[00:01] <nacc> cyphermox: yeah, we were up to ubuntu17 in zesty :)
[00:01] <nacc> luckily just back and forth on the tests mostly
[00:01] <cyphermox> I thought it was mostly tests and initramfs, yeah
[00:01] <nacc> yep it was an easy merge, in the end
[00:01] <cyphermox> oh ok
[08:35] <LocutusOfBorg> hello mwhudson golang-pty sync, right?
[08:35] <mwhudson> LocutusOfBorg: um
[08:35] <mwhudson> we have delta for that?
[08:35] <LocutusOfBorg> yes sir, and some ppc stuff
[08:35] <mwhudson> apparently we do
[08:36] <mwhudson> ah, ppc, everyone's favourite platform when it comes to ioctl numbers
[08:36] <mwhudson> oh no, this is ppc as in powerpc as in who cares any more
[08:37] <LocutusOfBorg> soooo will you sync it?
[08:40] <mwhudson> yes
[08:40] <LocutusOfBorg> thanks!
[08:42] <mwhudson> ah yes the war on built-using
[08:42] <mwhudson> LocutusOfBorg: done
[10:28] <LocutusOfBorg> Logan, sorry but I think qbzr still needs that delta :(
[10:30] <LocutusOfBorg> I'm readding it
[12:57] <sil2100> Laney: hey! Do you know if when I re-run an autopkgtest for a github PR, will it be automatically picked up by the PR itself?
[12:57] <sil2100> Laney: I got some failures caused by 'API rate limit exceeded for 91.189.89.216.' in tests that actually poll github through the API
[13:24] <smoser> so i did apt-get install ubuntu-gnome-desktop. select gdm3 at the dpkg prompt. log out, back in. it seems like all i have changes is the system theme. i still have unity panel on the left and the top. and changing background doesn't work.
[13:24] <smoser> is that known/expected ?
[13:25] <ogra_> smoser, why wouldnt it ... if you didnt select another s4ession you will end up in unity again
[13:26] <smoser> yeah, htat makes sense. but i was confused by the theme change.
[13:26] <ogra_> (if you expect to end up in gnome you need to pick gnome)
[13:26] <smoser> yeah.
[13:26]  * smoser logs out. thanks
[13:28] <smoser> ogra_, so, chose 'gnome'. background now is solid black. still unity launcher and panel.
[13:28] <smoser> what am i supposed to pick ?
[13:28] <ogra_> hmm, gnome should have brought you a gnome desktop
[13:29] <ogra_> without unity panel and all
[13:29] <smoser> other options were: gnome classic, gnome on wayland
[13:29] <smoser> and 'ubuntu'
[13:30] <ogra_> right, ubuntu should be unity ... gnome -> plain gnmome
[13:30] <smoser> xerros shows:
[13:30] <smoser> dbus-update-activation-environment: setting GDMSESSION=ubuntu
[13:30] <smoser> dbus-update-activation-environment: setting XDG_CURRENT_DESKTOP=Unity:Unity7
[13:33] <ogra_> what release is that btw ... it works fine for me (without changing to gdm btw) on my 16.04 laptop .... i installed gnome-desktop and just seletced it on lightdm
[13:46] <smoser> Orphis, it is/was Artful.
[13:46] <smoser> i think the issue was that i must selected the session, then typed the password wrong, then typed it right. i think it lost it in between.
[13:46] <smoser> ogra_, (sorry Orphis )
[13:46] <Orphis> Aaaaah
[14:02] <jbicha> smoser: what are you trying to test?
[14:03] <jbicha> if you just want the current state of Ubuntu Artful with GNOME, I think it might be better to just add gnome-shell to your Ubuntu install
[14:03] <jbicha> ubuntu-gnome-desktop will change more settings and install more apps but not all of that will end up being in the default Ubuntu 17.10
[14:04] <smoser> jbicha, well, i'm there now.
[14:04] <smoser> so, "all in"
[14:05] <jbicha> and currently, I think we're leaning towards lightdm instead of gdm but that's still TBD
[14:08] <smoser> jbicha, thanks/
[14:08] <jbicha> lightdm would end up *looking* more like gdm though
[14:13] <Laney> sil2100: Pretty sure it should be.
[14:13] <Laney> If you're hitting a rate limit, slow the test down?
[14:43] <gsilvapt> hello everyone
[14:52] <bdmurray> jbicha: Have you seen my comment on bug 1689093?
[14:59] <jbicha> bdmurray: yes, could we set the hook to unreportable just for certain releases so that the hook can be synced with the current development release too?
[15:04] <bdmurray> jbicha: Sure whatever is in the report e.g. DistroRelease, Package, ...
[15:06] <jbicha> bdmurray: how much in a hurry are you? I'm not very familiar with apport's code so it will take more time for me to do that
[15:07] <bdmurray> jbicha: Enough that'd I do it for you if you could pastebin me some criteria.
[15:13] <rbasak> tjaalton: for bug 1676845, wouldn't it be appropriate to check that vlc still actually works?
[15:13] <rbasak> Similarly opentk.
[15:14] <rbasak> arges, bdmurray: ^
[15:15] <rbasak> I do feel that this is something missing from SRUs currently.
[15:16] <rbasak> A more explicit test plan so we actually check that the result isn't broken during SRU verification. Even just a quick smoke. It seems inappropriate to me to land an SRU if not one single person is known to have looked to see.
[15:16] <tjaalton> rbasak: it works
[15:16] <rbasak> But that's not in our current process.
[15:16] <rbasak> tjaalton: thanks! I'll release Xenial now then.
[15:19] <bdmurray> rbasak: yes, that makes sense me. What cases do you think would require a quick smoke check?
[15:22] <jbicha> bdmurray: how about something like https://paste.debian.net/931810/ ?
[15:22] <rbasak> bdmurray: any situation where checking the bug isn't fixed doesn't implicitly include a smoke check.
[15:23] <rbasak> This vlc bug is an example - it should be verified just by checking dependencies, without actually running vlc.
[15:23] <bdmurray> it could be verified, but we should do more - agreed
[15:24] <rbasak> OTOH, a hypothetical "help button appears in wrong place" bug wouldn't need it because one cannot check that the help button is in the right place without running the app, and that constitutes a smoke test.
[15:25] <bdmurray> jbicha: what about gnome3-next?
[15:26] <jbicha> bdmurray: that's empty right now, but feel free to use the same template there if you want, https://launchpad.net/~gnome3-team/+archive/ubuntu/gnome3-next
[15:26] <rbasak> I'd suggest a [Test Plan] section, but I agree about not wanting to require additional paperwork. So how about making it an ~ubuntu-sru thing? Not required, optional for uploaders, but we'll add one if we feel it is needed, or if there isn't one, we feel it is needed but it isn't clear what it would be, only then would we ask and defer accept until the test plan is agreed.
[15:27] <jbicha> rbasak: just rename Test Case to Test Plan?
[15:27] <rbasak> Test Case is for the specific bug being fixed though.
[15:28] <rbasak> Test Plan is wider (but does of course include Test Case)
[15:28] <bdmurray> rbasak: I think the suggestion seems reasonable and depending on how well it works we could then make it a requirement.
[15:29] <rbasak> "Here's what I plan to do to ensure that I haven't regressed the package in some other way" vs. "Here's what I plan to do to ensure that the bug I'm fixing is actually fixed".
[15:29] <rbasak> bdmurray: thanks. Shall I propose this on ubuntu-devel@?
[15:31] <bdmurray> rbasak: That'd be great.
[15:32] <bdmurray> jbicha: and the ubuntu-gnome hook could be dropped in artful correct?
[15:33] <rbasak> ack
[15:33] <jbicha> bdmurray: no, the GNOME3 PPAs are still being used, at least for now
[15:34] <bdmurray> jbicha: okay, got it
[21:28] <justanotherbody> can anyone tell me where i could find the recipies used to compile the binaries for e.g., libstdc++ ?
[21:36] <nacc> justanotherbody: the src package the generates the binary packge
[21:43] <justanotherbody> can you by chance tell me where to find the source package on http://mirrors.mit.edu/ubuntu/ ? i can only seem to locate the .deb files
[21:44] <nacc> justanotherbody: i don't know if they mirror them or not
[21:44] <nacc> justanotherbody: but you can just use `pull-lp-source <srcpkg>`
[21:45] <justanotherbody> if i didn't have admin privileges to install `pull-lp-source` could you suggest another command?
[21:45] <justanotherbody> (i dont, and it is apparently not installed)
[21:46] <sarnold> apt-get source will download to a current working directory
[21:46] <nacc> sarnold: thanks
[21:46] <sarnold> that mirror doesn't look like it has any sources -- only debs
[21:47] <sarnold> oh that's because the gcc dir has way too many debs
[21:47] <sarnold> never mind
[21:47] <justanotherbody> ?
[21:47] <sarnold> note e.g. http://mirrors.mit.edu/ubuntu/pool/main/g/gcc-6/gcc-6_6.3.0-17ubuntu1.dsc
[21:47] <sarnold> that file describes the other sources used to build the gcc-6 source package and all its binary packages
[21:47] <justanotherbody> so i looked at once of these
[21:48] <justanotherbody> but i didn't understand how i could use it to locate the build script
[21:48] <justanotherbody> so im specifically after the actual build script
[21:49] <justanotherbody> my underlying problem is i need to locally compile a newer version of libstdc++
[21:49] <justanotherbody> for reasons relating to the fact i'm running ubuntu 14.04 (not by choice) and need a version newer than security updates allow for
[21:50] <justanotherbody> and adding a PPO or alternative apt installation mechanism isn't supported by anyone with superuser access
[21:50] <justanotherbody> so i figure i download the library adn compile from source - something ive done before
[21:50] <nacc> building your own libstdc++ is a recipe for pain, it feels like
[21:50] <justanotherbody> but i want to mirror the make arguments and e.g., CFLAGS env variables as closely as possible
[21:51] <justanotherbody> im not disagreeing
[21:51] <nacc> seems highly likely to break stuff :)
[21:51] <nacc> justanotherbody: use `apt-get source` and then see debian/rules in the srcpkg
[21:51] <justanotherbody> which is *exactly why* i'm trying to locate teh build script
[21:51] <nacc> justanotherbody: i mean by version
[21:52] <nacc> justanotherbody: the chagne in version, without rebuilding stuff that depends on it, is likely to break things taht try to load it
[21:52] <justanotherbody> i agree
[21:52] <justanotherbody> my hope is to isolate the lib such taht hte fewest possible things try and load it
[21:52] <nacc> justanotherbody: but in any case, `apt-get source...` and then see what d/rules does
[21:52] <sarnold> you may need to also use newer gcc to compile your tools
[21:52] <nacc> yeah
[21:52] <sarnold> abi changes happen :(
[21:52] <justanotherbody> sarnold: interesting, new gcc is provided
[21:52] <justanotherbody> *interestingly
[21:53] <justanotherbody> by my infrastructure team
[21:53] <sarnold> justanotherbody: if you're going for slightly-hacky you might have some success just unpacking newer debs manually
[21:53] <sarnold> justanotherbody: that'd be a thousand times faster than building your own and it might work well enough
[21:54] <sarnold> justanotherbody: download the .deb file, use ar x to extract it into a data.tar.gz and control.tar.gz files, tar xf the data.tar.gz fiile, and grab the library out of it that way
[21:54] <sarnold> and if it doesn't work for some reason, well, you'll have only spent ten minutes on it anyway
[21:59] <justanotherbody> nacc, sarnold: you have been very helpful. thank you much!
[22:00] <nacc> justanotherbody: yw
[22:00] <infinity> sarnold: What you're looking for is "dpkg-deb -x foo.deb path/" ... ar and tar is pretty unfriendly. :P
[22:00] <sarnold> good luck justanotherbody :)
[22:01] <sarnold> infinity: I can do ar x tar xf blindfolded in my sleep. it takes me ten minutes to find the right manpage to read if I want to skip those.
[22:01] <infinity> :P
[22:01] <infinity> sarnold: dpkg-deb -x unpacks data.  dpkg-deb -R unpacks data and control.  The latter being useful if you want to dpkg --build again after you tweak it.
[22:02] <infinity> sarnold: I mean, keep aring and taring all you want, but the above is probably more user-friendly to tell others. ;)
[22:02] <sarnold> infinity: don't even get me started on apt-cache rdepends, apt-rdepends, and reverse-depends....
[22:03] <sarnold> infinity: but I have to know them myself before I can tel others about them, heh
[23:09] <nacc> cyphermox: fyi,, filed LP: #1689963
[23:10] <rbasak> jbicha: fyi, I got the same es translation notice from Launchpad for gnome-calendar in Zesty. Not a problem from my perspective - just letting you know.
[23:11] <jbicha> ok, I'll fix that up