13:00:10 <stickster> #startmeeting Workstation WG
13:00:10 <zodbot> Meeting started Mon Jul  3 13:00:10 2017 UTC.  The chair is stickster. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:10 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
13:00:10 <zodbot> The meeting name has been set to 'workstation_wg'
13:00:13 <stickster> #meetingname workstation
13:00:13 <zodbot> The meeting name has been set to 'workstation'
13:00:15 <stickster> #topic Roll call
13:00:17 <stickster> .hello pfrields
13:00:20 <ryanlerch> .hello ryanlerch
13:00:20 <zodbot> stickster: pfrields 'Paul W. Frields' <stickster@gmail.com>
13:00:23 <zodbot> ryanlerch: ryanlerch 'Ryan Lerch' <rlerch@redhat.com>
13:00:31 <mcatanzaro> .hello catanzaro
13:00:32 <zodbot> mcatanzaro: catanzaro 'Michael Catanzaro' <mcatanzaro@gnome.org>
13:00:42 <juhp_> .hello petersen
13:00:43 <zodbot> juhp_: petersen 'Jens Petersen' <petersen@redhat.com>
13:01:09 * stickster looks for that elusive last person to make quorum
13:05:30 <stickster> otaylor should be on his way
13:05:42 <mcatanzaro> stickster: Tomorrow is the change deadline for F27... if you don't mind, I'd like to sneak some Anaconda discussion in at the start of the agenda today, is that OK?
13:06:12 <stickster> mcatanzaro: Sure, we can start with that.
13:08:15 <mcatanzaro> Dot hello otaylor? ;)
13:08:25 <mcatanzaro> (We're waiting for quorum.)
13:08:28 <otaylor> .hello otaylor
13:08:29 <zodbot> otaylor: otaylor 'Owen Taylor' <otaylor@redhat.com>
13:08:42 <mcatanzaro> ;)
13:08:59 <mcatanzaro> stickster, we have quorum!
13:09:03 <stickster> #chair otaylor mcatanzaro juhp_ ryanlerch
13:09:03 <zodbot> Current chairs: juhp_ mcatanzaro otaylor ryanlerch stickster
13:09:07 <stickster> #topic Anaconda
13:09:11 <mcatanzaro> Shame there's not more people here
13:09:12 <mcatanzaro> OK
13:09:26 <stickster> it is what it is... USA holiday tomorrow, so a lot of people took off in between for a long weekend
13:09:32 <mcatanzaro> Oh yeah
13:09:47 * stickster either dedicated or dunderheaded
13:09:51 <mcatanzaro> My job respects no holidays, so easy to forget
13:10:00 <mcatanzaro> Anyway
13:10:18 <stickster> Go ahead, and feel free to use #info liberally for meeting notes
13:10:20 <mcatanzaro> A few years ago (maybe even before Workstation WG, I don't remember how long ago), we discussed the problem of having two different ways of creating user accounts and redundancy between anaconda and gnome-initial-setup
13:10:35 <mcatanzaro> #info anaconda has significant redundancy with gnome-initial-setup
13:10:44 <juhp_> true
13:10:45 <mcatanzaro> rdieter: A few years ago (maybe even before Workstation WG, I don't remember how long ago), we discussed the problem of having two different ways of creating user accounts and redundancy between anaconda and gnome-initial-setup
13:10:49 <mcatanzaro> #chair rdieter
13:10:49 <zodbot> Current chairs: juhp_ mcatanzaro otaylor rdieter ryanlerch stickster
13:11:07 <mcatanzaro> At the time, we decided to fix this by disabling a bunch of settings in anaconda
13:11:27 <mcatanzaro> Then a year or two ago, we had some meetings with Anaconda developers and they provided a configuration file we can use to do just that.
13:11:30 <mcatanzaro> But we have never used it.
13:11:39 <stickster> Right, the one that lets us disable spokes on the hub
13:11:50 <juhp_> aha
13:12:03 <mcatanzaro> So F27 change deadline is tomorrow... I was thinking to take a look at that config file later today and see what's possible with it. And maybe put together an F27 change proposal.
13:12:08 <stickster> (thinking of hub-and-spoke model by which Anaconda exposes, adds, subtracts modules you use for an installation)
13:12:22 <mcatanzaro> I don't think it has to be a finalized proposal tomorrow, but an initial proposal for discussion on devel@.
13:12:31 <stickster> mcatanzaro: So the idea would be to do this for the Workstation install -- essentially turn off the first-user spoke?
13:12:46 <mcatanzaro> So we don't need to fuss over all the details today, I don't think, and can always change our minds and withdraw the proposal later.
13:13:04 <mcatanzaro> stickster: Basically yes, but more than that. We would turn off the first user spoke, but also other spokes that are redundant with gnome-initial-setup.
13:13:07 <juhp_> mcatanzaro: sounds good to me
13:13:17 <otaylor> mcatanzaro: what does that mean? root password?
13:13:26 <otaylor> mcatanzaro: I don't htink there's a *lot* that's redundant
13:13:35 <stickster> mcatanzaro: It would be a *very* good idea to give the Anaconda folks a heads up before you file this Change proposal, or at least at the same time, and orient the Change specifically on having the necessary discussions
13:13:43 <juhp_> otaylor: good question
13:13:50 <mcatanzaro> otaylor: The plan was to have no root password, like Ubuntu. Advanced users can set root password later if desired.
13:14:02 <mcatanzaro> stickster: True! One day notice is not a lot.
13:14:54 <stickster> IOW we don't want this interpreted badly, so we need to go out of our way to make sure it's not
13:15:17 <mcatanzaro> There are also unresolved questions. We've previously agreed that language and keyboard layout selection need to occur before or at the very start of the live session, so those should eventually be removed from Anaconda. But maybe not at first, since we don't have that set up yet.
13:15:19 <stickster> I'm not really sure how or why it would be, but it's too easy for such discussions to start off on the wrong foot
13:15:28 <stickster> and one day notice is not a good precedent here
13:15:47 <stickster> especially the day before a holiday, and also two days before (IIRC) a CZ holiday
13:15:55 <juhp_> but it is just submission deadline, right?
13:16:02 <mcatanzaro> And there was in particular a lot of disagreement as to whether to remove the date and time spoke. Most GNOME folks were thinking the g-i-s page would suffice, but some installer folks are concerned about files potentially being installed with incorrect times.
13:16:09 <stickster> juhp_: correct -- doesn't need to all be coded or figured out -- just the change needs to be submitted
13:16:20 <juhp_> okay
13:16:25 <otaylor> stickster: Well, I think you could phrase it (as you suggested) as a discuss-and-modify and not look like it's a complete list of changes
13:16:31 <mcatanzaro> juhp_: Yes, so there will be plenty of time to change whatever I write up later.
13:16:42 <juhp_> anyway agree that communication around this is good
13:16:46 <stickster> otaylor++
13:16:46 <zodbot> stickster: Karma for otaylor changed to 2 (for the f25 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
13:16:54 <stickster> Oh look, cookie party!
13:18:01 <mcatanzaro> Anyway, I guess what I want approved today is something along the lines of:
13:18:27 <mcatanzaro> #proposed mcatanzaro to prepare F27 change proposal to remove redundancy between anaconda and gnome-initial-setup
13:19:05 <otaylor> +1
13:19:12 <juhp_> +1
13:19:13 <mcatanzaro> And it's too late to provide a specific proposal now, so obviously final approval will have to come later. I just didn't want to surprise the WG by submitting a change proposal without a heads-up first....
13:19:27 <stickster> +1, with caveat of also providing heads-up to Anaconda team
13:19:38 <mcatanzaro> (I mean, I came unprepared, I have not looked at actually trying to use the config file yet. ;)
13:19:39 <ryanlerch> +1 from me too
13:19:54 <mcatanzaro> Right, heads-up to Anaconda team included
13:20:30 <mcatanzaro> rdieter: ^
13:22:48 <stickster> mcatanzaro: assuming you're +1 on your own proposal, we're good to go.
13:23:22 * otaylor missed the fact that the change deadline was right now. Will spend some time on https://fedoraproject.org/wiki/Changes/Graphical_Applications_as_Flatpaks and tag that for the wrangler  unless anybody thinks that's an awful idea.
13:23:33 <stickster> #agreed mcatanzaro to prepare F27 change proposal to remove redundancy between anaconda and gnome-initial-setup (+1: 5, 0: 0, -1: 0)
13:24:16 <juhp_> Yes time new devel cycle starts so early I think?
13:24:26 <juhp_> first time *
13:24:27 <stickster> juhp_: no, the current release cycle starts so late.
13:24:32 <stickster> er, ends so late
13:24:37 <juhp_> okay
13:24:48 <mcatanzaro> I really hope we can get a better spring schedule next year.
13:24:52 <stickster> F26 has been delayed so often now that we're butting far into the F27 devel cycle (it doesn't move)
13:25:12 * stickster not holding his breath, there's so much going on with modularity, CI, flatpaks, etc.
13:25:22 <mcatanzaro> Let's stay alert for when those schedule emails start flying by, we want to encourage the powers that be to set the release closer to March
13:25:38 <mcatanzaro> And if there's too much stuff, the stuff should just wait for the next release. ;)
13:26:26 <mcatanzaro> This year for the first time, I've seen people recommending Ubuntu as a good way to try the latest version of GNOME, because it beat us to GNOME 3.24 by three months... we can't let this become a pattern.
13:26:36 <mcatanzaro> OK, stickster, I think we can move on to next topic. :)
13:26:56 <stickster> mcatanzaro: not sure what you mean by schedule, but F28 is not likely to move from normal ~May 1 target
13:27:36 <stickster> If we were flexing schedule like that, you'd already be seeing a later F27. This is something mattdm has been pretty steady/adamant on.
13:28:03 <stickster> anyhoo...
13:28:39 <stickster> #topic Removing LibreOffice Draw from Workstation
13:28:41 <stickster> #link https://pagure.io/fedora-workstation/issue/16
13:29:05 <stickster> this should be pretty simple... doesn't this just need a change to either comps or the kickstart?
13:29:38 <otaylor> +1 to remove it from the default install.
13:29:40 <stickster> +1 from me, I put this in a category like dia or inkscape, not needed for most productifity users
13:29:42 <ryanlerch> yeah this is a +1 from me.
13:29:53 * stickster needs auto-hunspell apparently
13:30:21 <juhp_> +1
13:30:39 <stickster> once again mcatanzaro gets to +1 his own suggestion... :-D
13:31:13 <mcatanzaro> +
13:31:15 <mcatanzaro> 1
13:31:39 <mcatanzaro> It shouldn't be in kickstart... I'll remove it from comps
13:32:53 <stickster> #agreed Drop it (+1: 5, 0: 0, -1: 0)
13:33:02 <stickster> #action mcatanzaro file PR for fedora-comps to remove libreoffice-draw
13:33:10 <stickster> that was easy!  (*hits Staples button)
13:33:30 <stickster> #topic fedora-workstation-repositories not in default install
13:33:35 <stickster> #link https://pagure.io/fedora-workstation/issue/17
13:34:38 <stickster> Oh wait, it looks like there's no real question for us here today. Still on cschalle to work the policy through Council et al.
13:34:48 * stickster withdraws topic, this is an F27 matter at this point
13:35:12 <stickster> That's all we had for today on the agenda. So...
13:35:15 <stickster> #topic All other business
13:35:20 <stickster> Open floor!
13:35:34 <otaylor> If people want, I can give a brief summary of the status of the flatpak effort
13:36:03 <mcatanzaro> Sounds like you can go ahead
13:36:11 <ryanlerch> yes please!
13:36:20 <juhp_> +1
13:36:40 <otaylor> So, as I linked to above there's a change proposal draft at  https://fedoraproject.org/wiki/Changes/Graphical_Applications_as_Flatpaks and an accompanying detailed plan at https://fedoraproject.org/wiki/Workstation/Flatpaks
13:36:56 <otaylor> I've circulated these to a lot of affected people, though I haven't gotten a ton of feedback
13:37:45 <otaylor> I prototyped using the modularity infrastructure to build the RPMS for flatpak runtimes and apps - and that worked out very well - no infrastructure changes were needed
13:38:18 <otaylor> There are prototype modules https://src.fedoraproject.org/cgit/modules/flatpak-runtime.git/ and https://src.fedoraproject.org/cgit/modules/eog.git
13:38:24 <stickster> otaylor: does that include talking to the Bodhi folks (e.g. bowlofeggs)?
13:38:43 <otaylor> stickster: Yes, I talked to bowlofeggs
13:38:48 * stickster recalls a lot of the foundational work in Bodhi is probably done or on the way, adding content types is supposed to extend to flatpaks too
13:39:01 <stickster> i.e. shares a lot of the same code work that's been done (or close thereto)
13:39:29 <otaylor> stickster: I'll have to update the change proposal there. The basic situation there is that multiple content type support is underway and will be easy to extend to flatpak, however there is no plan for how package updates and container updates interelate
13:39:46 <stickster> otaylor: you meant "yet" ;-)
13:40:25 <otaylor> Well, yes, yet - it's something that needs to be done, but I'm thinking more likely f28 than f27 at this point, since I don't think there's even an idea of how it should work.
13:40:52 <mcatanzaro> Who is bowlofeggs?
13:40:54 <otaylor> I told bowlofeggs that I would try to round up some people to discuss that
13:41:02 <stickster> *nod. clearly we need *some* plan because this is going to get hugely confusing for users and docs folks very quickly
13:41:04 <mcatanzaro> (Who thought that would be a good IRC nick. :)
13:41:16 <stickster> mcatanzaro: there's a story behind it. That's Randy Barlow on the Fedora Engineering team
13:41:23 <stickster> it's a *great* nick!
13:42:36 <otaylor> Continuing with my status report - so after getting the packages built, then next step is to build them into a flatpak
13:43:17 <otaylor> which will be done using hte container build pipeline - I did a quick prototype of this locally, and it worked out pretty well, and have moved on to actually prototyping things in the atomic-reactor tools
13:43:53 <otaylor> I'll be talking to some people on that team to discuss scheduling- they have a lot planned in the next few months, it will be tight to get this in even with us providing the code
13:44:02 <otaylor> talking to them on Wednesday
13:44:25 <bowlofeggs> yeah we had talked about making flatpaks be distributable as containers. there are some questions about how that might be done, but if possible that would simplify things for bodhi
13:44:30 <juhp_> sounds great
13:44:33 <bowlofeggs> if not possible, we can still do it i'm sure, just more work
13:45:38 * stickster notes, we have 10 min remaining, I have to end this meeting at :55.
13:45:49 <otaylor> The next step is to get the built flatpaks into the fedora registry - this is theoretically simple because flatpaks are built as oci images, but complicated because oci image support is not actually there.
13:46:13 <otaylor> There are some patches and it doesn't seem to that hard . I'm optimistic that can get done.
13:46:51 <otaylor> Harder part is providing a index to all the flatpaks - something that gnome software needs. Cockpit needs the same thing for server-side containers, so will try to collaborate there
13:47:32 <otaylor> That the basic status - things are going well in what we've worked on, and we're talking to lots of people, but there's lots of moving pieces :-)
13:48:47 <juhp_> otaylor: thanks, very interesting
13:49:18 <ryanlerch> thanks otaylor
13:49:35 <ryanlerch> do we have to vote on this one? or have we done that already before?
13:50:11 <stickster> Nope, I think this is mostly FYI
13:50:37 <mcatanzaro> "Harder part is providing a index to all the flatpaks - something that gnome software needs."
13:50:47 <mcatanzaro> otaylor: Have you looked at what Endless does? It seems to work fine there.
13:51:02 <otaylor> mcatanzaro: endless is distributing flatpaks via ostree repository
13:51:06 * stickster has an item for AOB too, so let's try to wrap this up into "what's next" please
13:51:08 <ryanlerch> i also have one item too, that we might want to write a change proposal for
13:51:18 <stickster> Ah, I think ryanlerch and I are thinking same thing here
13:51:30 <stickster> my item and his item are the same item :-D
13:52:15 <stickster> otaylor: can you make sure the flatpak status and open matters go to the list for more consumption?
13:52:20 <stickster> most people aren't in this meeting, it seems
13:52:35 <stickster> ryanlerch: go ahead, we have 3 min until close
13:52:40 <stickster> make it count :-D
13:52:46 <otaylor> mcatanzaro: there's developed technology there. but for the OCI Image case - which we're' using to have build pipeline convergence with containers and makes things easier to mirror, there's nothing that works (alexl made a stab at something with his original oci image work, but it was making assumptions about how things would work that are different than the way they ended up working in the end)
13:52:57 <otaylor> stickster: will do
13:53:01 <ryanlerch> the following proposal has been around for a bit:
13:53:04 <ryanlerch> https://fedoraproject.org/w/index.php?title=Workstation/VisualIdentity
13:54:30 <stickster> yeah, next steps here I think are to solidify a real proposal, and we might want to have a "TBD" type Change page for this too... similar to mcatanzaro
13:55:08 <stickster> I doubt it's systemwide because the changes aren't outside the Workstation area, but would be good to cover bases
13:55:16 <mcatanzaro> It's systemwide
13:55:28 <stickster> well then, definitely worth the proposal now then
13:55:29 <otaylor> What would fesco be approving here
13:56:03 <stickster> I'm not sure whether it really merits a page. But it would suck to miss the deadline because someone decides post-facto we needed one
13:56:12 <ryanlerch> so the deadline tomorrow is for systemwide changes?
13:56:17 <mcatanzaro> Yup
13:56:20 <stickster> ryanlerch: yes
13:56:24 <stickster> In any case guys -- we have to close the meeting now, or someone needs to take the gavel. I need to go.
13:56:32 <otaylor> I'm fine as long as it doesn't actually commit us to changing stuff, I'm +0 on actually changing stuff until we see a real proposal and have more discussion around that
13:56:34 <mcatanzaro> Bye!
13:56:42 <stickster> otaylor: correct, just like with Anaconda proposal
13:57:05 <mcatanzaro> All backsies rights reserved!
13:57:07 <stickster> #agreed ryanlerch to work on Change page, proposal coming soon (q.v. Anaconda Change above)
13:57:12 <stickster> #endmeeting