17:00:53 #startmeeting F24-blocker-review 17:00:53 Meeting started Mon Feb 15 17:00:53 2016 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:53 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:00:53 The meeting name has been set to 'f24-blocker-review' 17:00:57 #meetingname F24-blocker-review 17:00:57 The meeting name has been set to 'f24-blocker-review' 17:01:01 #topic Roll Call 17:01:09 ahoyhoy folks! who's around for some quick blocker review? 17:01:20 only four bugs, shouldn't take long 17:01:37 * kparal is here 17:01:50 * pschindl is here 17:01:59 #chair kparal pschindl 17:01:59 Current chairs: adamw kparal pschindl 17:02:07 I'll do the secretarization. 17:02:10 thanks pschindl! 17:03:29 #topic Introduction 17:03:30 Why are we here? 17:03:30 #info Our purpose in this meeting is to review proposed blocker and nice-to-have bugs and decide whether to accept them, and to monitor the progress of fixing existing accepted blocker and nice-to-have bugs. 17:03:30 #info We'll be following the process outlined at: 17:03:30 #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 17:03:31 #info The bugs up for review today are available at: 17:03:35 #link http://qa.fedoraproject.org/blockerbugs/current 17:03:37 #info The criteria for release blocking bugs can be found at: 17:03:39 #link https://fedoraproject.org/wiki/Fedora_24_Alpha_Release_Criteria 17:03:41 #link https://fedoraproject.org/wiki/Fedora_24_Beta_Release_Criteria 17:03:43 #link https://fedoraproject.org/wiki/Fedora_24_Final_Release_Criteria 17:03:59 #info 3 Proposed Final Blockers 17:04:12 #info 2 Proposed Alpha Freeze Exceptions 17:05:40 starting with proposed final blockers: 17:05:41 #topic (1304681) Journal spam: Gdk-WARNING **: gdk-frame-clock: layout continuously requested, giving up after 4 tries 17:05:42 #link https://bugzilla.redhat.com/show_bug.cgi?id=1304681 17:05:42 #info Proposed Blocker, gnome-shell, NEW 17:07:21 this seems like kind of a difficult call 17:07:44 i'm probably -1 on the idea that log spam breaks 'basic functionality', the CPU usage is a bigger issue 17:08:51 I see this on my working laptop and it causes its fan running all the time. But I don't think it breaks any criterion. 17:09:28 i'm gonna propose we punt this on the basis it's a difficult call and we only have a few people today 17:09:47 sounds good to me 17:09:54 That seems right to me too 17:10:09 I'm afraid we're still not at that point where we can block for battery and fan noise issues 17:10:20 so I'd go with -1 I guess, but punt is ok too 17:10:49 kparal: i think there's a better argument if it's gonna happen to just about every install 17:11:12 that would be a good argument, yes 17:11:16 I think that lot of people would be quite disapointed if we release it with this bug. 17:11:59 proposed #agreed 1304681 - punt (delay decision) - this is a difficult judgment call and it's hard to make it with only a few (mostly QA) folks in attendance 17:12:08 pschindl: we don't have any criteria for disappointed people, yet ;) 17:12:24 ack 17:12:28 ack 17:12:42 #agreed 1304681 - punt (delay decision) - this is a difficult judgment call and it's hard to make it with only a few (mostly QA) folks in attendance 17:12:50 #topic (1033778) installer considers encrypted Apple Core Storage volumes as resizeable 17:12:50 #link https://bugzilla.redhat.com/show_bug.cgi?id=1033778 17:12:50 #info Proposed Blocker, python-blivet, NEW 17:13:48 adamw: btw have you noticed we have 3 proposed beta blockers as well? it was not mentioned in your summary text 17:14:39 kparal: er, really? i checked beta twice and didn't see anything 17:14:53 has cmurf been on a proposal spree again 17:14:54 adamw: they show up for me 17:14:57 i see 'em now, we'll add them 17:15:06 damnit, i need breakfast, this is supposed to be a SHORT meeting 17:15:14 so, this appears to be another cmurf vs. the anaconda devs bug 17:16:20 i'm kinda wary about wading in on it 17:16:46 -1. If anaconda doesn't try to resize it (and by the way destroy the disk), then it is just about wrong information. 17:17:07 it does try to resize it and destroy the disk, according to the report. 17:17:12 well, the partition. 17:17:25 "b. Installer always obliterates all data on the OS X PV when the user chooses the "shrink" option in Reclaim Space." 17:17:38 that's kinda bad 17:19:02 proposal: we ask the anaconda devs what they think 17:19:12 it certainly makes sense to disallow "shrink" for "unknown" filesystem 17:19:14 oh goody, i get to talk to the anaconda devs about cmurf again, that's always a fun time 17:19:32 kparal: yeah, that seems reasonable to me on the face of it, but i feel like there may be hidden depths here 17:19:37 hence why i want to ask the devs 17:19:44 please do, and let's punt this 17:19:50 I haven't managed to read the whole discussion 17:19:51 there's lots of stuff about GUIDs and partition types and lalala 17:20:47 proposed #agreed 1033778 - punt (delay decision) - on the face of it we're inclined to agree that this is a blocker, but we suspect there may be hidden depths, so we want to discuss with anaconda devs before making a decision (none present at meeting) 17:21:29 ack 17:21:40 ack 17:21:48 #agreed 1033778 - punt (delay decision) - on the face of it we're inclined to agree that this is a blocker, but we suspect there may be hidden depths, so we want to discuss with anaconda devs before making a decision (none present at meeting) 17:21:49 #topic (1306837) Anaconda's help is broken 17:21:49 #link https://bugzilla.redhat.com/show_bug.cgi?id=1306837 17:21:49 #info Proposed Blocker, yelp, NEW 17:22:22 sure, +1. not sure why the previous report got closed, but if it's still broken in current installer images, +1. 17:22:31 ack 17:22:33 +1 17:22:42 +1 17:23:06 proposed #agreed 1306837 - AcceptedBlocker (Final) - clear violation of "Any element in the installer interface(s) which is clearly intended to display 'help' text must do so correctly when activated." 17:23:26 ack 17:24:37 ack 17:24:49 #agreed 1306837 - AcceptedBlocker (Final) - clear violation of "Any element in the installer interface(s) which is clearly intended to display 'help' text must do so correctly when activated." 17:24:58 #topic (1306462) _ped.PartitionException: Unable to satisfy all constraints on the partition. 17:24:58 #link https://bugzilla.redhat.com/show_bug.cgi?id=1306462 17:24:58 #info Proposed Blocker, anaconda, NEW 17:26:04 hmm 17:26:10 "Crashes after accepting reclaim space" 17:26:20 the criterion does not apply to shrinking, it applies to a disk with literal empty unpartitioned space 17:26:46 "Reject or disallow invalid disk and volume configurations without crashing." might get it through, but i'd like a bit more detail 17:26:58 we need devs response first 17:28:24 punt and ping anaconda? 17:29:07 proposed #agreed 1306462 - punt (delay decision) - it's not clear precisely what the reporter did here (he cites the "unpartitioned space" criterion but refers to "reclaim space" also), thus we'd like more details and a developer response before making a determination 17:29:33 ack 17:29:43 ack 17:30:47 #agreed 1306462 - punt (delay decision) - it's not clear precisely what the reporter did here (he cites the "unpartitioned space" criterion but refers to "reclaim space" also), thus we'd like more details and a developer response before making a determination 17:31:02 oh, sorry, should have mentioned - we're on Beta blockers now. 17:31:07 that last one was proposed beta, so are the next two 17:31:08 topic (1306808) blivet.errors.FSError: mount failed: 32 17:31:08 #link https://bugzilla.redhat.com/show_bug.cgi?id=1306808 17:31:08 #info Proposed Blocker, anaconda, NEW 17:31:11 grrr 17:31:12 #undo 17:31:12 Removing item from minutes: INFO by adamw at 17:31:08 : Proposed Blocker, anaconda, NEW 17:31:16 #undo 17:31:16 Removing item from minutes: 17:31:23 #topic (1306808) blivet.errors.FSError: mount failed: 32 17:31:23 #link https://bugzilla.redhat.com/show_bug.cgi?id=1306808 17:31:23 #info Proposed Blocker, anaconda, NEW 17:31:55 this seems +1 17:32:51 eh 17:33:03 i wish cmurf would explain *exactly what he actually did* instead of leaving us to infer it 17:33:19 this seems like he created a completely empty btrfs volume outside of the installer then tried to install to it? 17:33:23 or something? he really doesn't explain 17:33:50 so let's ask for a proper reproducer and developer feedback? 17:34:48 the 'possible dupe' has a lot of other people hitting the same crash 17:35:03 so there's clearly something bad in that bit of blivet, but i'm not sure we understand exactly what it is yet 17:35:25 still, given the volume of people in 1253481 i'm actually inclined to +1 this on the 'reject invalid...without crashing' criterion 17:37:09 votes? +1 or punt? 17:37:13 punt 17:37:48 punt 17:38:28 ok then 17:39:05 proposed #agreed 1253481 - punt (delay decision) - this seems a strong candidate as there are many people hitting the 'potential dupe' bug, but reporter did not explain precisely what he actually did, we need that clarified to make a decision 17:39:23 ack 17:39:40 ack 17:39:48 #agreed 1253481 - punt (delay decision) - this seems a strong candidate as there are many people hitting the 'potential dupe' bug, but reporter did not explain precisely what he actually did, we need that clarified to make a decision 17:39:54 #topic (1259865) call `dnf mark install `on packages installed from PK 17:39:54 #link https://bugzilla.redhat.com/show_bug.cgi?id=1259865 17:39:54 #info Proposed Blocker, PackageKit, NEW 17:40:46 you might want to see comment 29-30 first for some trivial reproducer 17:41:09 and in comment 25 dnf developers are raising the priority for this 17:41:20 there's a lot of dupes for it 17:41:56 basically anything you update with packagekit might get removed by dnf at any point, if nothing else depends on it 17:42:11 even important system packages 17:42:33 does this affect things updated with Software? 17:42:37 yes 17:42:39 if so we might perhaps take it under "The installed system must be able to download and install updates with the default graphical package manager in all release-blocking desktops. " 17:42:46 still, that's a stretch... 17:43:05 the issue is the combination of using two different package managers 17:43:12 yeah 17:43:13 we don't have a direct criterion for it 17:43:20 well, we could cite the two together 17:43:22 but it can break the default app functionality :0 17:43:31 (alpha has "install using CLI tool") 17:43:38 that's true 17:43:42 autoremove kicks in when you run 'dnf update'? 17:43:57 or does it have to be explicitly run? 17:44:00 by default, it kicks in for any operation that removes something, I believe 17:44:07 hmm, ok 17:44:12 i think i can go +1 on that basis 17:44:15 pschindl? tflink? 17:44:33 certainly dnf remove, not sure about update 17:45:28 +1 17:45:37 i thought i had a todo item to require basic package manipulation (not just updates) to work safely, sigh 17:46:03 but it's easy to imagine a situation where you update e.g. libreoffice using gnome-software, and then remove libreoffice-latex plugin and it takes away whole libreoffice 17:46:09 that is still a safe example 17:46:14 it might get worse 17:46:33 kparal: yeah, problem is the criteria don't explicitly cover anything but updates anywhere 17:46:50 we don't technically require that package install or removal works, or doesn't explode your system, without a bit of criteria judo 17:46:53 still, let's fudge it for now 17:46:59 I pinged hughsie about this and he said he thinks kalev already fixed it, but it's not stable yet 17:47:25 proposed #agreed 1259865 - AcceptedBlocker (Beta) - we held that this violates the combination of "The installed system must be able to download and install updates with the default console package manager." (Alpha) and "The installed system must be able to download and install updates with the default graphical package manager in all release-blocking desktops." (Beta) as it's believed that doing the former after the latter could lead to the rem 17:47:25 oval of critical packages 17:47:27 grr 17:47:33 proposed #agreed 1259865 - AcceptedBlocker (Beta) - we held that this violates the combination of "The installed system must be able to download and install updates with the default console package manager." (Alpha) and "The installed system must be able to download and install updates with the default graphical package manager in all release-blocking desktops." (Beta) as it's believed that doing the former after the latter could remove critical 17:47:33 packages 17:47:35 grrrr 17:47:51 proposed #agreed 1259865 - AcceptedBlocker (Beta) - we hold this violates combination of "The installed system must be able to download and install updates with the default console package manager." (Alpha) and "The installed system must be able to download and install updates with the default graphical package manager in all release-blocking desktops." (Beta) as it's believed that doing the former after the latter could remove critical packages 17:47:53 there. 17:47:55 one has to love irc 17:48:17 ack 17:48:25 +1 and ack 17:48:35 #agreed 1259865 - AcceptedBlocker (Beta) - we hold this violates combination of "The installed system must be able to download and install updates with the default console package manager." (Alpha) and "The installed system must be able to download and install updates with the default graphical package manager in all release-blocking desktops." (Beta) as it's believed that doing the former after the latter could remove critical packages 17:48:53 ok, just a couple of alpha proposed FEs to finish on 17:49:00 #info Alpha FEs 17:49:04 #topic (1299210) No initial-setup on Fedora Rawhide 20160115 (qemu-kvm) 17:49:04 #link https://bugzilla.redhat.com/show_bug.cgi?id=1299210 17:49:04 #info Proposed Freeze Exceptions, initial-setup, MODIFIED 17:49:48 +1 17:49:56 this could actually be blocker if it affects ARM disk image 17:50:00 but +1 FE for sure 17:50:39 we're not yet frozen, is there any reason to go through FEs? 17:50:43 +1 sure 17:50:46 hmm, point 17:50:49 i'm all for ending the meeting 17:50:54 anyone really want to do 'em? 17:51:04 don't shout all at once :) 17:51:24 #info actually, we're not frozen for a while, so let's skip FEs for now 17:51:35 #topic Open floor 17:51:39 any other blocker-related business? 17:51:44 wheeeee 17:51:47 nope 17:51:48 * adamw lights the blue touch paper 17:52:37 thanks for coming out and punting everything, folks ;) 17:52:43 :D 17:52:43 #endmeeting