13:01:48 <petersen> #startmeeting Fedora Workstation WG
13:01:48 <zodbot> Meeting started Mon Jul 29 13:01:48 2019 UTC.
13:01:48 <zodbot> This meeting is logged and archived in a public location.
13:01:48 <zodbot> The chair is petersen. Information about MeetBot at
13:01:48 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
13:01:48 <zodbot> The meeting name has been set to 'fedora_workstation_wg'
13:02:01 <petersen> #meetingname workstation
13:02:01 <zodbot> The meeting name has been set to 'workstation'
13:02:22 <petersen> #topic Roll call
13:02:24 <langdon> .hello2
13:02:25 <zodbot> langdon: langdon 'Langdon White' <>
13:02:33 <petersen> Welcome
13:02:38 <mcatanzaro> .hello catanzaro
13:02:41 <zodbot> mcatanzaro: catanzaro 'Michael Catanzaro' <>
13:02:43 <petersen> .hello2
13:02:44 <zodbot> petersen: petersen 'Jens Petersen' <>
13:02:51 <cmurf> .hello chrismurphy
13:02:52 <zodbot> cmurf: chrismurphy 'Chris Murphy' <>
13:03:24 <petersen> #chair langdon mcatanzaro cmurf
13:03:24 <zodbot> Current chairs: cmurf langdon mcatanzaro petersen
13:03:51 <petersen> of course I just noticed 10min that I am chairperson today...
13:03:54 <petersen> ago
13:04:03 <mclasen> .hello mclasen
13:04:04 <zodbot> mclasen: mclasen 'Matthias Clasen' <>
13:04:17 <petersen> #chair mclasen
13:04:17 <zodbot> Current chairs: cmurf langdon mcatanzaro mclasen petersen
13:04:54 <petersen> That's 5 at least - anyone else present?
13:05:20 <mclasen> cschalle_: ?
13:05:31 * cschalle_ hi
13:05:39 <petersen> #chair cschalle_
13:05:39 <zodbot> Current chairs: cmurf cschalle_ langdon mcatanzaro mclasen petersen
13:05:53 <otaylor> .hello2
13:05:53 * aday listening in
13:05:54 <zodbot> otaylor: otaylor 'Owen Taylor' <>
13:06:16 <mcatanzaro> Welcome back otaylor!
13:06:37 <petersen> #chair otaylor
13:06:37 <zodbot> Current chairs: cmurf cschalle_ langdon mcatanzaro mclasen otaylor petersen
13:07:54 <petersen>
13:07:55 <mcatanzaro> Almost perfect attendance, just missing kalev
13:08:27 <otaylor> I think kalev is on vacation currently
13:09:02 <petersen> #topic (#100) Fill open WG slot
13:09:07 <petersen>
13:10:50 <langdon> well... given the challenge w/ 9, i definitely don't think we should shoot for 11 ;)
13:11:26 <petersen> I updated the membership list
13:11:44 <petersen> Well we could do it progressively....
13:12:08 <petersen> I guess we don't have to rush in 3 new members :)
13:12:24 <petersen> Any thoughts, suggestions?
13:12:25 <mcatanzaro> otaylor: You had previously mentioned Stephen Snow was interested in WG membership, but when discussed a couple meetings ago we discovered nobody knew him :) so decided to hold off
13:13:26 <mcatanzaro> Phil Wyatt had also expressed interest in joining, but he doesn't have a very long record of contributions (though I think that's mostly our fault; he stopped contributing because he didn't have commit access and his MRs were consistently ignored for no fault of his own)
13:13:30 <mcatanzaro> Do we have any other candidates?
13:14:03 <aday> i could be interested. good to have a ux person on board
13:14:41 <aday> it'd be good to have a chat with someone about responsibilities
13:14:57 <otaylor> Geoffrey Marr was the other person who expressed interest on the list (Red Hatter, Fedora QA)
13:15:10 <mcatanzaro> OK, three candidates
13:15:14 <mcatanzaro> Four
13:15:16 <mcatanzaro> :)
13:15:17 <petersen> 4!
13:15:44 <petersen> Perfect takes us to 12! haha
13:15:52 <otaylor> I suggested Stephen Snow because he was very active on discourse for silverblue, has been less active recently, but summer, etc.
13:16:38 <otaylor> clearly (from my persective) aday would be wonderful to have as a regular part of the group, rather than just a frequent invited guest, the main question is how serious we are about increasing the number of non-Red Hat people
13:16:38 <petersen> I think one suggestion was to have people join in a few meetings before committing
13:16:54 <cmurf> petersen: +1
13:17:21 <cmurf> I have someone in mind but haven't asked yet if they're even interested.
13:17:49 <mcatanzaro> "I think one suggestion was to have people join in a few meetings before committing" seems like a good idea
13:18:51 <otaylor> aday: (and anybody else) - responsibilities are - show up at biweekly meetings, take some action items and follow up, watch any tickets filed against pagure/fedora-workstation and follow up, and suggest topics you think that need covering for the meeting by filing tcikets against pagure/fedora-workstation
13:18:53 <cmurf> But he has a long list of contributions to Fedora, and is also involved in/familiar with other distribution practices. He'd help the WG stay diverse anyway.
13:19:13 <mcatanzaro> cmurf: Well first step is to ask him!
13:19:19 <petersen> Sounds good
13:19:25 <cmurf> mcatanzaro: I think so.
13:19:36 <cmurf> I didn't realize there were still positions to fill.
13:19:39 <aday> thanks otaylor
13:20:12 <mcatanzaro> I guess we can invite interested participants to attend the next meeting and just see who shows up to narrow the pool a bit. Extra credit to aday for being here proactively.
13:20:42 <cmurf> aday was here last meeting also, double extra credit :D
13:21:01 <petersen> Yeah we can invite them to join one or more of the next fews of meeting
13:21:09 <aday> i generally sit in; whether i have anything to contribute is another question :)
13:21:30 <petersen> aday++
13:21:31 <zodbot> petersen: Karma for aday changed to 1 (for the current release cycle):
13:21:35 <mcatanzaro> FWIW I agree having aday on permanently for his design input would be wonderful. Also agree it'd be nice to have more non-RH representation though. Anyway, if we have more than one really strong candidate like aday we can always increase more than one position.
13:21:42 <mclasen> yes, kalev is on vacation this week
13:21:58 * mclasen got pulled away here, sorry
13:21:58 <mcatanzaro> Who wants to send the invites?
13:22:01 <petersen> mcatanzaro: +1
13:22:56 <petersen> I am rather swamped right now
13:23:56 <cschalle> +1
13:23:59 <cmurf> I'm unconcerned about even/odd membership. Ties are rare. And a tie breaker by one is hardly consensus and we probably ought to shoot for consensus whenever possible.
13:24:07 <petersen> mcatanzaro: are you able to assist? :)
13:24:42 <mcatanzaro> cmurf: Agree
13:25:12 <petersen> Yeah if we can get 2 more I think is good
13:25:19 <mcatanzaro> I can find an email for Phil but I'd need contact details for Geoffrey Marr and Stephen Snow
13:25:38 <mcatanzaro> cmurf you can invite the person you are thinking of, of course. And aday already knows.
13:25:46 <otaylor> mcatanzaro: will pm
13:25:49 <cmurf> mcatanzaro: yes
13:25:57 <cmurf> I'm working on it now actually
13:26:12 <mcatanzaro> #action mcatanzaro to invite Phil, Geoffrey, and Stephen to the next WG meeting
13:26:24 <petersen> mcatanzaro: thanks a lot!
13:26:37 <mcatanzaro> #action otaylor to provide mcatanzaro with contact details ;)
13:26:53 <mcatanzaro> (emails I guess would be best)
13:27:05 <petersen> I think so
13:27:23 <petersen> #topic (#99) Installer proposed change: Simply reclaim disk space in Anaconda
13:27:37 <petersen>
13:28:25 <cmurf> This is mostly an FYI, this is happening, do we have an opinion on it?
13:28:45 <petersen> Any concerns for WS?
13:28:49 <cmurf> sgallagh has indicated on devel@ he will not vote approval for it without mockups
13:28:55 * mclasen watches failing to load
13:29:06 <cmurf> So we can punt. We meet next on August 12, correct?
13:29:16 <mcatanzaro> Aug 12, yes
13:29:40 * Son_Goku waves
13:29:52 <Son_Goku> .hello ngompa
13:29:53 <zodbot> Son_Goku: ngompa 'Neal Gompa' <>
13:30:07 <otaylor> well, if we have feedback, we should proivide it to fesco
13:30:14 * mclasen still not seeing the ticket
13:30:14 <cmurf> Hi Son_Goku
13:30:27 <cmurf> Looks like Neal Gompa got my email invitation already.
13:30:29 <petersen> Yeah but Aug 12 is right after Flock...
13:30:35 <langdon> im a bit of a toss up for the 12th.. end of flock and ill be in eu still..
13:30:37 <mcatanzaro> I'm torn here, on the one hand the current dialog could be confusing for nontechnical users and simplifying as far as possible is great. On the other hand, seeing what is going to happen before it happens seems pretty important... especially for partitioning which can screw up your system. Hi Son_Goku ;)
13:30:37 <otaylor> mclasen: there is some particular problem with you and pagure, it's weird...
13:30:46 <Son_Goku> cmurf, hey
13:31:02 <mclasen> naturally, I think it is all pagure's fault
13:31:46 <Son_Goku> I'm not particularly in favor of the anaconda proposed change
13:32:04 <cmurf> yeah I have a feeling with Flock ending on August 11th that there is a very good chance we will not have quorum on August 12 anyway
13:32:06 <petersen> cmurf: are those screenshots in a Drive folder?
13:32:10 <Son_Goku> mainly because I think we're simplifying it too much *and* losing the most common usage of the existing partitioning flows
13:32:12 <cmurf> petersen yes
13:32:18 <otaylor> I'm not in favor of punting non-technical users to the custom partitioning screen
13:32:18 <Son_Goku> cmurf, as it is, I'll be mid-flight on August 12
13:32:19 <petersen> cmurf: nod
13:32:22 <cmurf> pagure would not attach png or jpg
13:32:43 <mcatanzaro> Hm, having the meeting one day after Flock might not work well indeed. The next meeting after that is Aug 29, GUADEC, so that could be good or bad for attendance.
13:32:46 <Son_Goku> otaylor, we need the autopartitioner to handle Windows and macOS properly then
13:32:50 <petersen> Okay - just thought it might be easier to browse folder than files but anyway that okay
13:33:16 <mcatanzaro> We could just skip Aug 12 and return Aug 29, or shift by one week...
13:33:26 <Son_Goku> mcatanzaro, or shift by a day?!
13:33:27 <otaylor> who here is going to be attending flock?
13:33:35 * Son_Goku raises wing
13:33:38 <petersen> Yeah 19th could be better
13:33:43 <petersen> I will be there
13:34:09 * otaylor will not be there this year
13:34:16 * langdon will be at floxk
13:34:20 <langdon> flock even
13:34:37 <cmurf> or meet next week?
13:34:59 <cmurf> decent chance there will be a mockup this week, and fesco decides this friday or next
13:35:11 <petersen> aha
13:35:26 <cmurf> i don't see how they decide much later than that, beta freeze is august 29
13:35:38 <otaylor> fesco is back to meeting on mondays, btw
13:36:11 <petersen> Okay
13:36:12 <cmurf> another option is to discuss the mockup on desktop@, arrive at consensus on list, and add the WG "vote" if you will to the FESCo ticket
13:36:19 <Son_Goku> langdon, I guess I'll be seeing you quite a bit over the next two weeks...
13:36:26 <langdon> ha.. probably
13:36:38 * Son_Goku makes a note to sign up for desktop@
13:37:04 <cmurf> Son_Goku is my nominee for WG if that's not already clear.
13:37:45 <cmurf> anyway, without a mockup of this anaconda feature, some people won't be ready to form an opinion
13:38:01 <Son_Goku> we didn't already have a mockup? welp...
13:38:03 <cmurf> I'm on the fence, the existing UI is confusing for mortal users, so I can see why they want to get rid of it
13:38:24 <cmurf> but if they get rid of it, there is literally no UI for shrink in Custom partitioning
13:38:29 <cmurf> it's a leap of faith
13:38:36 <cmurf> it'll do it, but it's not obvious
13:38:37 <cschalle> I will be at flock
13:38:51 <Son_Goku> cmurf, well, sometimes that leap leads you to fall into the chasm...
13:38:52 <petersen> Cool
13:38:57 <cmurf> also in no case does Anaconda resize macOS volumes
13:38:58 <langdon> what mockup is missing give the mockup in the ticket?
13:38:59 <Son_Goku> and *that's* not obvious either
13:39:24 <cmurf> if there's a mockup in the ticket, I've missed it
13:39:41 <Son_Goku> cmurf, which is a bit unfortunate, given all the older mac hardware that Apple is abandoning in the next macOS release
13:39:51 <cmurf> yes
13:39:55 <langdon> ohh.. i think i may not have been reading it correctly.. like there is a mockup of part of the experience but maybe not all of it?
13:40:26 <mcatanzaro> I agree the anaconda custom partitioning is quite confusing.
13:40:42 <mcatanzaro> The UI that is being removed is a lot simpler and easier to use.
13:41:05 <cmurf> ahh ok you're referring to the WG ticket. Yeah those are not mockups those are today's UI in Anaconda and what's going away.
13:41:35 <petersen> I guess best to comment on the tracker ticket, when it is created
13:41:52 <cmurf> Even the UI being removed is very wordy and not at all obvious what to resize because so many other partitions that should definitely not be touched are presented for resize or deletion
13:41:58 <cmurf> petersen:
13:42:01 <cmurf> agreed
13:42:14 <mcatanzaro> Maybe we could say something like: the WG appreciates the desire to simplify storage configuration in anaconda, but has reservations about the proposed user experience change and encourages careful consideration by anaconda designers and FESCo.
13:43:31 <petersen> Or would it make sense to phrase the concerns as a question?
13:43:49 <langdon> i think the request for "what will the experience actually be" (e.g. mockups or walk throughs) are necessary
13:43:57 <petersen> Nod
13:43:59 <langdon> s/are/is
13:44:31 <petersen> Without more details it seems hard to make any statements about it
13:44:33 <cmurf> the feature proposal text tells us what the plan is, no simple resize/delete partition dialog, the dialog will instead ask the user for intent
13:44:52 <cmurf> and from that it will make a decision
13:45:19 <cmurf> but the expectation will be the user has already resized in some other tool; or they will be redirected from automatic partitioning to Custom partitioning
13:46:00 <mcatanzaro> BTW at risk of confusing the discussion with two topics at once: I've been thinking and I kinda think Son_Goku and aday are both much stronger candidates than the others who have been proposed, maybe would be better to just approve them and retract the proposal to invite the other candidates to join? To avoid putting on a sort of contest/competition unnecessarily...?
13:46:16 <cmurf> but i'm fine moving on to another issue, we're at +45
13:46:17 * mclasen can't see pagure, but thinks the direction is right - take partitioning out of the installer ghetto
13:46:47 <mcatanzaro> Of course the installer is the only place where partitioning can be done....
13:46:48 <langdon> cmurf: i saw that .. but i would hope it would also give hints.. like how much space is "use all available" .. and how many linux systems it thinks might be replaceing
13:47:21 <Son_Goku> langdon, where would it do that?
13:47:29 <petersen> I hope it is being discussed on devel list?
13:47:38 <cmurf> langdon: yeah maybe. giant walls of text in UI get my goat, and I think of translation effort, and confusion
13:47:38 <langdon> "this dialog with the following list of choices: "
13:47:39 <petersen> Shall we move on for now
13:47:47 <Son_Goku> mcatanzaro, I'm glad you have so much faith in me :P
13:47:50 <cmurf> petersen: yes let's move on
13:47:55 <mclasen> mcatanzaro: how so ?
13:48:31 <petersen> #topic (#98) Better interactivity in low-memory situations
13:48:40 <petersen>
13:48:52 <cmurf> oh dear... this one
13:49:54 <cmurf> I like the change. But the ducks are not in a row. I think we should vote to approve this work for Fedora 32 and that it should be a system-wide change.
13:49:58 <otaylor> cmurf: Isn't want hadess is working on inherently different than what you are working on
13:50:03 <cmurf> no
13:50:08 <cmurf> they conflict
13:50:18 <mcatanzaro> (Can we briefly go back to the WG membership topic again before we close the meeting? I know we're short on time.)
13:50:28 <otaylor> cmurf: that is, you were working on getting the live image to boot with zram/swap, while hadess was working on enabling it for the final solution
13:50:31 <cmurf> what I did was LiveOS only, to bring LiveOS in parity with netinstall - Anaconda activates swap on zram
13:50:32 <otaylor> system
13:50:41 <Son_Goku> yeah, I gotta leave soon (~10 min)
13:50:46 <mcatanzaro> cmurf: The implementation details conflict (my understanding), or the actual end goal?
13:51:32 <cmurf> what hadess is working on is a system level activation of swap on zram, and Anaconda no longer creating a swap volume on the installation target
13:51:33 <hadess> cmurf: there's a one-line change in anaconda so it doesn't conflict though
13:51:42 <petersen> mcatanzaro: I guess your suggestion is also okay - longer term maybe we should have a more formal process for membership though I don't know what that should be exactly
13:52:16 <cmurf> hadess: it does because anaconda tries to activate its own swap on zram on top of the already activated system one in your changes
13:52:37 <hadess> it does what?
13:52:44 <cmurf> so I get a bunch of log messages about /dev/zram0 can't be modified because it's already in use
13:53:31 <cmurf> anaconda's swap on zram isn't that smart it does no error checking, it assumes it can setup swap on the first zram device
13:53:57 <hadess> cmurf: i still don't understand why anaconda uses its own zram-swap service instead of the one in the zram package
13:54:13 <hadess> if it used the service in the zram package, then it would just work
13:54:18 <cmurf> that's beside the point and we have to take that up with anaconda folks to get them on board with a unified swap on zram solution
13:54:33 <hadess> why is it besides the point?
13:54:48 <petersen> Sounds like there is a good conversation to be had there?
13:55:00 <cmurf> because your work wasn't coordinated with mine
13:55:48 <hadess> so it's my fault that anaconda reinvented a zram-swap service?
13:55:50 <hadess> cool
13:56:01 <cmurf> and also your work isn't even done, you said last night that zram-generator isn't working, and yet you've submitted a PR to step on 99% of probinsons package
13:56:04 <cmurf> it's just weirdly done
13:56:28 <hadess> cmurf: which is marked as WIP in the code, pbrobinson knows that
13:56:32 <cmurf> hadess no one had a swap on zram service before anaconda, they've rolled their own and used it for several years
13:56:38 <mclasen> can we take the accusations out of this ?
13:56:47 * Son_Goku sighs
13:57:20 <hadess> mclasen: i'd seriously appreciate that as well, cmurf already said the same thing to me yesterday evening in the workstation channel
13:57:31 <mcatanzaro> OK well three minutes left
13:57:33 <cmurf> my complaint is these changes weren't discussed, they weren't tested, they do conflict.
13:57:46 <aday> let's focus on next steps
13:57:54 <aday> what needs to happen?
13:58:04 <mcatanzaro> I think cmurf and hadess need to further discuss this issue and coordinate a solution before bringing it back to the WG if necessary. :)
13:58:21 <mcatanzaro> OK? We agree leaving it broken is bad I'm sure.
13:58:26 <hadess> i think i'll let cmurf handle that one
13:58:46 <hadess> seeing as i only seem to make things worse, and i'll focus on upstream
13:59:02 <cmurf> I'm in favor of approving #98 for Fedora Workstation 32 as described.
13:59:09 <mcatanzaro> hadess: OK. It might be easiest to revert your changes as a first step and then cmurf can try again?
13:59:13 <Son_Goku> hadess, I don't think we can come to a reasonable solution if both of you don't work it out
13:59:22 <mcatanzaro> Oh yeah, approve #98 would be good... +1 from me.
14:00:04 <mcatanzaro> We are really out of time to start discussion on removing swap though, maybe defer to next week petersen?
14:00:14 <petersen> Alright maybe we can revisit this in the next meeting
14:00:16 <petersen> Yes
14:00:29 <cmurf> I always knew from the outset what I did was temporary for Fedora 31. If I had known anyone was interested in a broader solution, I would have worked on that effort and dumped my own.
14:00:30 <otaylor> I'm +1 on #98
14:00:32 <petersen> It sounds like a good F32 Change
14:00:51 <Son_Goku> though my voice doesn't matter, +1 for #98 for F32
14:00:51 <cmurf> But we do have to coordinate with anaconda folks, and make sure not creating a real swap  partition is sane for other use cases.
14:00:54 <cmurf> I'm not really sure it is.
14:00:58 <petersen> Okay please vote on the ticket
14:01:10 <Son_Goku> cmurf, and this means livecd-iso-to-disk needs fixing for f32 too :/
14:01:14 <petersen> I am also +1 for F32
14:01:33 <Son_Goku> not that I know anything about zram devices atm...
14:01:44 <petersen> I guess we need to make a hard stop?
14:01:47 <cmurf> Son_Goku: yeah maybe? anyway, another reason why I think this needs to go to FESCo as a system wide feature so all stake holders know about it.
14:02:01 <cmurf> can we have a meeting in a week?
14:02:01 <petersen> mcatanzaro: should we discuss your suggestion more?
14:02:13 <petersen> Maybe it is a good idea
14:02:48 <petersen> Any objections to moving the next meeting to the 5th August?
14:02:59 <Son_Goku> works for me
14:03:09 <mcatanzaro> Aw, three minutes late and three separate things to discuss :P
14:03:31 <langdon> i have a chance of being on a plane then too.. but i don't think it is till late eastern time
14:03:31 <petersen> Seems we have more things left to discuss
14:03:50 <cmurf> +1 meet in one week, same time, same place
14:04:10 <petersen> Okay I will send out minutes and mail about that
14:04:16 <petersen> Can someone update the calendar?
14:04:16 <mcatanzaro> #proposal Shift meeting schedule by one week (Aug 5 and Aug 19) to avoid problematic Aug 12 and Aug 26 dates
14:04:23 <petersen> +1
14:04:28 <cmurf> even better
14:04:28 <Son_Goku> +1
14:04:30 <cmurf> +1
14:05:14 <petersen> I feel we can go ahead with that plan as we already discussed about dates earlier
14:05:17 <otaylor> +1
14:05:29 <langdon> +1
14:05:47 <mcatanzaro> #proposal Defer earlier action item to proactively invite prospective WG candidates to join meetings since we have two strong candidates, but also consider whoever reads the minutes and joins at the next meeting
14:05:56 <petersen> Okay I think that is carried
14:06:08 <cmurf> +1
14:06:11 <petersen> +1
14:06:17 <langdon> +1
14:06:37 <mcatanzaro> #action mcatanzaro to adjust meeting schedule, all future meetings moved one week forward
14:07:02 <petersen> #agreed Shift meeting schedule by one week (Aug 5 and Aug 19) to avoid problematic Aug 12 and Aug 26 dates
14:07:13 <petersen> #undo
14:07:13 <zodbot> Removing item from minutes: AGREED by petersen at 14:07:02 : Shift meeting schedule by one week (Aug 5 and Aug 19) to avoid problematic Aug 12 and Aug 26 dates
14:07:44 <petersen> #agreed Shift meeting schedule by one week (Aug 5 and Aug 19) to avoid problematic Aug 12 and Aug 26 dates (+5)
14:08:06 <petersen> Okay that's wrap up here!
14:08:14 <petersen> Thanks everyone
14:08:21 <petersen> #endmeeting