13:09:27 #startmeeting Workstation WG 13:09:27 Meeting started Mon Apr 8 13:09:27 2019 UTC. 13:09:27 This meeting is logged and archived in a public location. 13:09:27 The chair is mclasen. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:09:27 Useful Commands: #action #agreed #halp #info #idea #link #topic. 13:09:27 The meeting name has been set to 'workstation_wg' 13:09:29 yeah, sorry, I am juggling 3 things atm 13:09:36 #meetingname workstation 13:09:36 The meeting name has been set to 'workstation' 13:09:50 #topic Roll call 13:09:55 .hello mclasen 13:09:56 mclasen: mclasen 'Matthias Clasen' 13:10:02 .hello petersen 13:10:03 juhp: petersen 'Jens Petersen' 13:10:32 .hello otaylor 13:10:33 otaylor: otaylor 'Owen Taylor' 13:10:34 (sorry to be late!) 13:10:36 .hello kalev 13:10:37 kalev: kalev 'Kalev Lember' 13:11:24 thats 4 13:11:26 one more ? 13:12:19 .hello cschalle 13:12:20 cschalle: Sorry, but you don't exist 13:14:10 .hello catanzaro 13:14:11 mcatanzaro: catanzaro 'Michael Catanzaro' 13:14:30 .hello ryanlerch 13:14:31 ryanlerch: ryanlerch 'Ryan Lerch' 13:15:22 do we have items for the agenda ? 13:15:40 I added one 13:15:49 that came up in irc apparently 13:15:52 mclasen: Can you do a report on the Silverblue test day? 13:16:13 I can point to the results page, thats about it 13:16:18 but yes 13:16:35 juhp: where did you add it ? 13:16:49 To the meeting tag 13:17:21 https://pagure.io/fedora-workstation/issue/54 13:18:15 #fedora-workstation: Any chance of https://pagure.io/fedora-workstation/issue/54 being tagged for tomorrow's meeting? The most immediately relevant point of that is the last comment, about just giving root a +10G bump for Fedora 30. 13:18:49 ok, lets start with that 13:19:01 #topic "Default disk partitioning layout for workstation" 13:19:19 Is it still possible for f30? 13:19:47 should be possible to change this week. next week starts the final freeze 13:19:50 juhp: I think that depends entirely on who you ask 13:20:08 okay 13:20:36 I'm pretty sure we can get 50G => 60G for F30, going to no-home-by-default harder 13:21:23 I have no clue where the defaults are set though. Somewhere in Anaconda, I guess? 13:21:53 anaconda team should know :) 13:22:30 looks like it's, pyanaconda/storage/partitioning.py in anaconda 13:23:11 Should we open an rfe in bz then? 13:23:11 do the numbers make sense ? 13:23:20 https://github.com/rhinstaller/anaconda/blob/f30-devel/pyanaconda/storage/partitioning.py#L48 13:23:20 or a PR? 13:23:38 if 50G was arbitrary, why go to 60, and not 100 ? 13:23:49 * mcatanzaro not convinced +10 GB is useful 13:24:33 If we stick with separate / then we simply need more than that 13:24:56 We seemed pretty determined to get rid of separate / last time we discussed this (of course can't do that a week before final freeze) 13:25:41 Yeah we really should 13:25:46 for f31? 13:26:15 mcatanzaro: what do you think the minimal increase is that is useful? 13:26:37 mclasen: I think the 50 GB limit is there so that / wouldn't end up taking up all of the space, and instead leave more space to /home 13:26:51 Specially with small discs (cough ssd) no /home is necessary I think 13:27:07 (I think everybody agrees on the single /, except for people who want to put homedirs in loopback mounted encrypted files...) 13:27:18 otaylor: I don't know 13:28:12 Also virt-manager vs gnome-boxes 13:28:22 So set aside F30... for F31, we can we do single / right? I know we'e waiting to do the LUKS change at the same time? 13:28:45 what is "the luks change" again ? 13:29:36 LUKS enabled by default 13:30:18 We have a ticket, no? 13:30:32 can we ignore F31 for now and just concentrate on the F30 issue to get the fix out? 13:30:44 ok, so where do we stand on the f30 change - do we just vote on 50 -> 60 and see if we can get that implemented ? 13:30:51 https://pagure.io/fedora-workstation/issue/82 13:30:55 or is there an alternative proposal ? 13:31:01 I'll note that we have fedora flatpaks enabled by default in F30 so that / requirements are going to be larger than in just rpm-based system 13:31:21 let's just vote on the 50 -> 60 change 13:31:54 kalev: should that be 50 -> 70 ? 13:32:04 I don't known :) 13:32:06 I don't know :) 13:32:11 aha! the alternative proposal 13:32:19 how big are the runtimes for reference? 13:32:24 or 50 -> 75 ? 13:32:41 numerology... 13:33:02 org.fedoraproject.Platform/x86_64/f29 is 1 GB 13:33:11 thanks 13:33:44 15GB? (or lower bids allowed? ;o) 13:33:51 s/or/are/ 13:34:34 Hmm well maybe 70GB is good enough? 13:35:11 lets vote on 70GB first, and proceed to lower numbers if necessary 13:35:18 +1 on 70GB from me 13:35:19 +1 for 70GB 13:35:23 +1 13:35:27 +1 13:35:43 +1 13:36:22 any more votes ? this seems accepted 13:37:06 otaylor: by the way, not sure if you saw: https://bodhi.fedoraproject.org/updates/FEDORA-2019-7d91eea013 enables the fedora flatpak repos by default 13:37:17 kalev: very cool 13:37:35 sadly this missed the silverblue test day 13:37:37 ok. any volunteer to file an anaconda bug / ticket / pr / whatever ? 13:37:54 I can go do that now if you update the workstation ticket to say it's accepted 13:38:13 * kalev has anaconda git repo checked out. 13:38:56 thanks 13:39:14 #action kalev will file an anaconda ticket 13:39:23 I updated the workstation ticket 13:39:34 #topic Silverblue test day 13:40:05 to summarize, a Fedora testday on Silverblue happened on Friday 13:40:11 the results are here: http://testdays.fedorainfracloud.org/events/63 13:40:37 I started looking over it this morning, there doesn't seem to be any huge issue that jumps out at me 13:41:00 one recurring topic is issues with firefox and flatpak repo files getting downloaded as text files 13:42:24 anybody else have impressions from the test day ? 13:43:36 looks like some people tried to use pkcon for tests 13:44:01 filed to 70 GB pr: https://github.com/rhinstaller/anaconda/pull/1939 13:44:09 kalev: thanks! 13:44:42 mclasen: gnome-software regressed just before the test day and the rpm-ostree backend stopped working due to rpm-ostree changing the libdnf version it bundles 13:44:53 too bad 13:45:16 Seems like there is some issue with adding repos from the command line vs. gnome-software picking them up? 13:45:27 yep, I should look into that 13:45:51 Sems like people hit the firefox problem, tried to work around it, and still didn't see stuff 13:49:58 installing chrome seems to have worked for most people 13:51:18 ohh nice 13:51:37 * mclasen is moving some items along in the silverblue kanban board according to the test results 13:52:11 sadly the gnome-software backend was non-functional during the test day, otherwise it should be just the matter of double clicking on the rpm to install chrome 13:52:33 yeah, unfortunate 13:52:43 so this affected gnome-software, but not rpm-ostree itself ? 13:53:16 yeah 13:53:51 had to rebuild gnome-software against the same version of bundled libdnf that rpm-ostree switched to 13:54:54 * mclasen quietly rolls eyes 13:55:07 we have 5 minutes left - any other topics to raise ? 13:56:11 kalev: they can't share? 13:56:46 juhp: they can and they do, but rpm-ostree bumped the soname and that broke gnome-software 13:57:11 I mean, bumped the bundled libdnf soname :) 13:57:54 * kalev is not very happy of how this is set up right now. 13:58:02 okay right 13:58:54 OK, I need to head off. I sent out mail(s) about doing another attempt at finding a new meeting time. Please add your info there, and we'll see if something shows up or we need to find someone to fill in for stickster 13:59:55 #endmeeting