/srv/irclogs.ubuntu.com/2016/05/05/#ubuntu-devel.txt

=== JanC is now known as Guest15498
=== JanC_ is now known as JanC
=== juliank_ is now known as juliank
=== Malsasa_ is now known as Malsasa
Logancjwatson: regarding https://launchpadlibrarian.net/257940143/buildlog_ubuntu-yakkety-amd64.eclipse-titan_5.4.1-1_BUILDING.txt.gz - it looks like the requirement of hardening-wrapper is due to the Ubuntu vendor module for dpkg that you wrote07:33
Logancjwatson: now that hardening flags are automatically exported by dpkg as of 1.16.1, it would probably be safe to remove the hardening-wrapper logic from https://anonscm.debian.org/cgit/dpkg/dpkg.git/tree/scripts/Dpkg/Vendor/Ubuntu.pm no? Lintian complains about hardening-wrapper being obsolete when I add a build dependency on it07:34
mwhudsonuh why is https://launchpad.net/ubuntu/+source/golang-pretty in main?08:27
mwhudsonboth in a "can someone tell me" way and "can someone tell me how to find out" way08:27
mwhudsonoh i think it's on a chain of build depends that ends up at juju-core or something08:28
zygabrendand: hey :)08:40
brendandzyga, hey08:40
zygabrendand: how are you doing? settling in in the new role?08:41
brendandzyga, getting to know maas08:42
brendandyeah08:42
zygacool, it's great to have you back :)08:42
Loganmwhudson: if you do an advanced search on bugs for the package and allow "Fix Released," you'll see Bug 126739308:43
ubottubug 1267393 in juju-mongodb (Ubuntu) "[MIR] juju-core, juju-mongodb, gccgo, golang" [High,Fix released] https://launchpad.net/bugs/126739308:43
Logan(there should always be an MIR bug associated with a package that was moved to main)08:44
mwhudsonah yes08:44
mwhudsoni guess an advanced search for "MIR" would be even more direct08:44
Logantrue :)08:44
mwhudsonoh well, i hope i didn't just break juju by syncing that package from debian :-)08:44
mwhudson(in fact it seems juju already ftbfs thanks to other things i did in debian...)08:45
Loganlooks like it has a depwait anyway08:45
Loganoh, golang-text needs to make it through the NEW queue first before it'll build08:47
Loganin any case, ideally there's an autopkgtest for juju that would fail if these new Go packages caused a regression08:48
cjwatsonLogan: while hardening-wrapper is obsolete, I'm reluctant to change that behaviour until it's actually gone.  In this case this is just a clear bug in eclipse-titan; it's using DEB_BUILD_OPTIONS but with the syntax and semantics documented for DEB_BUILD_MAINT_OPTIONS (see dpkg-buildflags(1))08:50
cjwatsonand DEB_BUILD_OPTIONS should not be set by packages anyway08:51
cjwatsonLogan: I'll fix this case and forward to Debian08:51
Logancjwatson: if DEB_BUILD_OPTIONS isn't supposed to be set by packages, and it doesn't have "hardening" by default, I don't see the point of maintaining that behavior (because when will it ever be set to that?)08:53
Loganthanks for the clarification, though! and I appreciate you fixing/forwarding the bug08:54
Loganactually I guess a user could just set it as an environment variable - is that the edge case you're supporting?08:54
cjwatsonLogan: The behaviour was specifically to help users setting that08:55
Logangotcha :)08:55
Loganhmm, maybe dpkg-buildflags(1) should be more explicit about the difference between DEB_BUILD_MAINT_OPTIONS and DEB_BUILD_OPTIONS08:56
Loganit just lists both of them as serving the same purpose but happens to use the MAINT one in the examples08:56
cjwatsonso I'm not totally disagreeing that we should eventually drop this behaviour from Dpkg::Vendor::Ubuntu08:56
LoganI guess it *should* be self-explanatory that the MAINT one is for package maintainers08:56
cjwatsonI'm just not quite sure about the timing, and it's not the right way to fix this particular bug08:56
Loganoh never mind about the documentation part, it's clarified later on in the manpage08:57
Loganbut yeah, agreed - the best way to fix this for now is to make sure the proper variable is being set in this package08:58
Loganthe logic in the dpkg module should probably only be removed once hardening-wrapper is killed off08:58
cjwatsonyeah08:58
Loganthanks again!08:59
Logancjwatson: btw, while I have you, nobody has merged my fix for the ubuntu-policy FTBFS yet (I wanted to make sure it was reflected in the repo instead of just uploading to universe): https://code.launchpad.net/~logan/debian-policy/fix-ftbfs/+merge/28985309:02
Loganand you seem like a likely candidate to review, even though you haven't touched it since 2009: http://bazaar.launchpad.net/~ubuntu-core-dev/debian-policy/ubuntu/changes09:02
cjwatsonyikes, I really must merge that thing properly09:03
Logan(although we should probably just get rid of this package because it's so obsolete)09:03
Logan(or that)09:03
cjwatsonwell it does actually have differences that are still applicable09:03
cjwatsonbut I'll merge your thing for now, thanks09:03
Logansure thing09:04
cjwatsonoh nice, ffmpeg sorted out, thanks doko09:08
cjwatson(and others I'm sure)09:08
rbasakapt bug? Bug 1571174.10:50
ubottubug 1571174 in squid3 (Ubuntu) "package squid3 failed to install/upgrade: dependency problems - leaving triggers unprocessed" [High,Triaged] https://launchpad.net/bugs/157117410:50
=== hikiko is now known as hikiko|off
pittiGood morning12:10
LocutusOfBorgdoko, I tried a lot with something like DEB_BUILD_OPTIONS=harderning=-pie but I don't see any -no-pie flag injected12:50
LocutusOfBorghow can I have that flag injected automatically?12:52
LocutusOfBorg"export DEB_BUILD_MAINT_OPTIONS=hardening=-pie,-pic,-format,-bindnow" <-- virtualbox now12:53
=== _salem is now known as salem_
rbasakLaney: thank you for spying on us :)13:33
rbasak(please continue)13:33
rbasakI'm struggling to follow how everything fits together.13:33
rbasakWhere are manually maintained packagesets kept?13:33
Laneyhi rbasak13:34
Laneythey are 'kept' in Launchpad13:34
rbasakThat makes sense. So they are manipulated only through launchpadlib, or is there a UI?13:34
Laneythere's a list here http://people.canonical.com/~ubuntu-archive/packagesets/yakkety/13:34
Laneytypically you use the edit-acl script in lp:ubuntu-archive-tools to manipulate them13:35
rbasakAh13:35
rbasakSomebody told me that already, but I've been looking in ~developer-membership-board only. Thanks :)13:35
Laneynod13:36
* Laney is here for advice if needed13:36
rbasakAppreciated :)13:36
rbasakDOes Launchpad keep a revision history of ACL changes?13:36
LaneyDon't think so13:37
rbasak(that would make touching this a little less scary)13:37
rbasakOK13:37
Laneyit's probably edit-acl -P ubuntu-mate -p flexiondotorg -S yakkety remove13:37
Laneyand then -p ubuntu-mate-uploaders ... add13:38
Laneyafter you make the team13:38
Laney(It might be that this a TB action - some things are, but I forget exactly what)13:38
rbasakI'd like to follow what's going on, but I think infinity made the initial change so it might be better for him to look at this particular change.13:39
rbasakJust to avoid any confusion.13:39
Laney'k13:39
flexiondotorgLaney, rbasak I think infinity add me as an uploader to the MATE packages in xenial and yakkety last week.13:39
rbasakflexiondotorg: Laney suggests doing it differently: https://lists.ubuntu.com/archives/devel-permissions/2016-May/000922.html13:39
rbasakflexiondotorg: it shouldn't make any difference to your actual access I believe.13:40
rbasakQuerying myself (or presumably any core dev) is interesting as it basically lists everything.13:47
xnoxpitti, cyphermox - http://packages.ubuntu.com/search?suite=yakkety&section=all&arch=any&keywords=ntp&searchon=contents14:23
xnoxthat's where original dhcp sets ntp server settings i am guessing14:23
xnoxvia /etc/dhcp/dhclient-exit-hooks.d/ntp14:23
xnoxi am shocked that ntp cannot be set in /etc/networking/interfaces =(14:24
pittibdmurray, arges: do you have some time to review my systemd SRU? we're getting more SRU requests, so would be nice to unblock this15:07
argespitti: i'll have time this afternoon if bdmurray doesn't get to it first15:08
bdmurraypitti: I'll look shortly15:21
seb128mdeslaur, hey, bug #1578576 seems a regression from the recent samba-regression-fix security update :-/15:27
ubottubug 1578576 in samba (Ubuntu) "ntlm_auth --helper-protocol=squid-2.5-ntlmssp report segfault" [Undecided,Confirmed] https://launchpad.net/bugs/157857615:27
mdeslauryeah, that's nice15:28
seb128mdeslaur, sorry :-( (just pointing it because I review recent reports and crossed it)15:28
mdeslaursamba regression whack-a-mole15:28
mdeslaurseb128: thanks15:29
seb128mdeslaur, yw!15:30
hallynpitti: hey, if/when you have a minute, http://paste.ubuntu.com/16241500/16:25
hallynpitti: ^ at line 195 is my check which does do a deb-system-helper disable libvirtd.service,16:25
hallynbut later on is debhelper-added code which appears to reenable it16:25
hallyn(starting line 264)16:26
hallyndo i need to do another step to tell it "no really, i want it disabled"?  is that what 'mask' would do?16:26
hallynmaybe i should just re-mask it16:32
hallynor i could manually move the libvirt-bin.masked dir to libvirtd.masked i suppose16:34
=== freyes_ is now known as freyes__
hallynpitti: ok but surely http://paste.ubuntu.com/16242185/ ought to work.16:55
hallynmaybe i need an update-state in there?17:00
hallynnope, doesn't help.  i'm at a loss.17:02
hallynsome food may help.17:02
=== freyes__ is now known as freyes
hallynmaybe i just need to find the right time to move /var/lib/systemd/deb-systemd-helper-enabled/libvirt-bin.service.dsh-also to /var/lib/systemd/deb-systemd-helper-enabled/libvirtd.service.dsh-also17:20
hallynreally there should be a dh_systemd command for that17:21
chilukcyphermox: https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/156956717:26
ubottuLaunchpad bug 1569567 in grub2 (Ubuntu) "Running update-grub does not update /boot/grub/grub.cfg with GRUB_CMDLINE_LINUX_DEFAULT from /etc/default/grub" [Undecided,Invalid]17:26
cyphermoxyay17:27
cyphermoxcjwatson: what are your thoughts on this? https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1569567/comments/617:41
ubottuLaunchpad bug 1569567 in grub2 (Ubuntu) "Running update-grub does not update /boot/grub/grub.cfg with GRUB_CMDLINE_LINUX_DEFAULT from /etc/default/grub" [Undecided,New]17:41
JordiGHsabdfl logged in 4 weeks ago last time. Was he around here?17:57
=== marcusto_ is now known as marcustomlinson
cyphermoxJordiGH: you're probably better off sending an email.18:05
JordiGHI don't want to contact him. Honestly just want to know if he comes around.18:06
dobeyyes, he is on irc plenty18:18
dobeyhe travels a lot more than most of us though18:18
JordiGHI see. Interesting.18:20
=== josepht` is now known as josepht
hallynmbiebl_: hey - are you around by chance?19:06
hallynmbiebl_: end of https://git.launchpad.net/~libvirt-maintainers/ubuntu/+source/libvirt/tree/debian/libvirt-daemon-system.postinst?h=2016-05-02/yakkety19:06
hallyni've added 'echo disabling >> /tmp/debug' to the postinst when it should disable it.  It is calling the deb-systemd-helper disable libvirtd.service19:07
hallynbut then the service is running anyway.19:07
hallynis there anything else i should do?  i don't understand the 'update-state' or 'mask' or 'unmask' options...  maybe onf of those, except i've tried them blindly with no apparent success19:07
cjwatsoncyphermox: haven't read the whole bug; I agree that 2) and 3) are likely good changes, at least; not quite sure why /etc/default/grub needs to be "[moved] to be early in /etc/default/grub.d" as I thought it was already always sourced first19:46
cyphermoxcjwatson: either or19:46
cyphermoxcjwatson: if we don't carry the values, then moving /etc/default/grub to grub.d will make it more obvious that there may be something to override what you've set19:47
cyphermoxI may have written that all wrong19:47
cjwatsoncyphermox: I'm very wary about moving configuration files around; not only is it inherently fiddly but there's a long tail of documentation and random web pages and such.  I think that should be avoided unless absolutely necessary19:48
cyphermoxI agree19:48
cyphermoxI would just fix curtin/maas/whatever to keep the values already sourced unless it's really not possible19:49
cyphermoxhrm, for some reason, I don't have the key repeat settings doing the right thing here.19:49
cjwatsonso yeah, if grub.d files are unconditionally clobbering settings when they could reasonably append, I consider that a clear bug in those packages19:50
hallynpitti: mbiebl_: so https://git.launchpad.net/~libvirt-maintainers/ubuntu/+source/libvirt/tree/debian/libvirt-daemon-system.postinst?h=2016-05-02/yakkety doesn't quite work but at least doesn't fail so i'll look at other failures until one of you can loan me a clue :)  thx20:01
argesbdmurray: i take it you reviewed systemd already?20:01
pittiarges: yes, he did20:23
argesah good20:24
=== hikiko is now known as hikiko|off
dokoLogan, please fokllow-up on the library transitions you just started21:46
Logansure, I was waiting for them to clear the NEW queue :)21:46
Logandoko: does the transition tracker need to be poked to update from the repo?22:19
dokoLogan, transition tracker is all manual :-/22:23
LoganI know22:23
LoganI added new ben files in the repo 20 minutes ago, but the transitions haven't shown up on the site22:23
Logan(and I'm pretty confident I formatted them correctly)22:24
=== salem_ is now known as _salem
clivejohi folks, what vervion of libc is yakkety using?22:43
sarnoldlooks like 2.23-0ubuntu3  https://launchpad.net/ubuntu/+source/glibc22:44
clivejolibc6 ?22:46
sarnoldyeah22:46
Loganit's had that SOVERSION for a long time22:46
sarnoldlibc5 is .. shrouded in mists. I for one can only vaguely recall those times.. dark times they were.22:47
sarnoldof course legends tell of libc4, grim and meagre.22:47
clivejowhat does error: ‘sqrt’ is not a member of ‘std’ mean?22:48
Logandoko: sorry to bug you again, but how exactly should dependency levels factor into which packages I transition first?22:48
Loganas far as I can tell, they all directly depend on the older version...22:49
dokofirst do level 1, wait until these are published, and then go on with the next level22:49
Loganwhat is the reasoning behind it, though? what do dependency levels mean?22:49
Loganclivejo: it means that you haven't included the necessary header for std::sqrt22:49
Logan(which appears to be <cmath>)22:50
sarnoldclivejo: feels like you forgot the #include <cmath> header: http://en.cppreference.com/w/cpp/numeric/math/sqrt22:50
sarnoldclivejo: .. but it might be something more subtle/annoying like a misused "use" statement or something.22:50
LoganI think glibc 2.23 removed a lot of implicit includes22:51
clivejoits KDE software and is building on other distro's, just cant get it to build on yakkety22:51
Loganlink? I can take a look22:51
clivejohttps://launchpadlibrarian.net/258045993/buildlog_ubuntu-yakkety-i386.plasma-desktop_4%3A5.6.3-0ubuntu1~ubuntu16.10~ppa3_BUILDING.txt.gz22:51
clivejosure22:51
clivejopackage is in https://launchpad.net/~kubuntu-ppa/+archive/ubuntu/staging-plasma22:52
Loganlooking22:53
Logandoko: pinging for above question (right after you replied to me)22:54
Loganclivejo: boom https://quickgit.kde.org/?p=plasma-desktop.git&a=commit&h=3a3bbc39d5cba8d77c89f6652c5b9c24c998049722:56
Logancherrypick that commit, and you should be golden22:57
clivejoI could kiss you Logan!!  Thanks so much!22:57
Loganhahaha no problem!22:57
dokoLogan, packages on higher levels may depend on packages in lower levels23:02
sarnoldLogan: boom :) nice find23:02
Logandoko: ahh, gotcha23:03
Logansarnold: thanks :)23:03
Logannow you all are required to endorse my imminent core dev application ;P23:03
sarnold\o/23:04
Unit193Logan: G'luck!23:04
clivejo+1 from me :)23:04
Loganthanks :) trying to time it properly23:04
LoganI've had a few people tell me I should apply23:05
Loganew, why does ubuntu-touch-meta depend on a shared library?23:42
LoganI shouldn't have to mess with germinate for a library transition23:44
Logangah sil2100 left just as I said that23:45
Logancould a core dev please merge this so that I can push an update to ubuntu-touch-meta? https://code.launchpad.net/~logan/ubuntu-seeds/ubuntu-touch-tinyxml2/+merge/29395523:54

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!