13:00:14 #startmeeting Workstation WG 13:00:14 Meeting started Mon Mar 13 13:00:14 2017 UTC. The chair is stickster. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:14 Useful Commands: #action #agreed #halp #info #idea #link #topic. 13:00:14 The meeting name has been set to 'workstation_wg' 13:00:17 #meetingname workstation 13:00:17 The meeting name has been set to 'workstation' 13:00:19 #topic Roll call 13:00:22 .hello pfrields 13:00:23 stickster: pfrields 'Paul W. Frields' 13:00:48 cschalle: mclasen: mcatanzaro: rdieter_work: ryanlerch: o/ 13:01:00 .hello mclasen 13:01:01 mclasen: mclasen 'Matthias Clasen' 13:01:07 .hello cschalle 13:01:07 cschalle: Sorry, but you don't exist 13:01:26 * stickster looks for otaylor juhp kalev 13:01:34 .fas schaller 13:01:35 stickster: uraeus 'Christian Schaller' - tobiasschaller 'Tobias Schaller' 13:01:44 ^ .hello uraeus should work 13:02:56 hi 13:02:56 #chair cschalle mclasen juhp mclasen mcatanzaro rdieter_work ryanlerch 13:02:56 Current chairs: cschalle juhp mcatanzaro mclasen rdieter_work ryanlerch stickster 13:03:42 cschalle: Was Owen away this week? ISTR something hazily about that 13:04:36 stickster, not sure 13:05:20 * stickster notices we only seem to have 4 of us here, but let's see if someone shows up late 13:05:51 #info Late breaking question on list for blocker bug 13:06:05 #topic BZ 1405539 13:06:07 #link https://bugzilla.redhat.com/show_bug.cgi?id=1405539 13:06:56 mclasen: Does this one fall into your area? QE is looking for an opinion on whether this is a blocker, and there seems to be some disagreement about that 13:07:14 I'll ask halfline about it when he comes in 13:08:01 I think this one looks serious, but I'm not sure we would expect enough incidence to block Alpha for this... perhaps we could fix, and require it to block Beta 13:08:36 I have been looking at it - I doubt it would be an Alpha blocker though 13:08:41 I have no idea how often people install and then later change primary keymap 13:09:07 I don't believe this is a new issue though ? 13:09:20 more in the 'it has always been this crappy' category ? 13:09:22 mclasen: it's not new, that's correct 13:09:25 right to both of you 13:09:29 how does it become a blocker then ? 13:09:49 mclasen: but in the past g-c-c allowed you to fix both keymaps (user + system) and it doesn't seem it works the same way now 13:10:06 definitely GOODTOFIX though IMO 13:10:08 that sounds like something _is_ new after all ? 13:10:25 I don't know when g-c-c simplified that bit 13:10:57 * stickster doesn't do much with keymaps either, and would yield opinion to e.g. juhp here 13:11:06 one suggestion made was that plymouth should show the system keyboard layout in use when prompting for the passphrase 13:11:50 .... text 13:11:53 -> fonts 13:11:56 -> pain 13:12:34 not sure if it is possible currently to set up system layout such that one can no longer type the original passphrase anymore? 13:12:40 mclasen: :) 13:12:41 I think all of this is more or less a lost cause, unless we get something like kmscon 13:13:03 until then, your best hope is to avoid all user interaction before gdm 13:14:02 I thought kmscon itself was a lost cause? 13:14:08 yes 13:14:44 I am not really hoping for a bulletproof fix but at least a way for user to be aware that something changed wrt to their initramfs layout 13:14:48 fading hopes 13:15:39 nothing but bandaid fixes until retirement 13:15:46 * mclasen not feeling generous this morning 13:16:09 It would be nice to tell the user something that helps them avoid the pain, since it's almost impossible to diagnose when it will bite them 13:16:23 w can't tell the user anything, because ... fonts 13:16:36 Well, in g-c-c we could, like "be careful changing your primary keymap if you have disk encryption" 13:16:53 mclasen: is fonts really the primary issue? 13:17:03 stickster: right 13:17:14 well, you need a whole font rendering stack too 13:17:31 at that point ,your initramfs rapidly starts resembling a full os 13:17:48 *nod 13:17:49 mclasen: ah currently no font in plymouth you mean? 13:17:50 thats the third one on the way up, after the firmware and the the bootloader 13:18:07 .hello catanzaro 13:18:08 mcatanzaro: catanzaro 'Michael Catanzaro' 13:18:44 It's a really nasty bug. :( I think we need more opinions from developers who are experienced with plymouth/initramfs/etc. 13:18:51 at least g-c-c could warn/tell users that they are also changing system layout, no? 13:19:08 Please don't make it an alpha blocker. 13:19:15 mcatanzaro: yeah 13:19:26 mcatanzaro: I doubt it would meet the criteria 13:19:27 juhp: If we can't fix anything else, that would at least make sense -- but I'm also -1 Alpha blocker fwiw 13:19:29 yes, control-center could warn 13:20:00 stickster: me too 13:20:09 I do think it should be a final release blocker. It is severe. Currently preexisting issues are allowed to be release blockers, it's how we manage to clean up nasty longstanding bugs like this. 13:20:26 mcatanzaro: agreed 13:21:18 mclasen: did I understand that you're also -1 Alpha blocker? Seemed pretty much that way 13:21:28 we never manage to clean them up, unfortunately 13:21:44 we only put another bandaid on top, to hold the previous one in place 13:21:51 "clean up" in this case is more like, "put a mop and a bucket in the hall for general use" 13:22:05 Regarding control-center. Warning that it is going to change your system keyboard layout doesn't make a ton of sense, since the whole point of not exposing the difference between system and user keyboard layout in the Region & Language panel is to remove the totally unnecessary distinction between system/user layout on single-user systems. 13:22:47 Warning that it can change your keyboard layout for disk encryption makes a bit more sense, but it doesn't make any sense to say "it can change your keyboard layout for disk encryption at some undetermined point in the future. you will receive no further warning when this happens." 13:23:51 mcatanzaro: short of wishing for a new console system that doesn't seem to be coming, do you have a better suggestion? 13:24:19 stickster: Not really. It's a thorny problem that probably needs to be discussed outside of this meeting; we don't have the right people here. It's also not a typical blocker because we don't agree on what the solution should be. But I do want it to be treated as a data loss bug. It's unreasonable to expect users to figure out their keyboard layout has changed without displaying keyboard layout to them. 13:24:25 *nod 13:24:47 And you know, if we have to put font rendering stack into the initramfs to make this work properly, then maybe that's what we have to do.... 13:24:49 right 13:25:15 mclasen: Shall I put this #action on you to discuss w/ plymouth + other folk? 13:25:20 OTOH I do recognize that it's rather rich to say "we can't release until this is fixed" without proposing how to fix it. :) 13:25:28 feel free to, I'll discuss it anyway 13:25:44 mcatanzaro: yep, it's disingenuous, so blocking on it is just spinning wheels for no reason 13:26:25 this also feels like the kind of issue that juhp and his team probably should own 13:26:29 #action mclasen discuss this bug with plymouth + other devs to see if we have some reasonable way of mitigating or alerting 13:26:34 stickster: thanks for bringing it up - I had been thinking to raise it tonight too 13:27:23 cschalle: yeah if we can help we would be happy to assist 13:27:54 #agreed juhp team to assist and possibly own (TBD) 13:27:58 Let's move on 13:28:10 #topic Workstation branding 13:29:23 So this is something that Matthias had pulled together into a wiki page, but then neither of us was really able to follow up... IIUC the next thing that should happen is pulling a couple design folks together to lay out objectives 13:30:07 ryanlerch from Fedora design can be one of the people involved in this, and I've talked to him about it in terms of high-level goals 13:30:26 #link https://fedoraproject.org/wiki/Workstation/Branding 13:32:15 mclasen: is there someone else from desktop who's the right person to get involved as well? 13:32:30 like aday or someone else? 13:33:09 * mclasen wishes link highlighting worked in polari 13:33:31 oh bummer... I'm getting it through vte because I'm on old school irssi 13:34:10 stickster, mclasen : lets assume aday, but we can have a follow up discussion afterwards 13:34:17 Which reminds me, ryanlerch still has a WG seat, did you forget...? I'm not sure when he last attended a meeting; I know timezones are really hard there. (Hi ryanlerch. ;) 13:34:30 Yeah, this time sucks even worse for him than for juhp 13:34:44 That's why he wanted to pitch in here :-) 13:35:12 I think it is a bit unfair to tag aday with anything that has design in it. jimmac has done branding work in the past, too 13:35:24 EDT now makes it slightly better :) 13:35:36 * stickster has no preconceived person targeted, and is OK with it being TBD 13:36:07 #action stickster finalize a group and initial meetup time (either on IRC or some other venue) 13:36:24 ^ sound OK? 13:36:29 The wiki page looks like a fine start :) 13:36:31 Yup. It'd be nice to include one of the GNOME designers. 13:36:40 agreed 13:37:26 sounds ok 13:37:38 One other priority from the wiki page: WG ratifying the goals 13:37:56 1. When looking at someone using Fedora Workstation one should in the most common usecases be able to quickly identify that they are using Fedora Workstation. Obvious exceptions include fullscreen playback of a movie 13:38:03 2. Fedora Workstation branding should tie in with overall Fedora Branding efforts 13:38:09 3. Branding does not have to be explicit, working group is fine with branding elements that provide distinguishability, but which will need to be popularized before they provide identification as Fedora. 13:38:50 These seem pretty reasonable to me, and none of them point to a specific implementation detail (which I think is good... leave that to designers) 13:39:30 * mclasen doesn't really agree with the first one, but thats long-standing and well-known at this point... 13:40:01 I'm not sure if the designers will be able to achieve these goals. 13:40:31 mcatanzaro, well then they have to tell us :) and suggest revised goals 13:40:31 My concern with 3. is anything we do to distinguish Fedora from GNOME can just be replicated by other distros, unless it involves a Fedora logo. 13:40:53 mcatanzaro, why would other distros copy our visual features? 13:40:54 mcatanzaro: possibly, but why would they want to? 13:40:57 *jinx 13:41:09 we could take out a design patent :) 13:41:11 If what we've done is an improvement over upstream, why wouldn't they? 13:41:50 *shrug, it would be odd I think... and presumes some level 13:41:52 I'm fine with sending the goals to designers to see what they come up with, of course. 13:41:52 oops 13:41:59 mcatanzaro, I don't really see this things as improvements, its kinda like having a car in multiple colours, is one really better than the other? 13:42:21 *shrug, it would be odd I think... and presumes some level of change has to be either improvement or regression 13:42:23 s/this/these/ 13:42:42 stickster, great minds thing alike :) 13:42:50 or not enough coffee 13:42:54 heh 13:43:19 Anyhoo, I'm willing to move on after committing to do something here, but don't want to shut off further discussion :-) 13:43:19 a great example of 'branding' gone wrong is the matrix clients - I'm sure the design agency they had on contract thought the mint green was an awesome idea 13:44:22 and its just a color, so it can't be a regression... 13:44:39 could be worse, neon green! 13:45:15 * stickster moves on 13:45:21 #topic Atomic Host based Workstation 13:46:17 So as for where we are here... I'm a bit lost from the long discussion on the list, and was hoping otaylor would be here to pull some next actions out 13:47:19 I agree that we want to have builds happening in Fedora more rapidly, so that Fedora can be a better suited home for the effort overall 13:48:37 stickster, owen could be away due to the birth of their second child 13:48:50 cschalle: Oh, I thought that was a while off! 13:48:59 my next step is to get kalev freed up and having him take ownership of this 13:50:02 There is a bunch of work going on in the Fedora Engineering team as well, toward a real continuous integration (CI) effort, which would provide us things like e.g. better Rawhide 13:50:47 mclasen: ^ so given that I think there should be a lot of openness to working with kalev to encompass a faster Workstation generation too 13:50:54 I don't see much point in discussing this topic without otaylor present. 13:51:22 Or, OK, if owen is out for a while and kalev is going to be taking over, nevermind then. 13:52:16 I guess the point here is just informational for the moment -- but I think we should be thinking about how to integrate this effort with the overall release efforts, as opposed to doing a separate implementation and then trying to land it 13:54:08 yea 13:54:10 IOW, how can we make sure building the AH based Workstation happens as quickly as we need it, as part of the standard build mechanics that we are trying to make happen for other key artifacts 13:54:57 If that's fair, we can probably agree to defer to more discussion w/ kalev & the rest of the team 13:55:25 well, even if it's not fair, we can do that :-) since I need to switch over to my next meeting :-) 13:55:51 anything else on this? if not, will end in 1 minute 13:56:00 sounds good 13:56:04 I got another meeting too 13:57:11 OK, thanks for coming everyone! 13:57:14 #endmeeting