15:02:28 <adamw> #startmeeting Fedora QA meeting
15:02:28 <zodbot> Meeting started Mon Jul 22 15:02:28 2013 UTC.  The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:02:28 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:02:29 <adamw> #meetingname fedora-qa
15:02:29 <zodbot> The meeting name has been set to 'fedora-qa'
15:02:31 <adamw> #topic Roll call
15:02:35 <adamw> morning folks!
15:02:40 * jskladan1 lurks from his cellphone
15:02:43 <kparal> hi
15:02:55 <adamw> who's around for qa meeting fun?
15:03:32 * satellit_e listening
15:03:34 * pschindl is here
15:03:46 * Martix is here
15:04:03 * tflink is here
15:04:30 * brunowolff is here
15:05:05 <adamw> alrighty
15:05:09 <adamw> anyone not here clearly doesn't like fun
15:06:05 <kparal> boooo
15:06:08 <adamw> before we leap into fedora 20 feature review, did anyone think of any major topics I'd forgotten?
15:06:59 <adamw> okay then
15:07:06 <adamw> #topic Fedora 20 Change review
15:07:42 <adamw> so, there's a metric buttload of Changes so far proposed/accepted for F20, I put as many of them as I could find on the list at https://fedoraproject.org/wiki/QA/Meetings/20130722#Fedora_20_Change_review
15:07:55 <adamw> i've been following the devel@ discussions and I think a few of you have too
15:08:05 <adamw> but wanted to check if we have any other concerns that haven't been raised yet with any of them
15:08:27 <kparal> we probably should re-visit ARM, since it got half-accepted
15:08:40 <adamw> well that's what last week was for
15:08:49 <adamw> i wanted to avoid spending another hour in that bikeshed
15:08:56 <adamw> but did you have something important to bring up about it?
15:09:08 <kparal> somebody said he could send us some hw
15:09:24 <kparal> since it got half-accepted, it might be a good time to have the hw
15:09:56 <tflink> as a dev, the web assets feature concerns me a bit, but that has nothing to do with qa
15:10:18 <adamw> kparal: take any hardware you can get your claws on, sure :)
15:10:39 <Martix> adamw: I have some info about Bluez5 and it's integration with Gnome and KDE
15:10:47 <kparal> adamw: so, who was the generous guy?
15:10:57 <adamw> hmm? i thought you knew
15:11:11 <kparal> nope. but I can read the logs
15:11:17 <adamw> oh, you mean last week. sure
15:11:18 <Martix> kparal: handsome-something?
15:11:25 <adamw> that's j_dulaney
15:11:29 <Martix> pirate?
15:11:40 <kparal> I can ping John and ask about it
15:11:51 <adamw> Martix: bluez5 isn't even listed as a Change, is it?
15:11:56 <adamw> seems like maybe it ought to be
15:12:08 <Martix> kparal: it is, I don't know if it was accepted
15:12:14 <Martix> adamw: ^
15:12:26 <tflink> would the kdump under secureboot affect anything other than kdump?
15:12:51 <kparal> it might not be a bad idea to send the ARM hardware to Beijing, so that the guys there enjoy some change
15:12:51 <Martix> adamw: not yet: https://fedoraproject.org/wiki/Changes/Bluez5
15:13:11 <kparal> apart from x86/64 testing
15:13:33 <kparal> but we can have it in Brno as well of course
15:13:38 <adamw> Martix: oh, i don't see it on devel@...
15:13:49 <adamw> tflink: I don't think so - that one's basically 'get something working that doesn't work now', aiui
15:14:03 <kparal> adamw: the other thing that jumps up on me is AppInstaller with hawkey instead of yum
15:14:05 <Martix> adamw: it was proposed, I saw it there
15:14:12 <robatino> i take it BTRFS by default is off the table this time? it used to be discussed for each release but haven't heard anything this time
15:14:33 <adamw> Martix: searching 'bluez' in devel archives just shows 'FYI: F20 Feature: Migrate to Bluez5', not an official thread from feature wrangler
15:14:37 <kparal> robatino: after so many corrupted filesystems I would be very surprised :)
15:14:40 <adamw> wonder if the wrangler's still working on it
15:14:47 <adamw> robatino: that's a good question actually
15:15:04 <adamw> always seems to be coming next release
15:15:47 <brunowolff> I noticed that RHEL 7 is supposed to be using XFS by default.
15:16:07 <brunowolff> It was in a roadmap presentation a month or so ago.
15:16:17 <Martix> adamw: I talked to both Gnome and KDE about bluez5 migration and KDE support doesn't exist yet, but they can have it in a month
15:16:19 <drago01_> adamw: gnome3.10 might move to bluez5
15:16:30 <drago01_> adamw: so it might get in that way fyi
15:16:52 <drago01_> brunowolff: how is this relevant for fedora?
15:16:54 <Martix> drago01_: right, bluez is maintained by Gnome developer
15:16:59 <adamw> drago01: well, the point is that if it's going in it ought to be listed as a Change, since it needs co-ordination with other c onsumers
15:17:32 <drago01_> adamw: https://fedoraproject.org/wiki/Changes/Bluez5
15:17:42 <brunowolff> It might indicate the amount of resources that Red Hat is willing to put into making BRTFS the default for Fedora.
15:18:44 <brunowolff> If Red hat isn't pushing for it, it is much less likely to happen.
15:19:00 <drago01_> Josef left Red hat a while ago
15:19:25 <adamw> anyway, we're getting a bit distracted
15:20:30 <adamw> #info https://fedoraproject.org/wiki/Changes/Bluez5 needs to be integrated smoothly with KDE 4.11 and GNOME 3.10 at least
15:20:57 <adamw> kparal: what was your concern about AppInstaller?
15:21:12 <adamw> #info btrfs-by-default appears to be off the table again for F20
15:21:13 <kparal> adamw: that we might end using two different dep solvers
15:21:17 <kparal> *end up
15:21:33 <adamw> I think fesco is considering that issue
15:21:46 <kparal> it would definitely complicate verification
15:22:07 <adamw> #info https://fedoraproject.org/wiki/Changes/AppInstaller may result in us using two depsolvers in official packaging tools
15:22:18 <adamw> verification of what?
15:22:37 * nirik notes it will be in this week's fesco meeting. The current rule is only one.
15:22:46 <kparal> of package installers working properly. I don't know.. for example install with yum and remove with AppInstaller, what happens?
15:23:05 <kparal> I just wanted to note that there might be some issues lurking
15:24:06 <kparal> but the criteria are written OK I think
15:24:21 <adamw> oh, I see. the criteria focus mostly on installing *updates*
15:25:00 <kparal> alright
15:25:05 <adamw> i think what's complicating things at present is the unresolved conflict between gpk-update-application and offline updates, if this feature resolves that it might actually be a _win_ in validation terms
15:25:10 <adamw> but yeah, something to keep an eye on
15:25:33 <adamw> https://fedoraproject.org/wiki/Changes/VisibleCloud would have obvious deliverable and criteria consequences
15:25:45 <kparal> if they keep both AppInstaller and gpk-update, we will have 4 ways to install updates :)
15:27:21 * kparal wonders if we have people to verify those cloud criteria, if they get written
15:27:22 <adamw> is anyone super excited to do some cloud testing? :)
15:27:27 <adamw> yeah
15:27:41 <adamw> I know dgilmore keeps an eye on the cloud images
15:27:47 <kparal> but we don't have people for xen etc verification anyway, so nothing new here
15:27:50 <adamw> but we'd really want a couple of other people to help with testing if they're going official
15:27:51 <tflink> I think it's important, but not sure I want to commit to anything
15:28:06 <adamw> it doesn't require much in the way of special resources, though, i don't think
15:28:43 <tflink> it might be worth noting that mattdm is having an intern work on automated cloud tests, or at least he was last time I spoke with him about it
15:29:46 <adamw> #info https://fedoraproject.org/wiki/Changes/VisibleCloud will likely require changes to criteria and extra validation testing
15:30:01 <adamw> I'm planning to file tickets for that stuff if the feature gets accepted, for the record
15:31:43 <adamw> https://fedoraproject.org/wiki/Changes/SSD_cache looks pretty vague but could be significant if accepted, I guess
15:31:51 <adamw> I think people on devel@ were confused about the scope of it
15:31:53 * satellit_e afk
15:31:54 <adamw> anyone know anything about that one?
15:32:15 <kparal> not a clue
15:32:38 <tflink> me neither
15:33:10 <adamw> guess we have to keep an eye on it then
15:33:13 <brunowolff> It seems to need changes to ananconda.
15:33:17 <adamw> nirik: has there been any fesco discussion of it yet?
15:34:13 <nirik> don't think so. I think its this week? yeah, some anaconda changes, but not much else really.
15:35:34 <adamw> #info https://fedoraproject.org/wiki/Changes/SSD_cache may result in anaconda changes and changes to configuration of deployed systems with SSDs
15:36:02 <adamw> https://fedoraproject.org/wiki/Changes/SDDMinsteadOfKDM replaces the login manager for KDE (one of the release-blocking desktops)
15:36:10 <adamw> should be no problem as long as it lands early enough to shake any bugs out
15:36:55 <Martix> adamw: back, Gnome is ready for bluez5, KDE need port of libbluedevil, which will take a month...haddes (bnocera) will ask releng, if F20 can have both bluez4 and bluez5, but the problem is, it can't be installed side by side -> this could lead to broken deps when installing second desktop (Gnome and KDE)
15:36:56 <adamw> https://fedoraproject.org/wiki/Changes/InstallerLVMThinProvisioningSupport is a semi-large change to partitioning in the installer - yay
15:37:11 <adamw> Martix: that seems like a substantial issue
15:37:48 <Martix> best approach is to port libbluedevil
15:39:01 <Martix> adamw: anyway, I'm going to do Bluetooth test day when both desktops will be ready
15:39:29 <adamw> nirik: can you take notes on the bluez issues for fesco? and maybe try to bang some gnome/kde heads together when the change comes up for discussion?
15:39:42 <adamw> thanks for looking into it, martix
15:40:02 <Martix> no problem :-)
15:41:51 <nirik> sure, yeah, that seems non ideal
15:42:00 <adamw> #info GNOME 3.10 is ported to Bluez5 and won't work with Bluez4, but KDE's libbluedevil is not yet ported to Bluez5 and may not be for a month: we would like FESCo to take a hand in dealing with that issue
15:43:10 <adamw> https://fedoraproject.org/wiki/Changes/Plasma-nm is another KDE change; looks like there's quite a bit of KDE churn for F20, we should probably test it a bit harder than usual
15:43:40 <Martix> adamw: I can take care of it
15:43:48 <adamw> thanks martix
15:43:59 <Martix> I was looking for new Plasma-NM
15:44:35 <Martix> *looking forward
15:45:04 <adamw> i guess it'd be good to know the timeframe on that change too
15:45:55 * adamw fires off an email
15:46:01 <adamw> okay...anything we missed?
15:46:24 <adamw> #info we can look into the question of which Changes we should try and 'validate' once FESCo finishes approving them
15:48:50 <adamw> okay, looks like that's it for Changes
15:48:54 <adamw> #topic open floor
15:48:57 <adamw> any other business, folks?
15:49:19 <adamw> note that the first release blocker meeting is scheduled for Wednesday
15:49:25 <tflink> say what?
15:49:28 <adamw> uh huh
15:49:34 <adamw> http://fedorapeople.org/groups/schedule/f-20/f-20-quality-tasks.html , tape it to your fridge
15:49:45 <adamw> we can skip it if it's not worth doing
15:49:57 <tflink> so we're doing blocker review meetings before all the features have been reviewed?
15:49:58 * nirik files a blocker bug on needing an extra month of time
15:50:21 <adamw> tflink: i don't see that those two are particularly related
15:50:26 <tflink> it's a good thing we don't have any major features proposed for F20, otherwise  I'd be worried
15:50:28 <adamw> we can review blockers so long as they've been proposed
15:50:56 <tflink> adamw: criticism of the continued pretending that we're actually going to branch as scheduled and be successful
15:51:12 <tflink> but then again, sarcasm doesn't work well over irc
15:52:20 <adamw> i'm still not sure what it is you're trying to communicate, but point is, the 'f20' (rawhide) tree exists and is testable and we have nightly lives and installer images and stuff and people are testing them and filing blocker bugs, so we can review the blocker bugs. seems straightforward
15:52:38 <adamw> the details of the development schedule seem not to be particularly significant to that process
15:52:44 * tflink is just frustrated, feel free to ignore him
15:52:47 <adamw> roger!
15:53:07 <adamw> i'd say schedule the meeting if it looks like anything would be achieved by reviewing the current proposed blockers, if not, don't.
15:53:58 <adamw> #info first F20 blocker review meeting is scheduled for this Wednesday (07-24): may happen or not depending on how many blockers are proposed
15:54:26 <adamw> anything else on anyone's radar?
15:56:42 <adamw> well then...
15:56:56 * adamw sets quantum fuse for as many minutes as it takes his cat to eat a moth
16:00:16 <adamw> thanks for coming, everyone!
16:00:18 <adamw> #endmeeting