16:00:48 <jsmith> #startmeeting FESCO (2016-12-16)
16:00:48 <zodbot> Meeting started Fri Dec 16 16:00:48 2016 UTC.  The chair is jsmith. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:48 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:00:48 <zodbot> The meeting name has been set to 'fesco_(2016-12-16)'
16:00:48 <jsmith> #meetingname fesco
16:00:48 <zodbot> The meeting name has been set to 'fesco'
16:00:48 <jsmith> #chair maxamillion dgilmore jwb nirik paragan jsmith kalev sgallagh Rathann
16:00:48 <zodbot> Current chairs: Rathann dgilmore jsmith jwb kalev maxamillion nirik paragan sgallagh
16:00:48 <jsmith> #topic init process
16:01:03 <nirik> hello
16:01:13 <Rathann> hi
16:01:15 * nirik just realized that the pagure outage is poorly timed with the fesco meeting.
16:01:22 <paragan> .fas pnemade
16:01:23 <zodbot> paragan: pnemade 'Parag Nemade' <pnemade@redhat.com>
16:01:36 <jsmith> .hellomynameis jsmith
16:01:37 <zodbot> jsmith: jsmith 'Jared Smith' <jsmith.fedora@gmail.com>
16:01:47 * Rathann has to step away for a few minutes
16:02:08 <jsmith> No problem Rathann...
16:02:18 <jwb> hi
16:02:19 <jsmith> nirik: We'll do the best we can :-)
16:02:23 <sgallagh> .hello sgallagh
16:02:24 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
16:02:49 <jsmith> Yay, we have a quorum...
16:03:51 <jsmith> We'll wait until 5 after for anyone else to join, then get started
16:03:58 <jsmith> We should have a relatively light agenda today
16:05:45 <jsmith> OK, let's get started
16:06:00 <jsmith> #topic #1646 No appropriate sudo directory for user scripts
16:06:00 <jsmith> .fesco 1646
16:06:00 <jsmith> #link https://pagure.io/fesco/issue/1646
16:06:25 <jsmith> We postponed this one last week, as we felt we didn't have enough representation to make an informed decision.
16:06:34 <jsmith> I'm not sure that's really changed much this week :-/
16:07:08 <jsmith> Thoughts?
16:07:41 <sgallagh> I think this one probably needs to wait until we reconvene after the holidays.
16:07:47 <nirik> I guess I am fine either way...
16:08:05 <nirik> I don't think adding them really adds that much, but doesn't really hurt anything either.
16:08:07 <jwb> i have no opinion either way
16:08:12 <zodbot> jsmith: An error has occurred and has been logged. Check the logs for more information.
16:08:12 <nirik> but yeah, we could punt again
16:08:20 <jsmith> Proposal: Defer this decision until January...
16:08:27 <Rathann> ok I'm here
16:08:37 <jsmith> I'm fine with punting until January
16:08:42 <paragan> +1 to defer
16:09:32 <Rathann> I'm fine with deferring
16:09:40 <Rathann> I've already expressed my opinion, too
16:09:51 <Rathann> on #1646 that is
16:10:21 <jsmith> #agreed Defer a decision on ticket 1646 until January (+1: 5, +0:1, -1:0)
16:10:34 <jsmith> #topic #1654 F26 System Wide Change: Enable systemd-coredump by default
16:10:34 <jsmith> .fesco 1654
16:10:35 <jsmith> #link https://pagure.io/fesco/issue/1654
16:11:37 <nirik> someone added a comment about a bug with uid handling
16:11:43 <jsmith> #link https://fedoraproject.org/wiki/Changes/coredumpctl
16:11:47 <nirik> https://bugzilla.redhat.com/show_bug.cgi?id=1401772
16:11:54 <nirik> https://github.com/systemd/systemd/issues/3855
16:12:46 <zodbot> jsmith: An error has occurred and has been logged. Check the logs for more information.
16:14:01 <sgallagh> I haven't been following this one very closely. Have the coredumpctl people sorted out the interactions with ABRT yet?
16:14:35 <nirik> sgallagh: according to the change page, yeah... just disable the abrt thing and have it use systemd-coredump
16:15:21 <jsmith> nirik: It seems like the UID thing still hasn't been resolved -- systemd folks don't want to read /etc/logins.defs at runtime
16:15:35 <nirik> yeah.
16:15:55 <sgallagh> OK, then I'm +1 to moving ahead with this and trusting them to work out the bugs or enact the contingency plan
16:16:48 <nirik> yeah, I'm +1 to the change also. I am not sure how they can fix the uid thing without upstream systemd, but perhaps they can come up with something. If nothing else those people can revert it by enabling the abrt thing.
16:17:25 <jsmith> Agreed. Not 100% happy with the way things are now, but I'll give them the benefit of the doubt. +1 from me.
16:17:40 <jwb> +1
16:18:56 <jsmith> paragan, Rathann: Thoughts?
16:19:55 <paragan> Sorry I was reading the Change page, did not get enough time to read about this change before but I see I am fine with this change provided issues will get fixed soon, so +1
16:20:10 <Rathann> as far as I understand, coredumpctl doesn't work due to bugs at the moment
16:20:44 <jsmith> Rathann: Link to more details?
16:20:45 <Rathann> but I'm +1
16:20:58 <Rathann> jsmith: it's mentioned in the wiki page
16:21:03 <Rathann> https://bugzilla.redhat.com/show_bug.cgi?id=1341829
16:21:17 <Rathann> well, one bug to be exact
16:21:25 <Rathann> and the other issue is what nirik linked
16:21:27 <nirik> is that the selinux one?
16:21:33 <Rathann> which I would also like to see fixed
16:21:41 <Rathann> yes, it's the SELinux one
16:21:54 <nirik> yeah. It would have to be fixed. ;)
16:22:25 <jsmith> Definitely would need to be fixed :-)
16:23:09 <Rathann> doh, Michael's comment to disable selinux on workstation is not excellent :(
16:24:21 <jsmith> OK, looks like we're in agreement
16:24:50 <jsmith> #agreed Change is approved, as long as SELinux issue is fixed appropriately (+1:6,+0:0,-1:0)
16:25:06 <jsmith> #topic #1635 F26 Self Contained Changes
16:25:06 <jsmith> .fesco 1635
16:25:06 <jsmith> #link https://pagure.io/fesco/issue/1635
16:26:11 <jsmith> Since pagure is down at the moment, I'm assuming that the only outstanding change is:
16:26:11 <paragan> https://fedoraproject.org/wiki/Changes/AuthconfigCleanup
16:26:13 <nirik> looks like authconfig cleanup
16:26:19 <nirik> yeah. :)
16:26:19 <jsmith> #link https://fedoraproject.org/wiki/Changes/golang-buildmode-pie
16:26:33 <jsmith> Oh, maybe we have two
16:26:45 <jsmith> Let's tackle the authconfig one first
16:26:59 <nirik> +1 to autoconfig
16:27:00 <jsmith> #link https://fedoraproject.org/wiki/Changes/AuthconfigCleanup
16:27:08 <paragan> +1 to autoconfig
16:27:11 <Rathann> +1 to authconfig
16:27:18 <zodbot> jsmith: An error has occurred and has been logged. Check the logs for more information.
16:27:19 <jsmith> +1 to authconfig -- no brainer
16:27:19 <Rathann> (not aut_o_config)
16:28:39 <jsmith> jwb, sgallagh: Thoughts?
16:29:17 <sgallagh> +1 for authconfig, though I suspect it's going to be more complicated than described.
16:29:25 <sgallagh> /me suspects he'll end up involved there.
16:29:47 <jsmith> Aye... no doubt it'll be a fair amount of work, even if it's relatively low risk
16:30:52 <t8m> sgallagh, do you have something particular in mind?
16:31:16 <t8m> sgallagh, I am not going to do any kind of rewrite, just deleting stuff
16:31:25 <sgallagh> t8m: No, it's mostly just general concerns with how interdependent sssd, authconfig and realmd have become.
16:31:34 <t8m> sgallagh, there will be a new tool written by someone else
16:31:46 <sgallagh> Ah, I may have misunderstood the scope, then
16:31:50 <t8m> sgallagh, basically I do not want to touch anything related to sssd
16:32:06 <t8m> sgallagh, in the authconfig code - it will be kept
16:32:13 <sgallagh> OK
16:32:20 <sgallagh> Either way, I think this needs to happen.
16:32:31 <sgallagh> (And I already voted +1)
16:32:39 <jsmith> OK, sounds like we're agreed.
16:32:49 <t8m> Thank you
16:33:02 <jsmith> #agreed The authconfig self-contained change is approved (+1:5,+0:0,-1:0)
16:33:15 <jsmith> Now let's discuss the golang-buildmode-pie change
16:33:29 <jwb> jsmith: apologies, phone rang
16:33:29 <jsmith> I'm +1 for it, seems straightforward to me
16:33:33 * jwb reads scrollback
16:33:35 <jsmith> jwb: No problem :-)
16:33:41 <Rathann> I note there's nothing about the impact of this change to the performance of the compiled code
16:33:58 <sgallagh> Rathann: What compiled code are you talking about?
16:34:04 <jwb> i think authconfig changes are going to be complicated, but i agree it's a +1
16:34:35 <Rathann> sgallagh: go code compiled into binaries
16:34:47 <Rathann> what else?
16:35:04 <sgallagh> Rathann: Sorry, I missed the change in topic.
16:35:10 <sgallagh> I thought you were still discussing authconfig
16:36:12 <Rathann> PIE requires an additional register at least on i686
16:36:46 <Rathann> and there's some overhead on other architectures as well
16:36:47 <jwb> i686 is literally becoming an alternative arch probably at the end of this meeting, and the combination of that + go is small
16:38:14 <jwb> the will naturally be overhead involved here.  both for PIE and the hardened flags.  are you concerned the performance impact is going to be more than when we did the same for gcc built code?
16:39:01 <jsmith> I think the security gain is more than worth the performance penalty here, for the record
16:39:14 <Rathann> well I don't have any numbers
16:39:25 <Rathann> but the change page totally ignores this
16:39:54 * nirik notes pagure is back up now.
16:39:57 <paragan> this change is yet in "ChangeAnnounced" Category
16:40:12 <jsmith> Now that paguire is back up, I see this issue isn't on the official list of things to discuss...
16:40:17 <jsmith> Shall we punt until the next meeting?
16:40:35 * jsmith was going off of the emails sent to the devel list while Pagure was down for maintenance
16:40:54 <jwb> if we're discussing this early, that gives Rathann the opportunity to raise the question on the list
16:42:33 * nirik nods
16:42:55 <jsmith> OK, let's punt then.
16:43:32 <jsmith> #agreed Wait until next meeting to vote on golang-buildmode-pie issue, so that Rathann can solicit more feedback on the mailing list.
16:43:47 <jsmith> #topic Chair for 06 Jan meeting
16:44:16 <jwb> i can chair
16:44:20 <jsmith> In last week's meeting, we decided to cancel the next two meetings due to the end of year, and pick up again in January
16:44:33 <jsmith> #info jwb to chair January 6th meeting
16:44:43 <jsmith> #topic Open Floor
16:44:59 <jwb> i'd like to bring up the announcement email on i686 i created in the ticket
16:45:11 * nirik was +1 in the ticket to the announcement.
16:45:14 <jwb> i've gotten a few +1s but we don't have consensus
16:45:29 <jsmith> I added my +1 today -- does that push us over?
16:45:49 <jwb> i think we need to get this out soon, so i plan on sending the email to devel-announce after we get consensus
16:45:56 <jwb> jsmith: let me lok
16:46:10 <jsmith> jwb: That puts us at +5, by my count...
16:46:13 <sgallagh> I think I also had a +1 in the ticket, but if not, here it is
16:46:36 <jsmith> jwb: That's +5, not counting your implicit +1
16:46:47 <jwb> jsmith: you are right.  though i have no idea why there are a bunch of duplicate +1s
16:47:08 <Rathann> mine got duplicated too
16:47:11 <jsmith> Me either :-(
16:47:25 <jsmith> Maybe that was addressed in the Pagure maintenance?
16:47:34 <Rathann> there was an error saying that I can't close this ticket because there are dependent tickets open
16:47:43 <Rathann> when I submitted my +1
16:47:46 <jwb> i'll give people the rest of today in case dgilmore or others want to chime in
16:47:48 <paragan> https://pagure.io/pagure/issue/1671
16:47:55 <jsmith> I got the same error, Rathann
16:47:59 <Rathann> :)
16:48:15 <Rathann> ok I need to drop off now, sorry
16:48:19 <jsmith> #info Reminder to add yourself to the Fedora Vacations calendar if you plan to be away
16:48:21 <Rathann> thanks for the meeting
16:49:42 <jsmith> Anything else for the open floor?
16:49:58 * jsmith lights the fuse....
16:51:18 <jsmith> #endmeeting