04:07:14 <cmurf> #startmeeting Workstation WG (2020-06-23)
04:07:14 <zodbot> Meeting started Thu Jun 25 04:07:14 2020 UTC.
04:07:14 <zodbot> This meeting is logged and archived in a public location.
04:07:14 <zodbot> The chair is cmurf. Information about MeetBot at http://wiki.debian.org/MeetBot.
04:07:14 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
04:07:14 <zodbot> The meeting name has been set to 'workstation_wg_(2020-06-23)'
04:07:16 <cmurf> #meetingname workstation
04:07:16 <zodbot> The meeting name has been set to 'workstation'
04:07:18 <cmurf> #chair cmurf
04:07:18 <zodbot> Current chairs: cmurf
04:07:45 <cmurf> #topic Rollcall
04:07:46 <cmurf> #info present: cmurf, aday, jens, felipe, langdon, neal, michael, langdon, mclasen, kalev, josef
04:07:48 <cmurf> #topic Minutes for 9 Jun meeting
04:07:50 <cmurf> #link https://meetbot.fedoraproject.org/teams/workstation/workstation.2020-06-09-23.32.log.html
04:07:52 <cmurf> #agreed No objections - approved
04:07:54 <cmurf> #topic Minutes for 16 Jun meeting
04:07:56 <cmurf> #link https://meetbot.fedoraproject.org/teams/workstation/workstation.2020-06-18-16.49.log.html
04:07:58 <cmurf> #info These only just made it to the mailing list, so it's deferred until next week.
04:08:00 <cmurf> #topic Announcements
04:08:02 <cmurf> #info None
04:08:04 <cmurf> #topic Follow-ups and status reports
04:08:06 <cmurf> #info Chris: started a draft nano proposal
04:08:08 <cmurf> Needs help from people who understand what the mechanism for the default editor is.
04:08:10 <cmurf> Neal: right now we don't have one.
04:08:12 <cmurf> Langdon: Is this the editor environment variable?
04:08:14 <cmurf> Michael: there are 2 methods, the environment variable and /usr/bin/editor which is Debian specific
04:08:16 <cmurf> Matthias: what's the UX goal?
04:08:18 <cmurf> Michael: you get nano when you do Git commit.
04:08:20 <cmurf> Langdon: what are the ramifications? Nano would end up in a bunch of places - in containers?
04:08:22 <cmurf> Chris: asks for consent on the change - kalev objects
04:08:24 <cmurf> Michael: suggests we just change the env variable rather than /usr/bin
04:08:26 <cmurf> Allan: flathub whitelist is waiting on council. some confusion over who should be dealing with it
04:08:28 <cmurf> (Jens: can we not call it whitelist?)
04:08:30 <cmurf> #topic Guidelines for preinstalled and non-removable apps
04:08:32 <cmurf> #link https://pagure.io/fedora-workstation/issue/125
04:08:34 <cmurf> #info Not ready yet - deferred
04:08:36 <cmurf> #topic Default partitioning: solving the competition for free space
04:08:38 <cmurf> #link https://pagure.io/fedora-workstation/issue/152
04:08:40 <cmurf> #info Main questions: are there other ideas/proposals forthcoming? Change
04:08:42 <cmurf> proposal deadline is June 30.
04:08:44 <cmurf> #agreed LVM thin partitioning, split partitions for /usr & /var, /etc & /home, not considered to be realistic. The choice is therefore between btrfs, xfs, ext4.
04:08:46 <cmurf> Michael: doesn't think that switching to XFS makes sense.
04:08:48 <cmurf> Owen: XFS is the default for other spins, so there is an argument for it
04:08:50 <cmurf> No shrinking is an issue, but an edge case. It's a legitimate option.
04:08:52 <cmurf> We don't need to whittle it down one by one.
04:08:54 <cmurf> Neal: shrinking can be an issue if someone wants to add Windows as a dual boot.
04:08:56 <cmurf> Owen: advantages of XFS - reflinks and consistency, more actively developed than ext4
04:08:58 <cmurf> Michael: any of the 3 options are better than what we have today
04:09:00 <cmurf> Chris: non-shrinkability of XFS would be an issue with homed and LUKS. With ext4, the user would have to log out then do an offline resize. system-homed could imply btrfs.
04:09:02 <cmurf> Owen: not entirely sold on systemd-homed.
04:09:04 <cmurf> Langdon: is there no encryption in btrfs? Neal: it needs to be used in combination with LUKS or dmcrypt. Native encryption is planned for this year.
04:09:06 <cmurf> Langdon: not possible to preserve /home with btrfs? Neal: you can preserve it just fine. When you reinstall with Anaconda you can do custom partitioning and mount at /home and not format it.
04:09:08 <cmurf> Chris: can we ask FESCo for their perspective on btrfs?
04:09:10 <cmurf> Josef: in his experience, RHEL storage looks at upstream and RHEL, not Fedora
04:09:12 <cmurf> Langdon: would prefer to speak to Red Hat storage/management for their input. Doesn't think that it's a FESCo decision.
04:09:14 <cmurf> Owen: the decision needs to be coordinated with RHEL storage. Doesn't mean we can't go with btrfs, but people need to be informed about it.
04:09:16 <cmurf> Owen: we need to go back with more details on the proposal.
04:09:18 <cmurf> Next steps:
04:09:20 <cmurf> 1. Resolve RHEL questions
04:09:22 <cmurf> 2. Owner for a alternative change proposal - xfs/ext4
04:09:24 <cmurf> 3. Question for maybes on the btrfs question - what concerns about btrfs, what commitments would you like to see for btrfs?
04:09:26 <cmurf> Owen will take 1 forward.
04:09:28 <cmurf> Michael: question about the change proposal - will the btrfs change proposal be submitted for F33? Neal would like to. Langdon: we can submit the change and continue in parallel with other conversations.
04:09:30 <cmurf> Matthias: do we need a competing change? Can we combine the proposal: plain partition with btrfs by default or ext4 if you want it?
04:09:32 <cmurf> #action There will be a btrfs proposal before the weekend
04:09:34 <cmurf> #topic btrfs by default
04:09:36 <cmurf> #link https://pagure.io/fedora-workstation/issue/153
04:09:38 <cmurf> #info A rough draft proposal is in the ticket.
04:09:40 <cmurf> What questions/concerns/commitments does the WG have for the proposal's owners? And does the WG have questions it would like to pose to FESCo?
04:09:53 <cmurf> #endmeeting