16:00:07 #startmeeting F26-blocker-review 16:00:07 Meeting started Mon May 8 16:00:07 2017 UTC. The chair is roshi. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:07 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:00:07 The meeting name has been set to 'f26-blocker-review' 16:00:08 #meetingname F26-blocker-review 16:00:08 The meeting name has been set to 'f26-blocker-review' 16:00:08 #topic Roll Call 16:00:28 who's around for some blocker review time!? 16:01:13 ahoyhoyhoy 16:01:16 * adamw is! 16:01:22 yeah buddy! 16:01:29 #chair adamw 16:01:29 Current chairs: adamw roshi 16:02:59 blocker bug best buds! 16:03:05 :D 16:03:48 i spammed a few channels 16:04:15 sgallagh: tflink: danofsatx: amita: nirik: pwhalen: satellit: ping 16:04:41 * tflink is otherwise occupied today 16:04:44 .hello sgallagh 16:04:45 sgallagh: sgallagh 'Stephen Gallagher' 16:05:19 #chair sgallagh 16:05:19 Current chairs: adamw roshi sgallagh 16:05:44 well, suppose we can get started with we three 16:05:53 #topic Introduction 16:05:53 Why are we here? 16:05:53 #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. 16:05:57 #info We'll be following the process outlined at: 16:05:59 #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 16:06:02 #info The bugs up for review today are available at: 16:06:04 #link http://qa.fedoraproject.org/blockerbugs/current 16:06:07 #info The criteria for release blocking bugs can be found at: 16:06:09 #link https://fedoraproject.org/wiki/Fedora_26_Alpha_Release_Criteria 16:06:12 #link https://fedoraproject.org/wiki/Fedora_26_Beta_Release_Criteria 16:06:15 #link https://fedoraproject.org/wiki/Fedora_26_Final_Release_Criteria 16:06:36 we've only 4 proposals to go through, so should be quick 16:06:40 first up... 16:06:40 #topic (1447777) Since curl-7.53.1-7.fc26 , curl TLS transactions in initramfs environment are broken 16:06:44 #link https://bugzilla.redhat.com/show_bug.cgi?id=1447777 16:06:46 #info Proposed Blocker, dracut, POST 16:08:36 welcome Southern_Gentlem 16:08:37 this seems a pretty obvious +1 16:08:43 #chair Southern_Gentlem 16:08:43 Current chairs: Southern_Gentlem adamw roshi sgallagh 16:09:18 It's not technically a violation as written. (The criteria only cite HTTP). That said, it's pretty important to fix. 16:09:26 seems that way for me 16:09:40 doesnt this effect dnf as well 16:09:57 adamw: can you take over for a second? 16:10:01 brb 16:11:09 sure sure 16:11:19 I'm +1 to block on this, I guess. 16:11:23 sgallagh: as the person who wrote that criterion, it's not intended to be 'HTTP not HTTPS' 16:11:25 +1 16:11:27 * pwhalen is here 16:11:34 +1 16:11:39 It's kind of academic at this point anyway. 16:11:40 it was just 'http not ftp or nfs' 16:12:14 we can clarify the criterion, but i'd say the intent would be for this to block 16:12:23 I voted +1 16:12:30 yep, just clarifying 16:13:44 proposed #agreed 1447777 - AcceptedBlocker (Beta) - this is a violation of "The installer must be able to download and use an installer update image from an HTTP server" (note: 'HTTP' there is meant as 'HTTP not FTP or NFS', it does not mean 'we don't expect HTTPS to work') 16:14:27 ack 16:14:31 ack 16:14:51 ack 16:15:19 #agreed 1447777 - AcceptedBlocker (Beta) - this is a violation of "The installer must be able to download and use an installer update image from an HTTP server" (note: 'HTTP' there is meant as 'HTTP not FTP or NFS', it does not mean 'we don't expect HTTPS to work') 16:15:35 #topic (1443938) Firefox 53 - Second arches build failure 16:15:35 #link https://bugzilla.redhat.com/show_bug.cgi?id=1443938 16:15:35 #info Proposed Blocker, firefox, NEW 16:15:48 back, thanks 16:16:23 hi kevin 16:17:05 +1 16:17:10 um 16:17:21 i'm confused, here. if the package didn't build, the old package should still be tagged. 16:17:25 What criterion is this violating, exactly? 16:17:27 so what's the blocker here? 16:17:55 maintainer made it intel only, we dont get armhfp xfce image anymore as a result 16:18:11 image or firefox? 16:18:13 oh crap, i see 16:18:20 Made Firefox intel-only? 16:18:22 roshi, ff 16:18:28 sgallagh: https://koji.fedoraproject.org/koji/buildinfo?buildID=881361 16:18:28 Hmm 16:18:39 the maintainer did a build with non-x86 arches disabled 16:18:43 he really should not have done that 16:18:52 happened in alpha too 16:19:03 sigh. we'll have to get someone to explain to martin 16:19:26 so +1, since it does not seem reasonable to say 'we'll ship ARM with some other browser or no browser' 16:19:56 Is there a strict requirement for a browser on the media? 16:20:11 I thought it was just "default apps must work" 16:20:56 sgallagh, and wouldnt ff be a default app? 16:21:01 alpha: "It must be possible to run the default web browser and a terminal application from all release-blocking desktop environments." 16:21:09 this implies there must *be* a 'default web browser' 16:21:15 it is - but ff is the default 16:21:35 +1 16:21:40 +1 16:21:52 i mean, technically we can file a bug saying "the xfce spin has no default browser on arm" and there could be a resolution besides "fix the ff build for arm" (i.e. switch xfce's default to something else) but do we really want to do that? 16:22:25 I think we probably have to rule this blocker decision that way 16:22:54 Because if FF really is broken upstream for some arches, we probably need to switch defaults. 16:23:27 proposed #agreed - RHBZ#1443938 - AcceptedBlocker - This bug is a conditional violation of the following criterion: "It must be possible to run the default web browser and a terminal application from all release-blocking desktop environments." 16:23:35 (Also, we could *really* hedge and declare elinks or something to be default on ARM...) 16:23:44 it can't be just for ARM, really. 16:23:44 :/ 16:23:52 we can't define package sets by arch in that way. 16:23:58 roshi: I'm not sure we are in agreement yet 16:23:59 it's whatever the default is in the DE 16:24:08 whatever it was, would also be the 'default browser' for xfce on x86_64. which isn't blocking, but still 16:24:14 roshi: right, which is a comps change. 16:24:18 I think we're in agreement that this *issue* blocks - just not on the fix 16:24:37 roshi: i think sgallagh is arguing that the blocker bug should be "there's no default browser on Xfce ARM" 16:24:50 which allows the resolution of "change the default browser", rather than "fix firefox". 16:24:51 Well, I want whatever our statement is to make it clear that there's more than one way to unblock 16:25:02 adamw: yes 16:25:10 * roshi is fine with file new bug and block on that - however, people will expect ff on arm 16:25:29 roshi, agreed, and ff is the default on a few desktops 16:25:29 which, i mean, i'm fine with, it just involves more work, and yeah, i think that would be a pretty shitty resolution 16:25:33 I don't want to hamstring us in case there is a genuine problem with upstream here 16:25:53 fair 16:25:56 it's pretty hard to argue we'd do this for x86_64, so we'd be treating ARM as a second-class citizen if we did it 16:25:56 perhaps we should see if it is. for alpha it was fixed 16:25:59 which we're not supposed to do 16:26:12 but I also don't see mozilla saying they don't care about the other arches and this problem lingering 16:26:23 right 16:26:29 but i'm okay to go with the 'file a different bug and +1 it' route. 16:26:39 the suggestion to change the browser seems premature 16:26:44 roshi: I can totally see them deciding not to support 32-bit entirely, for example 16:26:57 but not arm 16:26:58 pwhalen: it's not a *suggestion*, per se, just a recognition that the blocker issue is not technically 'firefox doesn't build' 16:27:20 sgallagh: why don't you file a new bug while we paint the bike shed :) 16:27:21 is what I'm saying - even if it is an upstream issue, there's momentum behind arm, I don't see them dropping it 16:27:31 I'm not saying we should change the browser, just that we shouldn't close that off as an option to avoid a slip 16:27:51 I don't see how blocking on this closes that door? 16:28:01 adamw: I'm currently on my phone at lunch, so I can't. 16:28:01 nor I 16:28:12 I mean, if we decide to go that way we just update the bug with a note to the discussion 16:28:12 fine, i'll do it 16:28:21 roshi: the criterion we list carries implications 16:28:39 roshi: the thing is, we *do* need a bug for "firefox doesn't build on ARM" 16:28:45 that in itself clearly *is* a bug 16:29:00 but we can't mark that bug as the release blocker and say "oh but the release blocking issue is _really_ that there's no default browser on Xfce ARM" 16:29:09 it's confusing and unclear 16:29:30 i'll file a new bug. 16:29:35 * roshi shrugs - I didn't think it was confusing 16:30:25 if the bug was "no default apps get installed" we'd block on this, and this bug is just a subset of that - since there'd be no ff to install 16:30:34 it's not a subset of that 16:30:40 because 'that' could be fixed without this being fixed 16:31:02 if we change the xfce package set to use, say, epiphany, then the bug "there is no default browser on xfce" gets fixed, but the bug "firefox does not build on arm" is *not* fixed 16:31:27 pwhalen: does xfce ARM actually fail to compose right now? i'd guess that'd be the result? 16:31:39 but doesn't the "default application" list come from xfce, not from what we decide? 16:31:46 actually midori is still "default" for some values... 16:31:47 No 16:31:59 adamw, yes, as does any desktop needing firefox 16:32:00 yeah, it does. 16:32:01 I mean, if you changed the comps for gnome to all kde apps, would we then say that the gnome defaults work because it's defined in comps? 16:32:20 roshi: i don't think upstream firefox defines a default browser. 16:32:21 er 16:32:22 It comes from whatever we put in the comps entry that is used to compose the media 16:32:23 upstream xfce 16:32:38 then the xfce sig? 16:32:51 both firefox and midori are shipped. 16:33:03 but if you click the little "browser" button you get midori I think... 16:33:09 roshi: yeah, which is who would make the decision for whether to drop FF if it couldn't be fixed. 16:33:43 The Xfce sig. But we added it because lots of people wanted it. 16:34:15 nirik: if thats the case, I'd argue Midori is the real default. 16:34:24 yes, I would think thats reasonable 16:34:29 And the blocker bug here is "XFCE spin doesn't compose" 16:34:56 sure, but we want to include firefox... so they should fix things 16:35:13 and in the past they have pretty quickly 16:35:35 Yes, I agree. But that's not *blocking* 16:35:46 Which is the only part we care about in this meeting 16:35:55 xfce not composing is blocking because of arm 16:35:56 I suppose. 16:35:59 I know 16:36:09 so midori is the default browser 16:36:12 * nirik has to leave... hackfest is heading to lunch 16:36:13 https://bugzilla.redhat.com/show_bug.cgi?id=1448923 16:36:19 file the bugs. I will deal with it somehow. 16:36:20 ff is a nice add-on 16:36:28 thanks nirik ! 16:36:33 so, move that we transfer the blocker nomination to that bug>? 16:36:35 i dont think midori is on the images 16:36:41 works for me 16:36:51 pwhalen: but we could change that. theoretically. the option exists. 16:36:53 adamw: agreed 16:37:05 no , there were issues with midori on arm. 16:37:21 We don't have to solve the problem here 16:37:25 ff was more active, the change was good for us 16:37:41 As long as the image starts to compose again, how it happens is irrelevant 16:37:50 (To the blocker decision) 16:37:58 pwhalen: i'm not saying that's what we *ought* to do. but the point is that the possibility exists, so the blocker bug should be "xfce fails to compose and its current default browser isn't available", not "firefox doesn't build". 16:37:58 would we change the broswer if this was x86? 16:38:37 pwhalen: Hypothetically, Chromium is in the repo now... 16:39:01 probably not, but it would still be more correct to have the blocker bug be an accurate definition of the actual criteria violation. 16:39:10 Yes, what he said 16:40:50 sgallagh: it's not quite "as long as the image starts to compose again", it's "as long as the image starts to compose again with a 'default browser'" 16:40:53 ok +1 to 1448923 as blocker, add a note to 1443938 referencing the blocker 16:41:41 adamw: Right, I abbreviated. 16:42:06 +1 like roshi just said 16:42:48 pwhalen, are you okay with that? 16:42:52 i dont think we should open the suggestion to change the default. 16:43:30 its the default on most desktops. we'd lose all of those for arm 16:44:30 We don't need to include that in the blocker decision message 16:44:39 I'm in strong support of fixing FF for this, fwiw 16:45:02 I just want it clear that the issues we are blocking on are "Doesn't compose" and (possibly) "has no functional default browser" 16:45:10 I see the "change the default" as counterproductive for our ARM presense 16:45:34 which might be why I'm confused as to why this is confusing, I guess 16:45:36 pwhalen: having the new bug as the blocker does not imply support for changing the default 16:45:40 I am also strongly in favor of fixing Firefox rather than changing the default 16:46:12 But I'd choose "switch to Midori" over "slip all of Fedora" if it became clear that Firefox could not be fixed in a reasonable timeframe 16:46:31 adamw, it was suggested as an alternative option in the bug which to me seems to early to consider 16:46:36 so we added a bug and a level of obfuscation from the actual issue (IMO) for a thing none of us actually wants to do (change the default) 16:46:39 *too 16:46:47 ? 16:46:55 that's what it's seeming like to me 16:47:27 roshi: well, yes, because sgallagh and i think it's important to accurately describe the blocking issue. 16:47:31 but we're going in circles here 16:47:52 hello adamw 16:47:53 * roshi would have just put the details in the acceptedBlocker comment to make it clear 16:47:54 hi everyone 16:47:55 we basically have something like +3/-1 for each bug, with a different +3 and a different reason for the -1. 16:48:00 welcome Amita 16:48:19 adamw: want to take a stab at the proposed? 16:48:31 well not really cos i don't know which one to propose. :P 16:48:40 can i just toss a damn coin or something? 16:48:41 that's what I asked you for 16:48:42 * sgallagh will make an attempt 16:48:47 :p 16:48:51 since I don't know either 16:48:52 if we're going to propose the other bug, we have to change the topic. 16:49:03 im ok with changing the title of it 16:49:15 so propose against the shadow bug we just created I guess 16:49:28 it doesn't make sense to me to change the title of the 'fails to build on arm' bug because we *do* need a 'firefox fails to build on arm' bug. 16:49:39 right 16:49:45 if we fix the 'no default browser / image fails to compose' bug by making firefox build on ARM again, we just close both bugs at once. 16:49:59 I thought we were meaning the title of the room right now so logs get generated correctly 16:50:13 Proposed: BZ 1448923 is accepted as a blocker on the grounds that it causes a release-blocking image to fail to compose. BZ 1443938 is rejected as a blocker because it does not violate any specific criteria. 16:50:15 that's what i meant, but i don't think it's what pwhalen meant :P 16:50:24 right, misunderstood :) 16:50:43 sgallagh: i agree with the sentiment, but we should do it by unproposing this bug with a #info and switching the topic to the new bug. 16:50:59 sigh, i love bug bureaucracy 16:51:07 adamw: Un-proposing or rejecting? 16:51:07 someone does 16:51:15 hey, I did promise blocker funtimes, remember? 16:52:44 sgallagh: meh, either way. 16:52:59 Proposed: BZ 1443938 is rejected as a blocker because it does not violate any specific criteria. 16:53:58 ack 16:54:07 ack 16:54:15 * roshi still thinks it's causally a blocker though :p 16:54:17 propose #agreed #1443938 - RejectedBlocker (Beta) - this is rejected in favour of the alternative proposal #1448923 , which more precisely reflects the actual criteria violation 16:54:20 well, it did 16:54:28 (just a bit cleaner than sgallagh's proposal) 16:54:35 ack 16:54:38 ack adamw 16:54:38 +1 16:54:40 ack 16:54:41 adamw: ack 16:54:54 #agreed #1443938 - RejectedBlocker (Beta) - this is rejected in favour of the alternative proposal #1448923 , which more precisely reflects the actual criteria violation 16:55:02 (crp by committee is always interesting ) 16:55:03 roshi: OK, I proposed the other one so we can go ahead and switch to it now? 16:55:07 Southern_Gentlem: right? :P 16:56:46 #topic (1448923) Xfce ARM image default browser is unavailable (image fails to build) 16:57:01 #link https://bugzilla.redhat.com/show_bug.cgi?id=1448923 16:58:21 +1 16:58:22 +1 blocker 16:58:35 +1 16:59:02 +1 16:59:14 (crp by committee is always interesting ) 16:59:16 +1 16:59:52 proposed #agreed - RHBZ#1448923 - AcceptedBlocker - This bug is a conditional violation of the following criterion: "It must be possible to run the default web browser and a terminal application from all release-blocking desktop environments." 17:00:17 mention the 'image fails to compose' criterion too? 17:00:30 * roshi thought about that 17:01:34 proposed #agreed - RHBZ#1448923 - AcceptedBlocker - This bug is a conditional violation of the following criterion: "It must be possible to run the default web browser and a terminal application from all release-blocking desktop environments." as well as the "image compose" criterion. 17:01:43 ack 17:01:49 ack 17:02:31 #agreed - RHBZ#1448923 - AcceptedBlocker - This bug is a conditional violation of the following criterion: "It must be possible to run the default web browser and a terminal application from all release-blocking desktop environments." as well as the "image compose" criterion. 17:02:52 well, glad that's decided 17:03:00 onto the next! 17:03:01 #topic (1348688) Anaconda cannot access LVM partitions in a LUKS-encrypted disk partition after decryption 17:03:04 #link https://bugzilla.redhat.com/show_bug.cgi?id=1348688 17:03:05 * adamw wipes forehead 17:03:06 #info Proposed Blocker, lvm2, POST 17:08:07 the criterion here would, I guess, be Beta " Correctly interpret, and modify as described below, any disk with a valid ms-dos or gpt disk label and partition table containing ext4 partitions, LVM and/or btrfs volumes, and/or software RAID arrays at RAID levels 0, 1 and 5 containing ext4 partitions 17:08:07 Create mount points backed by ext4 partitions, LVM volumes or btrfs volumes, or software RAID arrays at RAID levels 0, 1 and 5 containing ext4 partitions " 17:09:55 this one's new to me, but it reads like a violation of that... 17:10:06 that's what I was thinking 17:11:19 especially since it's claimed even the workaround doesn't work on f26 17:11:26 +1 17:13:10 so yeah, +1 for now 17:14:13 anyone else still alive out there, or are we suffering a plague of terminal blocker fatigue? 17:14:42 +1 17:14:45 lol 17:15:23 proposed #agreed - RHBZ#1348688 - AcceptedBlocker - This bug is a clear violation of the following criterion: "Correctly interpret, and modify as described below, any disk with a valid ms-dos or gpt disk label and partition table..." 17:15:42 patch 17:15:52 go for it 17:16:18 + 17:16:20 1 17:16:23 proposed #agreed - RHBZ#1348688 - AcceptedBlocker - This bug is a violation of the Beta criterion: "When using the custom partitioning flow, the installer must be able to: ... Correctly interpret, and modify as described below, any disk with a valid ms-dos or gpt disk label and partition table..." 17:16:43 ack 17:16:45 ack 17:17:00 ack 17:18:44 ooh, sorry 17:18:46 that was my proposal :P 17:18:50 #agreed - RHBZ#1348688 - AcceptedBlocker - This bug is a violation of the Beta criterion: "When using the custom partitioning flow, the installer must be able to: ... Correctly interpret, and modify as described below, any disk with a valid ms-dos or gpt disk label and partition table..." 17:19:37 * roshi just finished typing out your bits too :p 17:19:44 last one! 17:19:45 #topic (1227736) Minimal grub after a kernel update with gnome-software 17:19:48 #link https://bugzilla.redhat.com/show_bug.cgi?id=1227736 17:19:51 #info Proposed Blocker, plymouth, NEW 17:20:51 sorry I still need to investigate this issue ^ 17:21:00 ah, the Punted One 17:22:28 i did ping people on this one 17:23:19 zbigniew and ray (halfline) both responded and seem interested in addressing this 17:23:35 zbigniew says "I'd think it's rather FE or at most Final material, because the setup 17:23:35 is non-default, the fix is rather complex, and apparently the issue is 17:23:35 not that very common, and it's not a regression." 17:23:41 halfline, what's your opinion on blockeriness? 17:27:22 bueller? 17:27:49 welp, i'm -1 beta based on what we know from the bug report and the feedback i had by mail. 17:27:56 wfm 17:28:37 -1 blocker, it sounds like 17:28:59 can we make it +1 FE? 17:30:51 proposed #agreed - RHBZ#1227736 - RejectedBlocker - This bug doesn't qualify as a blocker due to the fact that it's a non-default installation method as well as how invasive the fix is. 17:30:59 ack 17:31:31 for something like this i think i'd want to see a proposed fix or at least have more concrete info on what it would be before voting for FE 17:31:36 true 17:31:39 good point 17:32:30 ack 17:33:02 #agreed - RHBZ#1227736 - RejectedBlocker - This bug doesn't qualify as a blocker due to the fact that it's a non-default installation method as well as how invasive the fix is. 17:33:08 that's all we have for today 17:33:21 there are some proposed FEs for Final, but I don't see a reason to do those now 17:33:31 ah 17:33:32 agreed 17:33:36 guess there's a FE for beta 17:33:49 #topic (1444654) Moving a file using Drag and Drop onto a folder works immediately the first time in nautilus, takes time subsequently. 17:33:52 #link https://bugzilla.redhat.com/show_bug.cgi?id=1444654 17:33:55 #info Proposed Freeze Exceptions, nautilus, NEW 17:34:00 we're close to freeze, so might as well 17:34:55 sure 17:35:10 +1 punt until we have a fix to look at 17:35:25 I really should use a mouse more to catch these things 17:35:35 huh, funny bug 17:35:46 as my drag and drop is mv 17:36:05 but yeah, i think i agree, this one would depend on how tricky the fix is 17:38:05 * roshi waits for a couple more votes... 17:38:28 nb: pwhalen: sgallagh: southern_gentlem: 17:38:29 +1 punt wfm 17:38:36 proposed #agreed - RHBZ#1444654 - Punt - Before we can vote on this we need a proposed fix so we can see how tricky it is. 17:39:14 ack 17:39:20 * nb looks 17:39:29 ack 17:40:26 ack 17:41:01 +1 punt 17:41:07 ack 17:41:38 #agreed - RHBZ#1444654 - Punt - Before we can vote on this we need a proposed fix so we can see how tricky it is. 17:41:42 #topic Open Floor 17:41:53 anyone have anything? 17:42:06 not I 17:42:14 nope 17:42:33 thanks for coming everyone! 17:42:39 * roshi sets the fuse... 17:42:40 3... 17:43:12 thanks roshi! 17:43:13 * adamw runs dramatically, in slow motion 17:43:25 cameraperson! wide shot! 17:43:40 2... 17:43:41 * pwhalen is diving behind some 'containers' 17:44:02 (little do they know I'm on an atomic host! bwahahaha) 17:44:04 1... 17:44:09 #endmeeting