13:02:29 <ryanlerch> #startmeeting Workstation WG Meeting
13:02:29 <zodbot> Meeting started Mon Sep 24 13:02:29 2018 UTC.
13:02:29 <zodbot> This meeting is logged and archived in a public location.
13:02:29 <zodbot> The chair is ryanlerch. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:02:29 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
13:02:29 <zodbot> The meeting name has been set to 'workstation_wg_meeting'
13:03:22 <stickster> probably want: #meetingname workstation
13:03:28 <kalev> morning
13:03:37 <ryanlerch> #meetingname workstation
13:03:37 <zodbot> The meeting name has been set to 'workstation'
13:03:37 <stickster> so it gets filed in the same place with others :-)
13:03:48 <ryanlerch> #topic roll call
13:03:52 <ryanlerch> thanks stickster!
13:03:58 <ryanlerch> .hello ryanlerch
13:03:59 <zodbot> ryanlerch: ryanlerch 'Ryan Lerch' <rlerch@redhat.com>
13:04:01 <juhp> .hello petersen
13:04:03 <stickster> .hello pfrields
13:04:03 <zodbot> juhp: petersen 'Jens Petersen' <petersen@redhat.com>
13:04:06 <zodbot> stickster: pfrields 'Paul W. Frields' <stickster@gmail.com>
13:04:13 <stickster> howdy all  o/
13:04:20 <aday> o/
13:05:16 * stickster looks for mcatanzaro otaylor rdieter
13:05:17 <ryanlerch> that's 4, how many do we need?
13:05:30 <stickster> ryanlerch: cschaller should be here at about :20
13:05:38 <juhp> 5 I think
13:05:57 <kalev> 7win 27
13:05:59 <kalev> oops
13:06:40 <stickster> if there's a discussion bit we could take care of that part early, then force latecomers to vote when they show up :-D
13:07:02 <ryanlerch> okies, first up let's try this one
13:07:30 <ryanlerch> #topic drop Workstation install tree https://pagure.io/fedora-workstation/issue/45
13:07:56 <ryanlerch> stickster: any update on this one, or untag it from meeting?
13:08:02 <rdieter> .hello rdieter
13:08:03 <zodbot> rdieter: rdieter 'Rex Dieter' <rdieter@gmail.com>
13:08:03 <rdieter> hi
13:08:44 <stickster> ryanlerch: We should probably set an 'f30' target on it because it's unlikely we're going to make this kind of big change to Anaconda pre-f29
13:09:04 <otaylor> here now, sorry to be late
13:09:11 <stickster> ryanlerch: but the anaconda team is apparently looking at doing a new implementatino of how install classes work
13:09:22 <stickster> *that's italian for "implementation"
13:09:29 <juhp> #chair if you want :)
13:09:59 <ryanlerch> #chair stickster juhp rdieter otaylor kalev
13:09:59 <zodbot> Current chairs: juhp kalev otaylor rdieter ryanlerch stickster
13:10:26 <juhp> stickster: lol
13:10:39 <stickster> #idea Tag the installclass change with 'f30' in our tracker so we can delineate between anything to be done quickly for F29 GA
13:11:27 <ryanlerch> stickster: +1
13:11:34 * stickster goes to create milestones
13:11:40 <ryanlerch> stickster: already done!
13:11:43 <juhp> So we don't want to drop for F29?  Just checking
13:11:43 <ryanlerch> :D
13:12:26 <ryanlerch> happy to move on from this one?
13:13:04 <juhp> Well it sounds better to have the kernel line install class of course, but all these trees take up space on the mirrors too
13:14:29 <stickster> juhp: I tend to think a more useful space saving for now would be clear out all pre-f21 trees and move to archive (and notify mirrors so they could grab the whole enchilada there if they care to)
13:14:39 <stickster> we haven't done that pruning in a while IIRC
13:14:40 <juhp> okay
13:14:56 <juhp> fair enough
13:15:01 <stickster> agreed this is worth doing, it's just that we won't get the concomitant anaconda change in time for f29
13:15:09 <juhp> Right
13:15:29 * stickster eof
13:15:56 <ryanlerch> #topic Replace plymouth theme https://pagure.io/fedora-workstation/issue/53
13:16:43 <stickster> oh wait, ha... just determined that the new mirror defaults are that they don't carry anything behind the supported releases :-D
13:17:09 <stickster> i.e. f26 trees and previous are blank in releases/
13:17:16 * stickster really eof
13:17:18 <ryanlerch> whoops sorry, jumped the gun a bit there
13:17:26 <stickster> nope, I just didn't shut up
13:17:37 <ryanlerch> any thoughts on the plymouth theme?
13:17:57 <aday> aruiz, issue is https://pagure.io/fedora-workstation/issue/53
13:18:14 <aday> ryanlerch, my last comment on the issue sums up my position
13:18:58 <aday> hans and jimmac should probably review the current state, but changing it for now seems like a positive
13:19:22 <stickster> Wouldn't mind see something fresher and prettier there. And having it be related to Fedora makes sense; there's plenty of warning to try something for f30 here
13:19:48 <aruiz> so, the only feedback I'd like to add is that hans has been working on flicker free boot and one of the things we've been looking at is keeping the UEFI vendor logo on UEFI systems instead of adding yet another transition
13:19:52 <juhp> Is there a screenshot or video somewhere of the Spinner theme?
13:20:03 <aruiz> Microsoft enforces a spec as to where lements should be placed
13:20:06 <aruiz> https://msdn.microsoft.com/en-us/library/windows/hardware/mt683761(v=vs.85).aspx
13:20:21 <otaylor> aruiz:so vendor logo => gdm ?
13:20:23 <ryanlerch> juhp: it is just a grey background, with a spinner on it
13:20:23 <halfline> juhp: it's the same one used in rhel7 basically (well rhel7 has a small watermark on it too)
13:20:23 <aruiz> so we want a spinner theme, but following the spec
13:20:34 <aruiz> otaylor, that's what we already have in F29
13:20:35 <ryanlerch> but no fedora logo anywhere
13:20:37 <juhp> halfline: okay
13:20:41 <halfline> otaylor: only if boot is less than 5 seconds
13:20:46 <aruiz> exactly
13:21:11 <stickster> Wouldn't mind see something fresher and prettier there. And having it be related to Fedora makes sense; there's plenty of warning to try something for f30 here
13:21:15 <stickster> oops, sorry
13:21:16 <otaylor> halfline: OK, that makes more sense - I wouldn't want to keep the UEFI logo over a system update or a  boot
13:21:27 <otaylor> or a spinning rust boot
13:22:05 <halfline> there's also a problem where some efi systems use 800x600 by default (ugh)
13:22:15 <ryanlerch> is it hard to put the fedora logo in here in the same position it is in GDM?
13:22:17 <halfline> so there's discussion of turning it off in that case too
13:22:56 <stickster> halfline: what's "it" in that sentence?
13:23:53 <aday> ryanlerch, in a slow boot situation, or in all cases?
13:24:00 <stickster> plymouth?
13:24:13 <halfline> ryanlerch: well i wouldn't say it's hard, but it would require source changes. putting it in the right corner could be done with just a text file change
13:24:52 <halfline> stickster: it → the persistent uefi logo
13:25:03 <stickster> ah, hm.
13:25:06 <ryanlerch> halfline: theme source changes or plymouth changes?
13:25:25 <halfline> well the theme source is part of the plymouth source
13:25:32 <ryanlerch> aday: does the plymouth theme show in slow boot situations?
13:25:39 <aday> maybe we should try and steer this discussion clear of actually designing the boot screen
13:25:56 <halfline> basically spinner uses a plugin called two-step that has a built in feature for a "corner image"
13:26:13 <stickster> we don't need to design it here, but let's not have anyone walk away thinking no one is interested in having a Fedora-specific boot screen, either.
13:26:15 <halfline> but two-step would need to grow a feature for bottom middle image
13:26:15 <aday> if the wg wants more guidance we can provide it
13:26:35 <halfline> but that's like 10 lines of code or something
13:26:38 <aday> if there are specific requests or interests a group of us can look into it and come back
13:26:41 <juhp> halfline: what's your suggestion then?
13:26:55 <halfline> i'm not suggesting anything, just adding information
13:27:05 <stickster> halfline++
13:27:05 <zodbot> stickster: Karma for rstrode changed to 6 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
13:29:14 <juhp> Fedora can boot in less in than 5s? 8-)
13:29:16 <stickster> Seems like the best thing for the WG to do/decide here would be to have a small group of interested folks talk via IRC/list to understand the relevant changes (and in which circumstances things work), and make a list of things that need to be done
13:29:24 <ryanlerch> so the basic proposal here is to drop the fedora charge theme, for a plymouth boot screen without a fedora logo?
13:29:26 <halfline> juhp: on some machines it can !
13:29:31 <stickster> juhp: it's pretty darn fast on my new laptop!
13:29:31 <juhp> Nice
13:29:51 <stickster> ryanlerch: I think the proposal is for that *in absence of anything else*
13:29:56 <halfline> one idea is we could do like what rhel 7 does
13:30:06 <juhp> I should measure mine but pretty sure it is more than 5
13:30:07 <aruiz> stickster, if that's the case then I'd like to chime in
13:30:09 <halfline> where it puts a highly stylized watermark over it
13:30:14 <halfline> not really a logo per se
13:30:20 <aday> right, this isn't necessarily a permanent change
13:30:24 <aruiz> Hans has been working on improving boot UX and we've already reached out to aday and jimmac in the past
13:30:42 <aruiz> Hans has been doing some of the initial work for F29 for silent boot and UEFI vendor logo -> GDM flickr-free
13:30:50 <aruiz> we had this in our to-do for the F30 cycle
13:30:59 <halfline> right, that's clearly the best way forward for systems that can do it
13:31:00 <aruiz> but it's pending visual design input
13:31:10 <ryanlerch> so it changes this release, then potentially again next release?
13:31:14 <juhp> halfline: should just make the watermark more subtle ;o)
13:31:29 <aruiz> my suggestion is, why don't we postpone this decision to F30 in case we fail to provide anything?
13:31:36 <aruiz> ryanlerch, that's my point
13:31:53 <ryanlerch> aruiz: i gotta agree with you here :)
13:32:11 <aruiz> I'm volunteering to steer the effort to have an implementation of something that is both visually pleasant and that identifies Fedora
13:32:19 <aday> i'd be fine with that
13:32:33 <juhp> F30 sounds realistic to me
13:32:37 * stickster hoping to avoid having things change without e.g. mattdm awareness, and then having an 11th hour disagreement over how things look
13:32:39 <aruiz> and Hans will have room to implement this (provided we get visual design assets and halfline reviews his patches ;-)
13:32:45 <ryanlerch> aruiz: i can work with you to help out too if you want!
13:32:52 <aruiz> ryanlerch, I'd love to
13:33:06 <stickster> aday++ ryanlerch++
13:33:06 <zodbot> stickster: Karma for aday changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
13:33:39 * stickster brb
13:34:07 <ryanlerch> #propoal aim for f30 to have a new plymouth theme, replacing charge
13:34:21 <ryanlerch> #proposal aim for f30 to have a new plymouth theme, replacing charge
13:35:09 <ryanlerch> +1 from me
13:35:15 <juhp> +1
13:36:50 <ryanlerch> anyone else?
13:37:24 <ryanlerch> stickster rdieter otaylor kalev
13:37:34 <rdieter> sorry, +1
13:37:49 <otaylor> +1
13:38:26 <kalev> +1
13:38:27 <cschalle_> +1
13:38:39 <otaylor> (assuming that we work to something - I don't agree with mcantazaro's proposal of just switching to spinner as the default course if nothing else happens)
13:39:05 <ryanlerch> otaylor: +1
13:39:24 <ryanlerch> That is #agreed then, i suppose
13:39:48 <ryanlerch> let's jump to the other plymouth issue in the queue::
13:40:06 <ryanlerch> #topic F29 plymouth regression https://pagure.io/fedora-workstation/issue/77
13:42:27 * stickster back
13:42:33 <ryanlerch> anyone know any more background on this one?
13:43:27 <cschalle_> sounds more like a bug report than a wg issue?
13:43:36 <stickster> aruiz: ^ maybe can shed some light? I haven't seen it and can't comment. Maybe I should move my new laptop to F29 to see what happens
13:43:51 <cschalle_> my laptop is F29 and I don't have the issue
13:43:59 <stickster> yeah, seems like a bug. an ugly one, but not clear how many people it's going to affect.
13:44:18 <aruiz> stickster, I'm assuming this is one for halfline?
13:44:28 <aruiz> ah, this is related to Hans work
13:44:32 <stickster> ¯\_(ツ)_/¯
13:44:33 <aruiz> I'll have to ping Hans about it
13:44:35 <otaylor> I'd like to see hans comment on that ticket
13:44:46 <juhp> halfline wrote in the upstream ticket https://gitlab.freedesktop.org/plymouth/plymouth/issues/68
13:44:48 <halfline> i already pinged him on the upstream ticket linked to it
13:45:08 <aruiz> halfline, do we know which machines are causing it?
13:45:13 <halfline> this is the 800x600 turn it off issue, i mentioned a few minutes ago
13:45:22 <aruiz> ah!
13:45:23 <aruiz> gotcha
13:45:30 <halfline> aruiz: i assume it's primarily workstations
13:45:48 <halfline> laptops will probably be hard coded to use the panel's native resolution
13:46:07 * stickster thinks there's nothing for the workstation WG to explicitly "decide" here, frankly.
13:46:10 <halfline> i guess firmware is just not doing ddc on the monitor to query it's preferred mode or osmething like that
13:46:24 <aruiz> I'll follow up with Hans, I wasn't aware of this one
13:46:24 <ryanlerch> the ticket is asking us to vote on reverting the change that caused this
13:46:35 <aruiz> ugh
13:46:36 <otaylor> halfline: or they didn't bother to make the logo screen work in multiple resolutions
13:46:59 <juhp> Can we defer to next meeting?
13:47:06 <halfline> otaylor: yea, thoug mcatanzaro sees it even if the logo is disabled
13:47:07 <otaylor> proposal late halfline and hans figure out the best course
13:47:13 <otaylor> *let*
13:47:14 <halfline> and its just bios messages
13:47:29 <stickster> I don't think the WG should be in the habit of reverting changes on the basis of a bug with uncertain impact
13:48:07 <aruiz> also, compromising the defaults because there's broken firmware out there is a bad plan too
13:48:08 * stickster +1 on otaylor proposal, we can wait & see on the right upstream fix
13:48:16 <aruiz> I'll ask hans to follow up
13:48:22 <ryanlerch> +1 on wait-n-see
13:48:26 <cschalle_> +1 on otaylor proposal for me too
13:48:28 <juhp> Yeah I am hoping this may just get fixed
13:48:30 <juhp> +1
13:48:58 <halfline> aruiz: yea, though it should be detectable if the firmware left the monitor in a non-preferred mode
13:49:02 <juhp> If there is still a serious problem we can revisit
13:49:06 <halfline> and turn it off in that case only
13:49:12 <aruiz> halfline, yep
13:49:33 <aruiz> I'm sure hans can take care of this, specially if mcatanzaro can assist with access to the hardware
13:49:56 <juhp> aruiz++
13:49:57 <zodbot> juhp: Karma for aruiz changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
13:50:26 <ryanlerch> #topic Workstation minimum system requirements on getfedora.org https://pagure.io/fedora-workstation/issue/76
13:50:48 <ryanlerch> okies, that seems sorted, on to the next...
13:51:33 <ryanlerch> this is for changing the suggested minimums system requitements for Workstation on getfedora.org
13:51:57 <stickster> I haven't tried Workstation on a 2GB RAM system that's not a virt system, but I can't imagine staying with a 1 GB minimum is sensible at all.
13:52:39 <juhp> yeah I stopped using 1GB VMs long ago
13:53:40 * stickster leaves +1 vote in the ticket
13:54:12 <ryanlerch> yeah, i'm +1 to upping this to 2GB minimum, and 20GB disk space
13:54:12 <stickster> anyone else?
13:54:22 <juhp> Disk space I am less sure about
13:54:40 <juhp> 20GB is certainly more than enough
13:54:46 <juhp> 15GB is too
13:54:53 <stickster> I think 20GB is probably the next usable round number.
13:55:07 <ryanlerch> cschalle_: juhp: kalev: otaylor: mclasen:
13:55:27 * mclasen walks in late, apologies
13:55:31 <juhp> Does it have to be round? :)
13:55:40 * stickster has to leave to moderate another meeting.
13:55:41 <cschalle_> +1 from me too
13:55:51 <stickster> juhp: of course, just like the spherical cow ;-)
13:55:58 <juhp> haha
13:56:04 <otaylor> +1
13:56:22 <juhp> well 20GB is not minimum but if we want to write that then
13:57:24 <ryanlerch> juhp: i think these are "reccomended" minimums really
13:57:28 <halfline> 20 is the default virt-manager allocates right? so it's probably pretty well tested
13:57:29 <otaylor> juhp: minimum is hard to define - in the end you need space to actually *do* something
13:57:41 <juhp> indeed
13:58:46 <aday> shouldn't the minimum be reflected in the installer?
13:59:00 <juhp> I would suggest 15GB
13:59:27 <ryanlerch> okies, almost at time...
13:59:45 <kalev> +1 from me too, sorry
14:00:19 <ryanlerch> lost count, but i think that is agreed?
14:00:56 <ryanlerch> anyone have anything else to add?
14:01:07 <ryanlerch> 5
14:01:14 <ryanlerch> 4
14:01:18 <ryanlerch> 3
14:01:21 <ryanlerch> 2
14:01:25 <ryanlerch> 1
14:01:29 <ryanlerch> 0
14:01:33 <ryanlerch> #endmeeting