18:06:02 <nirik> #startmeeting FESCO (2015-08-26)
18:06:02 <zodbot> Meeting started Wed Aug 26 18:06:02 2015 UTC.  The chair is nirik. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:06:02 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:06:03 <nirik> #meetingname fesco
18:06:03 <nirik> #chair ajax dgilmore hguemar jwb nirik paragan rishi thozza sgallagh
18:06:03 <nirik> #topic init process
18:06:03 <zodbot> The meeting name has been set to 'fesco'
18:06:03 <zodbot> Current chairs: ajax dgilmore hguemar jwb nirik paragan rishi sgallagh thozza
18:06:07 <paragan> Hi
18:06:09 <jwb> hello
18:06:13 <nirik> morning
18:06:15 <thozza> hi :)
18:06:26 <sgallagh> .hello sgallagh
18:06:30 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
18:06:51 <dgilmore> hola
18:07:03 <thozza> nirik: I always want to say "evening" when you say that, but I always forget :)
18:07:27 <debarshi> Thanks nirik
18:07:35 <nirik> https://freenode.net/faq.shtml#fst :)
18:09:14 <number80> .hello hguemar
18:09:16 <zodbot> number80: hguemar 'Haïkel Guémar' <karlthered@gmail.com>
18:10:00 <nirik> debarshi: you able to run things? or if your net is too bad, I can...
18:10:09 <debarshi> nirik: It is better if you continue.
18:10:13 <nirik> ok
18:10:14 <debarshi> I have a huge lag.
18:10:19 * nirik looks for the agenda
18:10:34 <nirik> #topic #1444 updates deliverables
18:10:34 <nirik> .fesco 1444
18:10:34 <nirik> https://fedorahosted.org/fesco/ticket/1444
18:10:36 <zodbot> nirik: #1444 (updates deliverables) – FESCo - https://fedorahosted.org/fesco/ticket/1444
18:11:33 <thozza> BTW jkurik is on vacation this week
18:11:35 <nirik> not sure where we are with this. there was a wiki page that was being worked on
18:12:02 <debarshi> thozza: Oh. I assumed that he would be here, so didn't explicitly invite him. :(
18:12:35 <nirik> I can't seem to even find the wiki page...
18:12:35 * debarshi looks at adamw
18:12:41 <adamw> hi.
18:12:47 <nirik> jreznik was working on it, but he's not around right now either.
18:13:50 <nirik> I guess we defer and try and have something before beta?
18:14:17 <debarshi> I think everyone who commented on the ticket should be here - roshi , jwb , dgilmore , adamw
18:14:30 <thozza> nirik: jkurik should be here next week
18:15:05 <nirik> right, but not sure if it jkurik was going to work on it now, or if jreznik was still going to finish it, or what
18:15:28 <number80> should be jkurik though
18:15:39 <dgilmore> I would rather the release deliverables sorted before we do this
18:15:41 <thozza> I can ask him next week
18:15:43 <dgilmore> but we need this also
18:15:58 <nirik> oh, I was thinking of the wrong ticket here.
18:16:18 <nirik> ok, proposal: defer a week and revisit
18:16:28 <debarshi> +1
18:16:36 <thozza> nirik: +1
18:16:45 <paragan> nirik, +1
18:17:05 <sgallagh> +1
18:17:11 <nirik> (and yeah, after release deliverables this might be easier to do)
18:17:38 <jwb> +1
18:17:46 <nirik> #agreed will defer a week and revisit (+6,0,0)
18:17:54 <nirik> #topic #1466 non-responsive maintainer exception process for skottler
18:17:54 <nirik> .fesco 1466
18:17:55 <nirik> https://fedorahosted.org/fesco/ticket/1466
18:17:56 <zodbot> nirik: #1466 (non-responsive maintainer exception process for skottler) – FESCo - https://fedorahosted.org/fesco/ticket/1466
18:18:21 <nirik> any news here? there was one issue with the nagios-plugins epel builds. ;( I will try and go push a fixed update for that soon.
18:18:47 <jwb> i pinged sam earlier today in #fedora-kernel to see if he was going to join the meeting
18:19:09 <jwb> spoke with him about this briefly last week.  his plan was to orphan most of his packages, particularly the erlang ones
18:19:14 <number80> someone told me that it broke EL6 though
18:20:37 <nirik> number80: there was a plugin in the epel6 nagios-plugins that was dropped in 2013... and so wasn't in this update. I can push a new update with obsoletes... but really you just need to remove that plugin. It's not any kind of big breakage.
18:21:34 <number80> nirik: I guess having the obsoletes would be fine but I was just reporting that :)
18:21:53 <nirik> number80: yeah, sadly it wasn't reported to me until after it went stable... but oh well.
18:22:10 <nirik> so, what do we want to do here?
18:22:18 <nirik> wait some more? go ahead with orphaning?
18:22:54 <number80> AFAIK, if he will orphan them, I don't mind waiting
18:22:54 <jwb> i can see either way i guess.
18:23:09 <nirik> if there's no urgent security issues I don't mind waiting
18:23:23 <debarshi> Will he also orphan nagios-plugins? That was the bone of contention.
18:23:32 <number80> erlang mass orphan is likely to impact rabbitmq and other large packages
18:23:53 <jwb> number80, he mentioned most of them are already co-maintained by the erlang sig or something
18:24:13 <number80> good to know
18:25:35 <nirik> so, revisit next week? orphan next week? something else?
18:25:52 <jwb> i don't want to keep revisiting this every week
18:25:58 <thozza> nirik: what will possibly change until next week?
18:26:06 * debarshi wants to resolve this sooner rather than later
18:26:11 <nirik> he could orphan things?
18:26:12 <jwb> so either we do nothing and close it as wontfix or we orphan i guess
18:26:13 <paragan> I think orphan this week only
18:26:22 <sgallagh> Proposal: If no status update by monday 1200 UTC, orphan.
18:26:32 <debarshi> sgallagh: +1
18:26:36 <thozza> sgallagh: +1
18:27:06 <paragan> sgallagh, +1
18:27:28 <nirik> ok, sure. +1
18:27:41 <sgallagh> +1 for the record
18:27:41 <thozza> sgallagh: the time definition is not specific enough though
18:28:00 <sgallagh> thozza: Monday, Aug. 31
18:28:13 <nirik> and you mean orphan nagios-plugins? or all his packages?
18:28:14 <number80> +!
18:28:25 <jwb> i was assuming all
18:28:25 <thozza> ohh, I missed "monday" in there.... sorry
18:28:28 <sgallagh> nirik: All of his packages
18:28:48 <sgallagh> nirik: He's had plenty of time to speak in his own defense.
18:29:18 <jwb> he did point out that his email is broken and that nobody even tried contacting him on irc
18:29:20 <sgallagh> If he wants to reclaim some of them, he can ask the remaining maintainers
18:29:26 <jwb> but yeah
18:30:00 <jwb> sgallagh's proposal passed
18:30:43 <nirik> #agreed if no status update from maintainer by 2015-08-31 12UTC, all their packages will be orphaned (+6,0,0)
18:30:49 <nirik> ok with everyone?
18:30:52 <number80> he has time to do orphan the packages he won't maintain so it should be ok
18:31:12 <thozza> nirik: yes
18:31:16 <nirik> #topic #1467 Progress at Change Checkpoint: Completion deadline (testable)
18:31:16 <nirik> .fesco 1467
18:31:16 <nirik> https://fedorahosted.org/fesco/ticket/1467
18:31:17 <zodbot> nirik: #1467 (F23 Changes - Progress at Change Checkpoint: Completion deadline (testable)) – FESCo - https://fedorahosted.org/fesco/ticket/1467
18:32:11 <nirik> so, I can speak to the passphrase policy. It's still waiting on a bit of implementation in anaconda... otherwise the changes are in. Also, overrides need to be added if any products want to.
18:32:54 <sgallagh> nirik: Has any progress been made on the "confirm weak" issue?
18:33:10 <nirik> sgallagh: I haven't seen anything since your last update in the bug.
18:33:13 <sgallagh> The last message I saw on that was... mine
18:33:22 <nirik> thats the last part that needs dealing with I think.
18:33:25 <sgallagh> ok, I'll ping davidshea about it
18:33:45 <nirik> thanks
18:34:10 <nirik> so, do we want to do anything here with these?
18:34:38 <nirik> I know max and dgilmore have been working on 2 week atomic and layered build... hopefully done before beta
18:35:01 <number80> should be
18:35:05 <paragan> so revisit this ticket after beta release and then close?
18:35:16 <number80> *nods*
18:35:17 <nirik> the dnf thing is ready for review, we should get someone to do so...
18:35:25 <sgallagh> Well, Beta Freeze is the official time to invoke the contingency plans
18:35:33 * debarshi looks at the review
18:36:36 <debarshi> Ok, I volunteer to do the review tomorrow.
18:37:00 <sgallagh> FWIW, I've tested the process personally using the COPR builds
18:37:01 <debarshi> Haven't done a package review in a while, so not a bad time to do one.
18:37:09 <sgallagh> It works quite well
18:37:16 <nirik> cool. ;)
18:37:32 <nirik> #info will follow these last changes closely and visit next week
18:37:34 <sgallagh> (Even tolerating some... peculiarities in my installed package set that fedup would not have)
18:37:40 <number80> debarshi: you're probably a lazy sponsor then :)
18:37:47 <nirik> #topic #1469 i686 as a non-blocking architecture
18:37:47 <nirik> .fesco 1469
18:37:48 <nirik> https://fedorahosted.org/fesco/ticket/1469
18:37:49 <zodbot> nirik: #1469 (i686 as a non-blocking architecture) – FESCo - https://fedorahosted.org/fesco/ticket/1469
18:38:21 <nirik> jwb: have any intro/info to add here?
18:38:47 <jwb> i don't believe i have any additional information
18:38:55 <jwb> the product WGs all don't care about i686 iamges
18:38:57 <number80> Cloud WG would support this
18:39:14 <number80> (proposal, not the i686 images)
18:39:52 <nirik> so, how will we know if image work enough to ship?
18:39:53 <sgallagh> jwb: Well, the Workstation group would continue to block on i686 if-and-only-if the i686 kernel was equivalently maintained.
18:40:04 <jwb> it hasn't been since feb
18:40:08 <jwb> so they don't care.
18:40:22 <sgallagh> But since that doesn't look like reality, that condenses to "don't ship i686 media"
18:40:39 <jwb> nirik, it's looking more like we just won't ship it
18:40:41 <jwb> period
18:40:54 <jwb> at least for the Editions
18:41:05 <debarshi> I guess the only sticking point for Workstation in terms of new hardware are those Intel tablets that Bastien was talking about.
18:41:06 <jwb> if someone wants to spit out a netinstall iso and test it, fine i guess
18:41:06 <sgallagh> jwb: Well, there's been some ambiguity about a nebulous "generic netinstall" retaining support for installing on 32-bit
18:41:07 <nirik> I'm not opposed, but it seems a bit abrupt for f23... since we shipped a alpha with it already.
18:41:08 <debarshi> But I am not sure whether the WG is officially interested in those.
18:41:12 <sgallagh> But I haven't seen anyone volunteer to create that
18:41:38 <jwb> debarshi, see my last comment in the ticket.  i like bastien well enough, but his interest in the tablets is not something the entire project should be beholden to
18:42:10 <jwb> nirik, we can target f24.  i can be patient
18:42:11 * number80 is his neighbour and he's not nice
18:42:23 <jwb> number80, taht wasn't necessary at all.
18:42:34 <sgallagh> number80: Off-topic and unnecessary
18:42:47 <number80> well, from my PoV, nobody wants to support the i686 image
18:42:50 <nirik> I'm completely +1 for dropping all the i686 images/media in f24. ;)
18:43:26 <number80> (my apologies btw)
18:43:32 <sgallagh> Proposal: Fedora will ship no i686/32-bit x86 install media in Fedora 24
18:43:38 <number80> +!
18:43:42 <number80> +1
18:43:44 <dgilmore> kinda related here is the idea of changing the definition of what is primary and secondary to where we ship the output, in this case we could have i686 be a secondary arch, even if only for some subsets of Fedora
18:43:54 <dgilmore> sgallagh: -1
18:44:15 <debarshi> jwb: I understand the last part of your argument.
18:44:32 <nirik> dgilmore: care to expand on that -1?
18:44:41 <dgilmore> nirik: the line above it
18:44:50 <nirik> ok, so:
18:45:02 <dgilmore> make 32 bit deliverables be secondary
18:45:19 <dgilmore> put them in /pub/alt or /pub/secondary
18:45:29 <number80> do we have people willing to maintain this secondary arch?
18:45:31 <sgallagh> Ok, perhaps that was poorly phrased.
18:45:31 <jwb> dgilmore, there is nothing in sgallagh's proposal that prevents that
18:45:44 <dgilmore> jwb: there kinda is
18:45:50 <sgallagh> Proposal: Fedora will ship no i686/32-bit x86 install media in Fedora 24 as a primary/blocking deliverable.
18:46:11 <dgilmore> sgallagh: that i can vote +1 on
18:46:13 <debarshi> dgilmore: We can't make i686 truly secondary because of skype.
18:46:21 <nirik> ?
18:46:24 <nirik> +1 to the proposal
18:46:32 <dgilmore> debarshi: I am not saying that it should entirely be secondary
18:46:42 <paragan> +1 to sgallagh proposal
18:46:43 <dgilmore> that some of the deliverables be secondary
18:46:46 <number80> debarshi: you mean we have to keep multilib support
18:46:47 <sgallagh> debarshi: We are sort of talking about redefining what "secondary" means
18:46:54 <jwb> +1
18:46:58 <debarshi> number80: yes
18:47:01 <debarshi> sgallagh: right
18:47:03 <sgallagh> Not that it's built like ppc today or anything
18:47:04 <number80> debarshi: off course
18:47:22 <sgallagh> Multilib alone means it has to still be built in the primary Koji instance
18:47:32 <thozza> sgallagh: +1
18:47:35 <nirik> personally, I am not sure I like the idea that microsoft dictates what we do... but perhaps it's just me.
18:47:38 <number80> sgallagh: still +1
18:47:45 <dgilmore> relaistically x86 is the only multilib arch now
18:47:58 <jwb> nirik, s/skype/third party applications using 32-bit libs
18:48:00 <number80> nirik: well, not limited to just microsoft
18:48:00 <dgilmore> I would like to see us look at ways to drop multilib
18:48:10 <dgilmore> but that is a down the road discussion
18:48:15 <jwb> dgilmore, that would require something like app containers
18:48:18 <nirik> ok, personally, I am not sure I like the idea that a closed source 3rd part dictates what we do... ;)
18:48:22 <dgilmore> jwb: right
18:48:23 <sgallagh> Yeah, unfortunately the world that allows that is still a ways off
18:48:29 <dgilmore> jwb: way out of scope for now
18:48:37 <nirik> right, sorry, sidetracking.
18:48:52 <nirik> so, thats +6 to the proposal?
18:48:55 <jwb> yes
18:49:02 <sgallagh> I'm +1 if you didn't account for it
18:49:15 <nirik> #approved Fedora will ship no i686/32-bit x86 install media in Fedora 24 as a primary/blocking deliverable. (+7,0,0)
18:49:27 <nirik> so, do we want to make any changes for f23? or keep it as it is?
18:49:49 <sgallagh> I'll vote -1 on changing our deliverables post-Alpha
18:50:00 <jwb> i'd be fine with doing this for f23, but i can see how others would be uncomfortable with that
18:50:06 <debarshi> Since we are changing the meaning of 2ndary architecture, don't we have to do something to make that change "official"?
18:50:06 <sgallagh> (But will not shed any tears if I'm outvoted)
18:50:10 <number80> I guess it could be extended to F23 cloud as currently i686 doesn't build
18:50:16 <nirik> well, we could just not block on them... but not sure if thats helpfull
18:50:38 <jwb> debarshi, we just did that thing.  we voted.
18:50:47 <dgilmore> I would rather not do it. but it really depends on what we decide. since change the tools to put things in different locations is still in early planning
18:50:50 <sgallagh> I'd like to note that changing our deliverables has impact on Websites as well
18:51:02 <sgallagh> I'd prefer to give them a full release cycle to deal with that
18:51:12 <nirik> we didn't demote i686 to secondary... we just said we don't want to ship any images in f24...
18:51:20 <nirik> so it will continue to be a primary as far as building in koji, etc.
18:51:23 <debarshi> jwb: Ok. :)
18:51:50 <dgilmore> we did not build 32 bit cloud for alpha
18:52:20 <jwb> i believe it is up to the WGs to define their deliverables for f23.  if it wasn't shipped for cloud, i don't see a need to ship it in beta/final either
18:52:54 <number80> jwb: the PRD was not updated yet on this point but everyone agreed on not shipping 32bits images
18:52:54 <nirik> so that ties into the release deliverables...
18:53:06 <number80> (for cloud)
18:53:23 <jwb> nirik, the WGs should already be tied into that
18:53:29 <dgilmore> nirik: which we did not do :(
18:53:33 <nirik> ok, so sounds like we want to keep f23 status quo? or does anyone want to float a proposal?
18:53:51 <dgilmore> I think keep f23 as is
18:53:51 <nirik> dgilmore: yeah, there's still an outstanding ticket for it...
18:53:53 <number80> status quo
18:54:04 <dgilmore> but look at changing up the world for f24
18:54:39 <sgallagh> /me remains slightly on the side of sticking with Alpha through F23.
18:54:57 <number80> Moreover, we need to give some space to the marketing folks
18:55:24 <number80> they may want to prepare some communication over the 32bits change
18:55:24 <nirik> ok.
18:55:38 <nirik> anything else on this then?
18:56:09 <nirik> #topic #1472 Investigate mysterious enabled systemd presets
18:56:09 <nirik> .fesco 1472
18:56:09 <nirik> https://fedorahosted.org/fesco/ticket/1472
18:56:10 <zodbot> nirik: #1472 (Investigate mysterious enabled systemd presets) – FESCo - https://fedorahosted.org/fesco/ticket/1472
18:56:30 <nirik> so, how do we want to handle this one?
18:56:32 <dgilmore> I think many of them make sense
18:56:54 <sgallagh> I was planning to go through them one-by-one before the meeting, but fires happened.
18:57:09 <nirik> yeah. been a week of them. ;)
18:57:16 <number80> I'm against abrtd enabled by default especially according the serious security issues related to it
18:57:20 <nirik> perhaps we could get some folks to go through them and propose?
18:57:37 <sgallagh> I wish I could find the original wiki that Lennart referenced.
18:57:56 <sgallagh> I linked to the only wiki I could locate
18:58:12 <sgallagh> Can we at least vote on the policy proposal in my last comment?
18:58:29 <sgallagh> I think that would be useful information for future use
18:58:49 <debarshi> number80: We do have a setting to disable automatic crash reports, don't we?
18:59:11 <nirik> sgallagh: +1 to those items... would be good to have in there...
18:59:15 <number80> debarshi: the solution suggested in RHEL is to disable it
18:59:28 <jwb> sgallagh, i see nothing wrong with your proposal other than it is impossible to enforce.
18:59:48 <sgallagh> jwb: Why would it be impossible to enforce?
18:59:59 <sgallagh> There's a *very* limited number of people with permission to edit the 'fedora-release' package.
19:00:13 <nirik> number80: have a link to the security issue? (I dimly recall something in the past, but can't remember details)
19:00:33 <number80> nirik: from RH knowledge base => https://access.redhat.com/articles/1415483
19:01:03 <jwb> sgallagh, this is scoped to comments only in that package, or to presets installed by any package?
19:01:10 <number80> as long as it's not fixed, I'm not comfortable in having enabled by default
19:01:18 <jwb> sgallagh, because you said the commit to systemd was what enabled these, not the fedora-release package
19:01:27 <sgallagh> jwb: Packaging policy forbids any other package in the Fedora collection from shipping presets
19:01:34 <jwb> and yet...
19:01:41 <sgallagh> jwb: That was prior to them being moved to fedora-release (recently)
19:01:59 <sgallagh> At the same time that policy was put in place.
19:02:11 <jwb> well, i will be skeptically optimistic then
19:02:12 <sgallagh> Previously it was forbidden for any package but systemd to have them.
19:02:36 <nirik> number80: those seem to be closed-errata?
19:03:07 <number80> nirik: solution was to rewrite abrtd so I guess, it's not fixed
19:03:26 <nirik> well, all the bugs I see are closed with updates...
19:03:31 <nirik> unless I missed some
19:03:58 <number80> *sigh*
19:04:11 <number80> I guess, they closed the ticket with minimal fixes
19:04:22 <nirik> anyhow...
19:04:26 * number80 will investigate this further
19:04:50 <nirik> any other votes on sgallagh's policy proposal ?
19:04:51 <thozza> sgallagh: I'm ok with the policy proposal
19:05:00 <jwb> sure +1
19:05:00 <thozza> sgallagh: +1 on the proposal
19:05:12 <sgallagh> +1 to my own proposal (unsurprisingly)
19:05:18 <nirik> still +1 here
19:05:20 <number80> +1 to the proposal
19:05:20 <paragan> +1
19:05:44 <sgallagh> Anyone want to split up the investigation of the services with me?
19:06:14 <nirik> #agreed We will use the policy proposal in comment 2 of ticket 1472 moving foward for presets. (+6,0,0)
19:06:18 <dgilmore> sgallagh: i will if I have the time, but I doubt I will
19:06:39 * nirik does think he will have time either
19:06:49 <paragan> sgallagh, I can help for few services
19:06:56 <number80> sgallagh: I'll take part of it too
19:07:21 <nirik> cool
19:07:23 <sgallagh> Thanks
19:07:40 <nirik> #action sgallagh number80 and paragan will review sevices and we will revist list next week.
19:07:43 <jwb> the only one on that list that really looks weird is lttng-sessiond
19:07:55 <sgallagh> jwb: Well, ladvd is pretty odd too
19:08:02 <sgallagh> That's the one that spawned this investigation
19:08:19 <number80> ack
19:08:41 <jwb> acpid probably needs to go
19:08:52 <jwb> but it isn't weird
19:09:10 <sgallagh> I'm going to recommend that further comments on this should go to the trac ticket.
19:09:21 <nirik> #topic next weeks chair
19:09:31 <nirik> anyone want it?
19:09:35 <jwb> sure
19:10:00 <nirik> cool.
19:10:05 <nirik> #action jwb to chair next week
19:10:11 <nirik> #topic Open Floor
19:10:16 <nirik> any items for open floor?
19:10:28 <sgallagh> /me waves to Phoronix who will no doubt report on the i686 decision
19:10:59 <number80> sgallagh: free advertisement
19:11:10 <number80> not sure, we want it though
19:11:53 <nirik> if nothing else, will close out in a minute.
19:11:59 <number80> thanks gentlemen
19:12:48 <sgallagh> /me wonders who number80 is referring to. Definitely not him.
19:13:03 <nirik> ok, thanks for coming everyone!
19:13:05 <nirik> #endmeeting