16:01:24 <jsmith> #startmeeting FESCO (2017-07-28)
16:01:24 <zodbot> Meeting started Fri Jul 28 16:01:24 2017 UTC.  The chair is jsmith. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:24 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:01:24 <zodbot> The meeting name has been set to 'fesco_(2017-07-28)'
16:01:25 <jsmith> #meetingname fesco
16:01:25 <zodbot> The meeting name has been set to 'fesco'
16:01:25 <jsmith> #chair maxamillion dgilmore jwb nirik jforbes jsmith kalev sgallagh Rathann
16:01:25 <zodbot> Current chairs: Rathann dgilmore jforbes jsmith jwb kalev maxamillion nirik sgallagh
16:01:25 <jsmith> #topic init process
16:01:31 <jsmith> Hello everyone...
16:01:43 <sgallagh> .hello sgallagh
16:01:45 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
16:01:49 <jsmith> .hello jsmith
16:01:51 <zodbot> jsmith: jsmith 'Jared Smith' <jsmith.fedora@gmail.com>
16:02:02 <nirik> morning
16:02:02 <sgallagh> Apologies for my spotty attendance the last few meetings
16:03:18 <jsmith> I really hope we have a quorum today, because there's a lot of tickets to go through
16:03:28 <jsmith> sgallagh: No worries -- summers are tough for most of us
16:04:27 <dgilmore> hola
16:04:40 <jsmith> OK, that's four of us....
16:05:03 <dgilmore> :( necesito cinco
16:05:24 <sgallagh> You can count my votes double, if you prefer ;-)
16:05:34 <dgilmore> id prefer not to
16:05:46 <sgallagh> (That was a joke)
16:09:06 <dgilmore> I know
16:09:07 <dgilmore> :)
16:10:16 <sgallagh> On a FESCo-related topic: nominations for seats ends today.
16:10:26 <dgilmore> indeed
16:10:37 <sgallagh> Does anyone know if jwb was planning to run? I think he's been on PTO for the whole nomination period.
16:11:11 * nirik notes there's 4 nominations for 4 seats currently. ;)
16:11:59 <dgilmore> sgallagh: I have no idea
16:12:17 * dgilmore did nominate again but almost didn't
16:12:36 <sgallagh> Yeah, I was the same. I considered taking a break.
16:12:41 <Rathann> Hi
16:12:44 <sgallagh> But the lack of nominees was disheartening
16:12:55 <Rathann> .hello rathann
16:13:00 <dgilmore> we have 5
16:13:05 <zodbot> Rathann: rathann 'Dominik Mierzejewski' <dominik@greysector.net>
16:13:14 <jsmith> Alrighty.... let's get started then
16:13:23 <sgallagh> Rock on
16:13:28 <jsmith> #topic Follow-ups
16:13:28 <jsmith> #topic #1736 - Don't automatically close security bugs on Fedora EOL
16:13:28 <jsmith> .fesco 1736
16:13:28 <jsmith> #link https://pagure.io/fesco/issue/1736
16:13:30 <zodbot> jsmith: Issue #1736: Don't automatically close security bugs on Fedora EOL - fesco - Pagure - https://pagure.io/fesco/issue/1736
16:14:26 <sgallagh> So, I don't think a perfect situation exists, but we can certainly improve the current state by simply increasing the version of any bug with the Security flag set.
16:14:44 <jsmith> As I understand it, we were waiting for feedback from the FPM and see what work needs to be done to the scripts
16:14:49 <dgilmore> there really is not a good situation here
16:14:55 <jsmith> Agreed
16:15:54 <sgallagh> dgilmore: Can you elaborate?
16:16:09 <sgallagh> I think the proposal to at least not close bugs tagged by Security makes sense.
16:16:24 <nirik> I noticed something while working on ImageMagick yesterday...
16:16:40 <dgilmore> sgallagh: well closing sucks. but leaving open they can not be fixed anyway, they could be moved to a newer release if appropriate
16:16:45 <nirik> the actual cve bugs get closed, but the security response ones don't... so you can find the others via that
16:17:57 <nirik> hum, or perhaps someone manually reopened them
16:18:00 <nirik> nevermind
16:18:02 <sgallagh> dgilmore: Perhaps then we just add a needinfo? flag on security bugs, so at least the owners get nagged about it between the announcement and the closing?
16:18:17 <dgilmore> sgallagh: perhaps
16:18:27 <jsmith> sgallagh: I like that idea
16:18:36 * sgallagh is in favor of improving the situation even if we don't solve it completely.
16:18:54 <dgilmore> given that we can not build or ship anything once a release is EOL, it needs evaluation for impact on other releases
16:19:06 <dgilmore> sgallagh: sure
16:19:19 <jsmith> What do you think -- do we feel comfortable voting, or do we want to kick this one down the road another week?
16:19:25 <nirik> so, perhaps at EOL move to current release and add a note?
16:20:05 <sgallagh> jsmith: It seems like we have several possibilities to discuss, so maybe we enumerate them in the ticket and revisit next week?
16:20:05 <nirik> "This bug was due to EOL with fedora N, but since it's a security bug it's being kept open, please evaluate if it affects Fedora N+1"
16:20:24 <nirik> we might also discuss on list, might be more ideas there
16:20:31 <jsmith> sgallagh: OK, I'm fine with that.
16:20:42 <sgallagh> nirik:  plus needinfo: pkg-owner@fp.o ?
16:20:44 <dgilmore> nirik: I think that is preferable to just leaving them open
16:21:09 <nirik> well, sure, but I doubt that will do much
16:21:14 <dgilmore> same
16:21:15 <jsmith> Proposal: Enumerate various ideas in the ticket and on the list, and revisit next week.
16:21:21 <nirik> +1
16:21:24 <jsmith> +1
16:21:31 <dgilmore> +1
16:21:38 <sgallagh> nirik: Well, in some cases the primary maintainer might be AWOL but someone else might notice if that happened.
16:21:47 <sgallagh> +1
16:21:59 <dgilmore> I need to run to the car dealership in 30 minutes, will be offline for about 10-15 mins
16:22:01 <nirik> but they have had a long time to notice no?
16:22:13 <dgilmore> indeed they have
16:22:38 <sgallagh> Depends on if they have watchbz set, maybe
16:22:43 <Rathann> +1
16:23:25 <jsmith> #agreed #1736 Enumerate various ideas i nthe ticket and on the list, and revist next week (+1:5, +0:0, -1:0)
16:23:41 <jsmith> #topic #1737 - Proposal: i686 SIG needs to be functional by F27 release date
16:23:41 <jsmith> .fesco 1737
16:23:41 <jsmith> #link https://pagure.io/fesco/issue/1737
16:23:42 <zodbot> jsmith: Issue #1737: Proposal: i686 SIG needs to be functional by F27 release date or we drop i686 kernel from F28 - fesco - Pagure - https://pagure.io/fesco/issue/1737
16:24:33 <jsmith> Other than a comment from dgilmore, I haven't seen any more discussion about what might make the SIG "functional".
16:24:35 <nirik> I still think the one measueable thing is that they fix all blockers on that arch and so there's a release.
16:24:35 <jsmith> Thoughts?
16:25:06 <jsmith> nirik: That's fair...
16:25:07 <sgallagh> nirik: Except i686 isn't blocking
16:25:16 <sgallagh> So that's trivially easy to accomplish :)
16:25:17 <nirik> right...
16:25:31 * nirik rephrases, but I am sure you know what I meant.
16:25:38 <jsmith> I was just going to say ''then what's your definition of blocking"
16:25:59 <dgilmore> nirik: indeed. best thing is people stepping in and getting things done
16:26:00 <sgallagh> nirik: I assume you meant "things that would block the release if i686 was a blocking arch"
16:27:06 <sgallagh> Of course, that's still a little slushy, because there are conditional blockers, etc.
16:27:08 <nirik> "The i686 sig will test against the fedora release critera, file bugs and address them as if it was a blocking release"
16:27:23 <dgilmore> without a i686 kernel we can not realistically deliver anything i686 except an Everything repo and updates respos
16:27:25 <nirik> yeah... but its a place to start
16:27:44 <dgilmore> the value of which is limited
16:27:49 <nirik> and it's easy to see it failed if they don't fix the anaconda bug thats been there for months
16:27:58 <sgallagh> nirik: Right, I'm just worried this will end up on the blocker crew plates to figure out if they're meeting their criteria
16:28:39 <sgallagh> dgilmore: Well, theoretically we could do container base images without a kernel as well
16:28:48 <dgilmore> sgallagh: we can not
16:28:49 <sgallagh> s/base//
16:28:56 <nirik> yeah, we don't want more work there...
16:28:58 <dgilmore> sgallagh: we can not build them without a 32 bit kernel
16:29:18 <dgilmore> they run anaconda in a vm
16:29:34 <nirik> we could ask the sig to report monthly to fesco and just have fesco decide if they are functional and what would be needed to be that?
16:29:41 <sgallagh> dgilmore: I said "theoretically" on the grounds that tooling could exist without requiring a kernel.
16:29:44 <sgallagh> Not that it currently does.
16:29:53 <dgilmore> last I tested a system with 64 bit kernel and 32 bit userspace it did not work well
16:30:12 <dgilmore> anyway enough noise by me
16:30:49 <sgallagh> Honestly, I'm trying to just be helpful, but I've been a fan of killing off i686 since 2010, so I'm not going to fight particularly hard :)
16:31:46 <dgilmore> sgallagh: sure, I think there is still value in 32 bit x86 for many use cases
16:32:14 <dgilmore> including 32 bit vms for memory constrained use cases
16:32:56 <dgilmore> I know there is new iot hardware out there thats 32 bit only
16:32:56 <jsmith> Does someone want to make a proposal?
16:32:57 * Rathann still has an old Atom-based netbook, which won't run x86_64 code
16:32:59 <sgallagh> dgilmore: To be clear, I don't think 32-bit as a concept is a bad thing. I just don't see Fedora as a project putting any real effort into it.
16:33:15 <dgilmore> sgallagh: I disagree
16:33:32 <dgilmore> well I agree we are not, but I think we should
16:33:44 <dgilmore> but it may be that we limit the use cases supported
16:33:52 <sgallagh> That was going to be the next thing I said.
16:34:07 <sgallagh> If IoT is a good use-case, maybe we focus on that.
16:34:12 * nirik tossed out 2 ideas, but not sure either is really ready to be voted on.
16:35:04 <dgilmore> nirik: I think it needs a bit more time to solidify a bit
16:35:13 <jsmith> Proposal: Keep discussing in the ticket and on the list, and revisit next week
16:35:20 <nirik> jsmith: +1
16:36:06 <sgallagh> Counter-proposal: Things proceed as they are and we let the kernel maintainers tell us after F27 release whether they feel a sufficient force has stepped up.
16:36:38 <dgilmore> +1
16:36:50 <sgallagh> I mean, at the end of the day this is a resourcing issue. So I think we should just trust that the people who know best will tell us if they can or cannot manage.
16:36:51 <dgilmore> +1 to jsmith
16:37:14 <dgilmore> sgallagh: I think FESCo needs to be more actively involved
16:37:26 <nirik> sgallagh: an interesting idea... but it could mean kernel is ok but the rest is all busted...
16:37:47 <sgallagh> dgilmore: Well, FESCo taking any position here amounts to an endorsement (or rejection) of the arch.
16:37:56 <Rathann> Sorry, I have to drop off now. Will vote in the tickets later.
16:38:00 <sgallagh> If we say "go build this SIG with these rules", we're saying i686 is important.
16:38:19 <sgallagh> ... and we're below quorum again
16:38:31 <jsmith> Yup..
16:38:46 <sgallagh> OK, I guess we defer to next week by way of lack of quorum :-/
16:39:06 <jsmith> #agreed #1737 Defer to next week due to lack of quorum.
16:39:25 <jsmith> And, since we have a lack of quorum, I'm not sure there's much use in going through the other tickets....
16:39:29 <dgilmore> sgallagh: not true
16:39:45 <jsmith> Thoughts/
16:39:59 <jsmith> ?
16:40:30 <dgilmore> sgallagh: FESCo being involved just means FESCo is monitoring things and does not need to get status as it is engaged and helping those navigate things that want the work done
16:40:39 <dgilmore> jsmith: ack
16:40:53 <nirik> :(
16:41:29 <jsmith> #topic Next week's chair
16:41:41 <sgallagh> dgilmore: Fair enough
16:41:43 <dgilmore> I can
16:41:44 <jsmith> I'm happy to volunteer again next week...
16:41:55 <dgilmore> unless jsmith really wants to
16:42:07 <jsmith> dgilmore: Well, I didn't get to do much this week.
16:42:16 <jsmith> Why don't I take next week as well, and you can have the week after that.
16:42:29 <jsmith> Sound fair?
16:43:00 <dgilmore> sounds good
16:43:04 <jsmith> #topic Election reminder
16:43:04 <sgallagh> I'll actually take the 11th, as I'm going to be away the 18th
16:43:09 <dgilmore> gracias amigo
16:43:26 <sgallagh> And I owe a chair due to a sudden conflict last week
16:43:39 <jsmith> Just a reminder of the upcoming elections -- do you part -- nominate yourself or a friend, and participate in the elections.
16:43:57 <dgilmore> #info 16:43 < jsmith> Just a reminder of the upcoming elections -- do you part -- nominate yourself or a friend, and participate in the elections.
16:44:15 <jsmith> #topic Open Floor
16:45:35 <jsmith> I'll set the countdown for a couple of minutes -- if there are no items for the open floor, we'll adjourn until next week.
16:48:07 <dgilmore> I have nothing
16:48:15 <dgilmore> actually I do
16:48:38 <dgilmore> https://pagure.io/releng/issue/6898#comment-450892
16:49:13 <jsmith> How hard is it to do another mass rebuild just for one arch?
16:49:17 <jsmith> Just curious...
16:49:23 <dgilmore> the tools guys submitted a bug fix to binutils, and want us to rebuild everything binary
16:49:32 <dgilmore> jsmith: it is impossible
16:49:37 <jsmith> Or logistically, does it mean doing it for all arches?
16:49:41 <dgilmore> nvrs are linked for all arches
16:49:42 <jsmith> That's what I was afraid of...
16:49:54 <jsmith> Do we really have a choice?
16:50:11 <dgilmore> it should be under half of the packages
16:50:19 <dgilmore> anything noarch will not need rebuilt
16:50:25 <jsmith> Right...
16:50:25 <dgilmore> we do not
16:50:40 <jsmith> Then I guess it is what it is, and we try to get it done ASAP
16:50:49 <jsmith> and hope it doesn't affect the schedule too much
16:51:51 * nirik nods
16:53:14 <dgilmore> yep
16:53:46 <jsmith> OK, anything else for the open floor?
16:53:53 <jsmith> If not, I'll end the meeting in one more minute
16:53:55 <dgilmore> nope
16:58:20 <jsmith> #endmeeting