council
LOGS
14:00:03 <mattdm> #startmeeting Council (2017-11-29)
14:00:03 <zodbot> Meeting started Wed Nov 29 14:00:03 2017 UTC.  The chair is mattdm. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:03 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:00:03 <zodbot> The meeting name has been set to 'council_(2017-11-29)'
14:00:05 <mattdm> #meetingname council
14:00:06 <zodbot> The meeting name has been set to 'council'
14:00:07 <mattdm> #chair mattdm jkurik jwb langdon robyduck bexelbie stefw
14:00:08 <zodbot> Current chairs: bexelbie jkurik jwb langdon mattdm robyduck stefw
14:00:21 <jwb> hi
14:00:21 <mattdm> #topic Introductions, Welcomes
14:00:25 <mattdm> Good morning everyone!
14:00:29 <mattdm> Or afternoon or whatever
14:00:34 <jkurik> .hello2
14:00:35 <zodbot> jkurik: jkurik 'Jan Kurik' <jkurik@redhat.com>
14:00:39 <mattdm> This is a special meeting to talk about Fedora Modular Server
14:00:43 <asamalik> good everything, everyone :)
14:00:48 <mattdm> and in particular
14:00:50 <mattdm> #link https://pagure.io/Fedora-Council/tickets/issue/149
14:00:57 <mattdm> #info F27 Modular Server, to ship or not to ship
14:01:09 <bexelbie> .hello bex
14:01:09 <zodbot> bexelbie: bex 'Brian (bex) Exelbierd' <bex@pobox.com>
14:01:10 <asamalik> .hello2
14:01:12 <zodbot> asamalik: asamalik 'Adam Samalik' <asamalik@redhat.com>
14:01:37 <mattdm> sgallagh, are you able to join?
14:01:43 <mattdm> also we kinda need langdon
14:01:50 <langdon> .hello2
14:01:51 <zodbot> langdon: langdon 'Langdon White' <langdon@redhat.com>
14:01:54 <mattdm> magic!
14:02:01 * langdon just didn't notice the start :)
14:02:04 <sgallagh> In another meeting
14:02:16 <mattdm> sgallagh: complete overlap?
14:02:52 <robyduck> .hello robyduck
14:02:53 <zodbot> robyduck: robyduck 'Robert Mayr' <robyduck@gmail.com>
14:02:59 <mattdm> I take that as a yes, I guess :)
14:03:16 <mattdm> #topic Agenda
14:03:21 <langdon> mattdm: sgallagh's opinions should be pretty well represented in the ticket
14:03:23 <sgallagh> mattdm: I will be here.
14:03:33 <mattdm> sgallagh: should we wait a few minutes?
14:03:43 <sgallagh> I just rescheduled the other meeting for later today.
14:03:59 <mattdm> #info This is a special meeting to consider what to do about Fedora Modular Server in light of planned architectural changes
14:04:08 <mattdm> #info see the ticket for details
14:04:13 <mattdm> I suggest:
14:04:18 <langdon> lots and lots of details :)
14:04:33 <jwb> langdon: the discussion was good!
14:04:41 <mattdm> 1. Overview of current state & summary of Server WG meeting (sgallagh, langdon)
14:04:54 <mattdm> 2. Decision making on the specific question of what to do
14:05:14 <mattdm> 3. Wider discussion on modularity as time permits
14:05:21 <mattdm> with #3 broken down further when we get there, perhaps
14:05:24 <langdon> jwb: sorta.. like the discussion is good.. but a believe a lot has been answered elsewhere.. and I would prefer to address "why modularity" as updates to the docs.. if people feel they are unclear
14:05:24 <mattdm> does this make sense?
14:05:43 <mattdm> let's put that in #3 :)
14:05:57 <jwb> langdon: agreed.  doc updates with a reminder/refresher is probably needed
14:06:07 <langdon> i think #2 has multiple choices ... but maybe you meant that?
14:06:20 <langdon> do we have adamw? or is it too early?
14:06:35 <langdon> s/choices/aspects
14:06:37 <sgallagh> langdon: It's 6am his time, so probably too early
14:06:49 <mattdm> yeah, #2 will include all of the choosing
14:06:57 <mattdm> let's jump into #1 though :)
14:07:04 <mattdm> #topic Overview of current state & summary of Server WG meeting
14:07:22 * mattdm hands over the floor, waits
14:07:27 <langdon> to?
14:07:38 <mattdm> you or steve :)
14:07:45 <langdon> so.. i would point to https://pagure.io/Fedora-Council/tickets/issue/149#comment-481875
14:07:54 <langdon> for the server sig opinion
14:08:27 <langdon> i think my opinion is that we ship "something" to gather feedback.. but.. i can see that not being the "server edition"
14:08:36 <langdon> which is really more of a #2 question
14:09:07 <sgallagh> If needed, I can summarize the reasoning why Server SIG came to those decisions
14:09:18 <mattdm> sgallagh: yeah that'd be helpful real quick
14:09:25 <mattdm> also I'm going to #info the stuff from the ticket
14:09:28 <mattdm> #info Server SIG does not want to ship a formal GA of Fedora 27 Modular Server
14:09:36 <mattdm> #info Since the GA RC compose passed the Server tests, Server WG would like to just release the Server Edition content from that compose as the official Fedora 27 Server Edition
14:09:42 <sgallagh> Our biggest concerns are:
14:09:44 <mattdm> #info Recommendation: The Server WG recommendation is not to bother with a Boltron 2 release, but this is ultimately up to the Council and the Modularity WG to decide.
14:10:51 <sgallagh> 1) Negative public response to the limited content set. We hadn't realized previously just how few modules we'd end up having. The lack of access to the complete Fedora Package Collection is a pain point.
14:12:20 <sgallagh> 2) The technical complexity of managing the modules is excessive and is highly unlikely to be supplemented by community packagers.
14:12:49 <sgallagh> Thus remaining a draw on the resources of the modularity members that could otherwise be used towards getting to the F28/"hybrid compose" approach.
14:12:54 <mattdm> yeah those are pretty big
14:13:20 <mattdm> So, as I understand it, there's a Modularity.revised plan which will address _both_ of these?
14:13:32 <langdon> to add to 2) particularly in terms of maintainence
14:14:10 <sgallagh> It will address 1) by having access to the complete package collection. It will mitigate 2) by eliminating many of the interdependencies that have plagued the modules thus fa.
14:14:12 <sgallagh> *far
14:14:17 <langdon> basically from the outset.. we have wanted to have an "everything else" module that would solve 1) .. and maybe 2) .. but the overhead of doing it as a "real" module turned out to be very high
14:14:34 <sgallagh> s/very high/insurmountable/
14:14:56 <sgallagh> There has simply never been a moment where all of Fedroa built successfully at the same time.
14:15:00 <langdon> our proposal is to "cheat" a bit and use traditional rpms as a "pseudo module" and allow the "everything repo" to act as an "everything module" ... for anything not modularized yet
14:15:04 <mattdm> Would it be fair to say "we thought it would be hard, but it turned out to be impossible when we actually got to it"
14:15:33 <mattdm> How much work is this change? Like, in terms of human-hours?
14:15:36 <langdon> yeah.. one of those .. on paper it seemed reasonable.. the intricacies of all the rpms and their packaging made it a non-starter
14:16:03 <langdon> well.. the hybrid compose is a bunch of effort ... but the modular part is actually lower once the hybrid works
14:16:17 <langdon> pungi doesn't really like the idea of mixing traditional rpms and modules
14:16:20 <sgallagh> mattdm: I don't like the word "impossible", but as in this case we're likely talking about timeframes referencing the average lifespan of *planets*, I'll settle for it :)
14:16:26 <mattdm> ha
14:16:36 <langdon> sgallagh: unpossible then?!?
14:16:58 <langdon> i don't want to speak for rcm.. but i have heard numbers like 1 man month for hyrid compsoe
14:17:02 <langdon> *compose even
14:17:21 <mattdm> I'm going to put a pin in "a bunch of effort" for later in the meeting
14:17:29 <sgallagh> Yeah, we're definitely talking timescales for hybrid compose that are better aligned with the 28 schedule.
14:17:33 <mattdm> For now, does everyone on the council understand the state of things and the server WG recommendations?
14:18:19 <robyduck> yes
14:18:21 <jkurik> mattdm: I believe I do
14:18:43 <robyduck> may I add a thought?
14:18:47 <mattdm> robyduck: please do!
14:18:56 <robyduck> We have a Modular Server Beta release, so we cannot just ship something. If we ship Modular Server, it needs to be final. If not, and the Server WG wants to ship what we have *now* to get feedback, let's call it a Beta-2 stadium. This way we keep a GA release (trad Server F26) and a testing release.
14:19:08 <jwb> yes
14:19:27 <jwb> er
14:19:31 <robyduck> would that be far away from what the Server WG wants?
14:19:37 <bexelbie> yes, but ... I want the docs updated to better
14:19:41 <jwb> my yes was to the "does the council understand" question
14:19:41 <langdon> robyduck: modularity-wg wants the feedback more than server sig ;)
14:19:43 <bexelbie> help others understand why we are doing this
14:19:50 <sgallagh> robyduck: Yes, we're talkinga bout abandoning the Modular Server GA for 27
14:19:51 <bexelbie> we have to get that story clearer
14:20:08 <bexelbie> especially because of the statement that the goals haven't changed but the architecture does sound very different in construction
14:20:13 <bexelbie> even if the end state is the same
14:20:22 <mattdm> robyduck: Yeah. :) They're asking to not ship the Modular Server GA at all, and release the traditional Server F27 bits on getfedora (they're already on the mirrors anyway)
14:20:39 <langdon> s/asking/planning
14:20:53 <sgallagh> robyduck: Server SIG and Modularity WG are not in complete agreement.
14:21:08 <langdon> well.. neither is modularity-wg :)
14:21:09 <robyduck> sure, but should we keep the Mdular Beta state?
14:21:18 <sgallagh> Server SIG doesn't want to ship the Modular Server as it stands today because it is underpowered and we're going to redesign it for F28.
14:21:24 <robyduck> I guess that would help to get more feedback
14:21:31 <sgallagh> It seems wasteful to burn goodwill on testing a dead pathway.
14:21:43 <langdon> robyduck: the concern is that keeping that (f27 m beta) useful would be a lot of work that could be spent on f28 M
14:21:53 <mattdm> Yeah, the problem is: if the feedback is on something we know we aren't going to do, it's not so useful
14:21:59 <sgallagh> robyduck: ^^ what langdon said
14:22:12 <robyduck> langdon: ok got it, the throw this away ;)
14:22:17 <langdon> if we are moving to #2 .. i like the idea that we discussed of "boltron-2" == "f28 very very very early alpha"
14:22:21 <sgallagh> Now, to be fair, feedback on the user experience of dealing with modules *is* useful. That experience shouldn't change.
14:22:27 <mattdm> hmmm, I think maybe we are getting into the decision making part of the meeting :)
14:22:35 <mattdm> #topic In which we make a decision
14:22:37 <asamalik> A simplified viz of the difference between the current and new approaches: https://asamalik.fedorapeople.org/modularity-hybrid.png
14:22:42 <sgallagh> However, we already have a LOT of feedback on that, thanks to the UX sessions at FLock
14:22:45 <mattdm> #link https://asamalik.fedorapeople.org/modularity-hybrid.png
14:23:04 * langdon grumbles not that all of it was adopted
14:23:56 <langdon> we would like to do more feedback sessions and i have talks proposed.. so i would like "something" to present..
14:23:57 <mattdm> langdon: do you want to put the open questions in the form of questions?
14:24:05 <langdon> hence boltron-2 is important to me
14:24:07 <bexelbie> I am uncomfortable with this being called a preview of F28
14:24:19 <langdon> bexelbie: hence.. it is called boltron-2
14:24:24 <mattdm> or should I just make a proposal?
14:24:25 <langdon> or something funnier for mattdm
14:24:37 <langdon> mattdm: i didn't understand what you were asking me for
14:24:57 <mattdm> It's true, I will not be happy if it doesn't have a funny sequel name.
14:25:14 <mattdm> So, I think it's already decided that:
14:25:26 <mattdm> a) we won't ship the F27 modular stuff any further -- it's getting dropped
14:25:53 <mattdm> b) assuming no last-minute gotchas, we'll declare F27 Server bits (traditional) as official after all
14:26:04 <sgallagh> mattdm: a) is what Langdon is still arguing
14:26:32 <mattdm> and there's an open question about what to do to get Langdon something to demo in presentations and talks (like FOSDEM)
14:26:41 <mattdm> not to mention to get further user involvement and feedback
14:26:47 <sgallagh> I had a proposal for that yesterday
14:26:50 <sgallagh> May I?
14:26:56 <mattdm> sgallagh: go
14:27:12 <langdon> sgallagh: actually.. i am way more concerned about mattdm's characterization.. not so much shipping f27-m
14:27:30 <sgallagh> To the best of my knowledge, the Bodhi updates system for modules works with the F27 Modular Server Beta.
14:27:50 <sgallagh> So if we REALLY need some updated packages for demoing, we can install from the beta and use Bodhi to provide updates to it.
14:28:10 <sgallagh> This doesn't require any new composes/RCM work or to make another installable release.
14:28:49 <sgallagh> It does put resources on maintaining things in the 27 path that could be used elsewhere, but that decision would be confined to the people doing the work
14:29:06 <sgallagh> EOF
14:29:08 <jwb> would that extend to adding NEW modules to f27?
14:29:34 <sgallagh> jwb: I haven't personally tested that, but I *think* so?
14:30:05 <jwb> i think that'd be important.  the modules for f27 are rather limited even for demoing?
14:30:33 <mattdm> and then presumably we would drop that as soon as rel-eng (is this threebean? mboddu? someone else?) has an F28-modular-thing-based-on-the-new-arch?
14:30:39 <bexelbie> Does this need to be able to be used by anyone other than the demo-er?  If not, they could do manual spins right?
14:30:45 <bexelbie> since we are really demoing concept and UI?
14:31:36 <langdon> bexelbie: would prefer something more "official" .. else we are back to original boltron
14:31:58 <bexelbie> but it seems a big request to put resources on an architectural dead-end
14:32:11 <langdon> however.. i am as concerned as sgallagh about effort in to a dead end.. so I would much prefer to see "very early alpha" as the demo platform
14:32:12 <bexelbie> we are redoing it so if the work doesn't contribute to that we should minimize
14:32:33 <asamalik> I would say it *should* be easy to handcraft a demoable container, using the new hybrid approach, before DevConf/FOSDEM.
14:32:51 <sgallagh> langdon: Well, nightly builds will be started as soon as possible
14:33:03 <sgallagh> Those are probably "official-enough"
14:33:20 <mattdm> sgallagh: nightly builds of the F28 thing?
14:33:30 <sgallagh> mattdm: yes
14:33:40 <mattdm> and is "as soon as possible" after the "month of human-hours" langdon mentioned earlier?
14:33:54 <sgallagh> mattdm: To clarify, the nightly builds I'm talking about will be the standard Fedora builds.
14:33:59 <mattdm> and do we have one or ten people digging this particular post-hole?
14:34:08 <sgallagh> One of the benefits of moving to this approach is that we won't need a separate set of composes for Modular.
14:34:18 <langdon> at present, we do have one.. we may be able to get more people focused on it
14:34:21 <sgallagh> It will just be added onto the traditional ones as supplementary
14:34:32 <langdon> and... in the interim.. we can just use two repos
14:34:42 <langdon> one w/ traditional and one w/ modules
14:35:00 <langdon> i have actually prototyped this and it works.. but we have some open questions on the build side
14:35:18 <sgallagh> I think we need to have an internal discussion about getting access to resources in order to have a demoable hybrid compose in time for DevConf.cz/FOSFEM
14:35:21 <sgallagh> *FOSDEM
14:35:31 <mattdm> I'm feeling *more* confused rather than less
14:35:33 <sgallagh> But I think that's achievable
14:35:52 <sgallagh> mattdm: I'll try to explain
14:35:54 <mattdm> sgallagh: "internal" = red hat?
14:35:54 <jwb> mattdm: same
14:36:04 <sgallagh> Yes, RHT
14:36:33 <sgallagh> With the current approach, anything represented as a module has to be built from the Module Build Service as a unit.
14:36:58 <mattdm> so far so good :)
14:36:59 <sgallagh> With the new approach, we will take existing RPMs and add metadata to "group" them into modules
14:37:19 * langdon calls ^^ pseudo-modules
14:37:26 <sgallagh> So we can add this metadata atop a traditional compose, rather than rebuilding the packages
14:38:00 <sgallagh> These pseudo-modules can then be consumed by MBS to build what we call App-Stream modules, which are the things we want to be swappable
14:38:09 <mattdm> and as I understand it, there will be a pseudo-module which consists of "every package in the fedora general collection"?
14:38:18 <sgallagh> Like Node.js, Ruby/Rails, etc.
14:38:35 <sgallagh> mattdm: well, more like "every package not in another named module", but close enough
14:38:48 <langdon> mattdm: correct.. i think i proposed "host", "platform", and "everything else"
14:38:50 <mattdm> sgallagh: every package with an f28 branch?
14:38:56 <sgallagh> mattdm: yes
14:38:56 <asamalik> mattdm: the shadow square in my picture basically
14:39:43 <sgallagh> mattdm: Whereas most modules will use arbitrary branching to indicate their upstream version
14:40:04 <mattdm> Okay. I understand all this.
14:40:05 <sgallagh> (modules meaning real modules, not pseudo-modules in that sentence)
14:40:18 <mattdm> I guess what I don't understand is the timeframe.
14:40:41 <langdon> so.. to add to that ...  we can model this scenario with two repos.. one traditional w/ or w/o module metadata and one modular buitl with mbs.. until we can build one "hybrid repo"
14:41:22 <mattdm> is building a hybrid repo even desirable? why not _keep_ it split like that?
14:41:42 <langdon> because.. at some point.. we will want a module directly in the "main repo" ...
14:41:49 <sgallagh> mattdm: simplicity of composing only one "Fedora"
14:41:56 <langdon> so that we can swap it with some other module
14:41:57 <asamalik> and we can already build modules against non-modular content - the bootstrap module pointing to the f27-trad package set - we actually mostly use this to build existing modules... the "new" thing is doing the compose
14:42:51 <mattdm> sgallagh: I'm not "fedora = one repo" is desirable. can we keep that in mind but not answer it here? :)
14:42:52 <langdon> i think the point is valid though.. we don't *need* the hybrid repo .. but it is definitely desirable
14:42:54 <sgallagh> asamalik: That's more of a hack taking advantage of an implementation detail of Koji than an intended feature :)
14:43:36 <sgallagh> mattdm: Whether the output is a single repo, I think we want it all to be one *compose process*
14:43:41 <asamalik> sgallagh: yes, a hack that became a common practise and the future :P
14:43:47 <mattdm> Okay, so, is the open question: can something from this be put together to a) demo from and b) collect feedback on by the same timeframe we expected F27 Modular Server?
14:43:56 <sgallagh> Which generates the psuedo-modules appropriately for each output repo
14:43:58 <mattdm> sgallagh: maybe?
14:44:08 * mattdm is not an adherent of the religion of the one true compose
14:44:26 <mattdm> because it's 9:45 and I have another meeting in 15 minutes
14:44:32 <langdon> ha
14:44:48 <langdon> suffice to say.. we can do without it.. but i think we will want it soon'ish
14:44:49 <sgallagh> mattdm: I'm not prepared to commit to the same timeframe, no
14:45:05 <sgallagh> I'd consider the DevConf/FOSDEM timeframe more realistic.
14:45:37 <mattdm> Okay, so, that means the question is
14:46:20 <bexelbie> From a planning question we need to answer this - but at a council level it seems we just need to approve messaging about what we are expecting to have happen so that people undrestand why we are releasing F27
14:46:54 * dgilmore perks his ears at hacks and koji
14:46:56 <mattdm> which of a) release an f27 updated modular server for demoing b) make langdon take the beta f27 and demo it with updates until f28 version is available, or c) focus everything on the f28 modular server?
14:46:56 <bexelbie> and language like "continuing releases to show UI, demos and functionality" covers it
14:47:53 <langdon> dgilmore: discussing a hybrid compose of modules and traditional rpms.. and the effort that might be required
14:48:00 <sgallagh> mattdm: I'm not sure b) or c) are really Council decisions.
14:48:12 <sgallagh> I think the Council mostly just needs to decide what official releases we're doing.
14:48:23 <dgilmore> langdon: okay, so not likely a hack, but a useful change
14:48:36 <sgallagh> b) and c) are really just "if someone does it, then it gets done"
14:49:07 <sgallagh> dgilmore: We were talking about the current "bootstrap" module, which is basically hacked up by us doing manually tagging because the module isn't actually buildable properly.
14:49:13 <mattdm> sgallagh: Okay, so, back from way before, I think we're okay with
14:49:16 <mattdm> Server SIG does not want to ship a formal GA of Fedora 27 Modular Server
14:49:17 <mattdm> Since the GA RC compose passed the Server tests, Server WG would like to just release the Server Edition content from that compose as the official Fedora 27 Server Edition
14:49:25 <sgallagh> The fact that it worked was not so much "designed" as a "happy accident"
14:49:39 <langdon> sgallagh: yes and no.. i am representing fedora a bit in these "positions" so i would like the "plan" validated by "fedora"
14:49:40 <sgallagh> From our perspective; it's proper Koji behavior from RCM's side
14:49:41 <mattdm> Is everyon on the council good on those two points?
14:49:48 <bexelbie> yes
14:50:07 <robyduck> +1
14:50:09 <mattdm> Okay, then...
14:50:14 <langdon> +1
14:50:18 <jkurik> yes, +1
14:50:26 <bexelbie> Proposal: Release the GA sping of F27 server mentioned above as the F27 server release; give modularity WG permissoin to continue to release boltron derivatives for dmeo/testing as needed
14:50:41 <mattdm> bexelbie: +1
14:50:44 <bexelbie> that seems to cover it from our perspective .. and makes no commitments about f28
14:50:48 <mattdm> I was just typing sommething similar
14:50:57 <bexelbie> type it better :D
14:51:07 <langdon> Proposal: Release the GA spin of F27 server mentioned above as the F27 server release; give modularity WG permission to continue to release boltron derivatives for demo/testing as needed
14:51:09 <bexelbie> and fix the spelling of permission
14:51:24 <jwb> +1
14:51:26 <mattdm> +1
14:51:30 <bexelbie> +1
14:51:31 <jkurik> +1
14:51:34 <robyduck> +1
14:52:00 <sgallagh> Addendum: Announce that F27 Modular Server GA will not be forthcoming.
14:52:01 <mattdm> okay, so I will write a Fedora Magazine post. I guess robyduck to update the website
14:52:04 <bexelbie> I would encourage Modularity or Server to come back to the Council if there are resource concerns we can help advise about
14:52:24 <mattdm> do we want to plan this for next tuesday?
14:52:28 <mattdm> #topic Timing
14:52:34 <jkurik> and what about Go/No-Go for the already delivered F27 Server bits (traditional) ? Are we comfortable to wave it as it is, or do we want to have the Go/No-Go for it ?
14:52:39 <sgallagh> mattdm: I think we should have a formal Go/No-Go decision (tomorrow?)
14:52:51 <mattdm> sgallagh, jkurik sounds good to me
14:52:59 <mattdm> robyduck: what do you need from a websites perspective?
14:53:02 <robyduck> +1 for formal decision and release date
14:53:17 <mattdm> jkurik: can you schedule that meeting?
14:53:17 <robyduck> not much, it is less work than Modular Server GA
14:53:21 <jkurik> I will not be able to run the Go/No-Go this week due to many meetings
14:53:34 <mattdm> robyduck: so, yay? :)
14:53:35 <sgallagh> With the caveat that a No-Go tomorrow probably means no F27 Server, because Server SIG does not want to attempt to build a new release candidate.
14:53:35 <jkurik> I can schedule it, but I will not be present
14:53:44 <robyduck> yay
14:53:51 <robyduck> :)
14:53:58 <jwb> sgallagh: where would you like that announcement to be and from whom?
14:54:09 <mattdm> sgallagh: we'll cross that bridge if there happens to be a river there
14:54:20 <sgallagh> mattdm: Sure, but I figured I'd mention it
14:54:26 <mattdm> *nod*
14:54:47 <mattdm> I'll do the announcement -- many people have already volunteered to help :)
14:54:53 <jkurik> sgallagh: may I ask you to run the Go/No-Go tomorrow ?
14:54:54 <sgallagh> jwb: I'll volunteer to co-author the announcement mattdm is writing as the Server Edition representative.
14:55:06 <jwb> ok
14:55:09 <sgallagh> jkurik: Can you email me a link to the SOP doc?
14:55:17 <sgallagh> Or the template or whatever you use normally?
14:55:23 <sgallagh> If so, yes
14:55:25 <jkurik> sgallagh: will do, thanks
14:55:32 <mattdm> #topic last things
14:56:08 <mattdm> asamalik, langdon -- I think it would be very, very helpful if https://docs.pagure.org/modularity/docs.html would get a section on "what this solves"
14:56:15 <mattdm> with some user stories
14:56:45 <bexelbie> And that messaging needs to get more broadcast than just that page
14:56:47 <mattdm> possibly the things I wrote at https://pagure.io/Fedora-Council/tickets/issue/149#comment-481780
14:56:52 <mattdm> rephrased into user stories
14:57:22 <mattdm> and I think it needs a prominent "why not just containers?" FAQ
14:57:26 <langdon> The problem is "channel" more than content.. We have tons of content.. We just can't market it well
14:57:42 <mattdm> because whenever I talk about this to the press, I spend most of my time going in circles over that
14:58:21 <langdon> Did u get my last? Connection is wonky
14:58:25 <mkolman> +1 from my personally for a list of actual benefits
14:58:46 <mattdm> langdon: channel? yes. communication is hard :)
14:58:57 <mattdm> but I think there *is* a problem with conent
14:58:59 <mattdm> content
14:59:04 <mkolman> so that people know all the necessary effort/overhead is done for good reasons
14:59:06 <bexelbie> I have struggled with this messaging at booths too
14:59:32 <mattdm> also I want to throw in that this new plan looks more and more like
14:59:36 <mattdm> #link https://mattdm.org/fedora/2015rings/
14:59:37 <langdon> Perhaps.. Of course w/o feedback on the content we don't know what is missing :/
14:59:51 <langdon> Mattdm pfft
15:00:01 <asamalik> mattdm: yeah
15:00:08 <mattdm> langdon: ^ the things I said above are definitely missing
15:00:52 <jwb> langdon: one of the questions i keep seeing is whether or not modularity aims to provide parallel availability AND installability
15:01:19 <langdon> I am not sure I agree.. They may be missing from where you are looking but that doesn't mean missing.. We literally have a hundred pages of content and videos and talks etc
15:01:24 <mattdm> jwb yeah that also need to go in the FAQ
15:01:34 <mattdm> langdon: they're missing from the thing that says "modularity docs"
15:01:57 <mattdm> "go find the overview -- it's in these hundreds of pages of content and videos, some of which is dead ends and stuff" does not help
15:01:59 <bexelbie> langdon, that much content is not addressable
15:02:11 <langdon> Right
15:02:14 <mattdm> I know some of it is in https://docs.pagure.org/modularity/
15:02:24 <mattdm> in asamalik's pretty pictures
15:02:41 <mattdm> but I think we also need simple textual explanations in https://docs.pagure.org/modularity/docs.html
15:02:57 <mattdm> (those two links look similar but are very different)
15:03:02 <mattdm> and anyway need to end meeting
15:03:04 <mattdm> thanks everyone
15:03:13 <mattdm> bexelbie: can I ask you to close the ticket with the thing we decided?
15:03:35 <bexelbie> yes
15:03:36 <asamalik> I'll definitely look into that
15:03:41 <bexelbie> did we get a #agreed on it?
15:03:56 <bexelbie> #agreed Release the GA spin of F27 server mentioned above as the F27 server release; give modularity WG permission to continue to release boltron derivatives for demo/testing as needed
15:04:15 <mattdm> bexelbie: thanks
15:04:29 <mattdm> bexelbie: yes. although no # happened, there were enough yesses and no nos
15:04:32 <mattdm> #endmeeting