, could you please sanity-check my changes before I upload it for building? There are a couple of new files, I want to make sure you agree with where I put them: https://paste.ubuntu.com/p/rcWrc4YDGK/ [18:49] mdeslaur: that looks reasonable to me [18:49] thanks rbasak! [18:49] mdeslaur: though I would like you to use QUILT_REFRESH_ARGS="-p ab --no-timestamps --no-index" please [18:49] Saves extra noise in diffs [18:50] mdeslaur: I'm assuming it passes dep8? [18:50] rbasak: ah, thanks for the quilt config, I'll do that [18:51] rbasak: haven't tested that yet [18:52] mdeslaur: Hi! [18:53] Skuggen: hi there! [18:53] One thing: We support with_protobuf=system (as opposed to using the bundle-built ones), but we should probably test it out more to be sure, and we can make that change later [18:53] I think the protobuf version we build upstream is the same as the one in eoan [18:53] Ah, so maybe we want that in Focal? [18:53] Skuggen: sounds like a change for focal [18:54] and only as long as he protobuf version won't change with point releases [18:55] Skuggen: also, not sure if you care or not, but I had to disable the complete test suite on s390x on bionic with 5.7.28 as it was hanging: https://launchpad.net/~ubuntu-security-proposed/+archive/ubuntu/ppa/+sourcepub/10686865/+listing-archive-extra [18:56] Yeah, I _think_ we'll change to use system protobuf for ubuntu in upstream packages as well, so should stay stable going forward. [18:57] About the ppc64el patch, you're sure it's not needed any more? I haven't really followed up on it, just see that our bug isn't closed [18:57] Skuggen: most of the patch was already applied...a few parts weren't, but the package built and the test suite ran fine [18:58] Hm, maybe we have a duplicate bug for this I'm not aware of. I'll look into it and maybe get it closed, thanks :) [18:58] About s390x, do you know if there's any particular thing that hangs? [18:59] unfortunately, I haven't had time to investigate further... [18:59] I think we only enforce the test results for amd64 and i386 for 5.7 builds [18:59] * mdeslaur orders s390x off of amazon [18:59] Hehe, yeah I don't really have anything except i386, amd64 and arm64 available :) [19:00] and it runs fine with xenial and disco... [19:01] I'm still awaiting autopkgtest results to make sure nothing regressed other than the test suite [19:01] Could it be related to the openssl change? [19:01] Though if it works on older releases maybe it's a change in a dependency that causes it [19:02] I thought that may be the case, but I rebuilt the old one with openssl enabled, and it worked [19:03] Skuggen: I'll figure out how to access an s390x builder box when I get a minute and if I figure it out, I'll let you know [19:03] Thanks! [19:05] Skuggen: another question: with 8.0, when --skip-grant-tables is used, mysql 8.0 won't bind to a network port...is there any way to override that? [19:06] "If the server is started with the --skip-grant-tables option to disable authentication checks, the server enables skip_networking to prevent remote connections. " [19:06] this is causing php7.3 to FTBFS [19:07] Yeah, I saw. You could try --skip-networking=off, but I also mentioned it to rbasak that it didn't seem like skip-grant-tables was actually needed there [19:07] Since --initialize-insecure was used, the root user should be passwordless [19:07] Skuggen: I tried --skip-networking=off and it didn't work [19:08] oh hrm, interesting [19:08] ultimately, we're starting an instance of mysql for nothing...none of the mysql tests even run since we moved to php7.0... [22:21] ricotz: Hm. I think I may have a rather different opinion as to what goes into a bugfix-only release than the vala developers 😕 [22:34] RAOF, hmm? [22:43] Oh, there just seen to be rather a lot of changes which don't fix a bug. [22:43] (and are not obviously in the service of fixing a bug) [22:43] RAOF, could you give an example? [22:45] prevent asserts and correcting bindings might be easy, but fixing memory leaks are not that obvious to see in the plain diff, so are fixing c-warnings in the generated code [22:47] RAOF, e.g. "Support static methods in error-domains" doesn't sound like bug-fix, but it actually is [22:51] Aha! [22:53] RAOF, irony? [22:55] Oh, no. Just that it makes more sense. [22:56] ok, the method implementations weren't emitted at all before [22:56] I still want to know what the testing plan is (in the bug, please) [22:56] of course the diff has grown a lot, given the age of the bug report [22:57] Unless I've missed it! [22:57] it is/was in the description which included a mass-rebuild at that time [22:57] And quite a lot of the diff is generated code. [22:58] use the link in the last comment [22:58] https://gitlab.gnome.org/GNOME/vala/compare/0.40.8...0.40.17 [22:59] (the mass rebuild link appears to be broken?) [23:01] I already commented including this information [23:03] Ah, thanks. I didn't see that. [23:05] RAOF, https://launchpad.net/~ricotz/+archive/ubuntu/vala-sru/+packages [23:06] Ta [23:08] Did you do any smoke testing on the resulting packages? [23:10] RAOF, only a few in the past, many packages have a testsuite [23:10] and a lot of the package are only generating a vapi from their gir [23:26] RAOF, g2g, eod here [23:27] Thanks for your clarifications!