16:04:25 <handsome_pirate> #startmeeting Fedora QA meeting
16:04:25 <zodbot> Meeting started Mon Feb 24 16:04:25 2014 UTC.  The chair is handsome_pirate. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:04:25 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:04:36 <handsome_pirate> #meetingname fedora-qa
16:04:36 <zodbot> The meeting name has been set to 'fedora-qa'
16:04:54 <handsome_pirate> #chair roshi danofsatx-work satellit tflink adamw
16:04:54 <zodbot> Current chairs: adamw danofsatx-work handsome_pirate roshi satellit tflink
16:05:08 <handsome_pirate> #topic Roll Call
16:05:13 * cmurf brings out a bolus of thick concentrated coffee+prescription Red Bull and injects adamw intraveneously
16:05:13 <danofsatx-work> aqui
16:05:15 * roshi is here
16:05:16 * satellit here
16:05:20 * pwhalen is here
16:05:22 <handsome_pirate> #chair cmurf
16:05:22 <zodbot> Current chairs: adamw cmurf danofsatx-work handsome_pirate roshi satellit tflink
16:05:30 * mkrizek is here
16:05:35 <handsome_pirate> #chair pwhalen mkrizek
16:05:35 <zodbot> Current chairs: adamw cmurf danofsatx-work handsome_pirate mkrizek pwhalen roshi satellit tflink
16:06:11 <cmurf> Does chair mean I have to do something useful other than give adamw a hard time?
16:06:22 <roshi> it means you're someplace to sit
16:06:28 * tflink is present
16:06:32 <danofsatx-work> I think it means we get to sit down during the meeting.
16:06:39 <handsome_pirate> cmurf:  Not really, but if you want to #info something
16:06:43 <handsome_pirate> #chair tflink
16:06:43 <zodbot> Current chairs: adamw cmurf danofsatx-work handsome_pirate mkrizek pwhalen roshi satellit tflink
16:06:48 <danofsatx-work> noob's have to stand
16:06:49 <roshi> or, in WWE fashion, handsome_pirate gave us all the chair
16:06:55 <adamw`> ahoy hoy
16:06:56 <handsome_pirate> lol
16:06:58 <adamw`> NO SITTING
16:07:05 <cmurf> I don't think I've ever info'd anything.
16:07:09 <handsome_pirate> #chair adamw`
16:07:09 <zodbot> Current chairs: adamw adamw` cmurf danofsatx-work handsome_pirate mkrizek pwhalen roshi satellit tflink
16:07:13 <cmurf> OH I SEE THE BOLUS worked right on time
16:07:14 <handsome_pirate> he he
16:07:18 * roshi is perpetually standing
16:07:20 <danofsatx-work> you found an outlet in the snow? wow, I'm impressed....
16:07:44 <handsome_pirate> adamw:  Want to take over? :)
16:07:50 <handsome_pirate> If not ...
16:07:51 <adamw> danofsatx-work: i'm in the lodge
16:07:54 <adamw> handsome_pirate: eh, go right ahead
16:08:01 <adamw> i'll chip in if needed
16:08:05 <handsome_pirate> #topic Fedora.next plans
16:08:21 <handsome_pirate> #info https://fedoraproject.org/wiki/Workstation/Technical_Specification
16:08:41 <handsome_pirate> Has everyone looked over the workstation spec?
16:09:20 * satellit worried about possible loss of persistence on lives...list discussions
16:09:33 * roshi has
16:09:46 * adamw has too
16:09:57 <adamw> i don't see anything that raises immediate alarms...
16:10:21 <roshi> btrfs as default wasn't what I was expecting, but cmurf should be happy :)
16:10:22 <handsome_pirate> So far, it largely looks like it will be standard Gnome DE
16:10:31 <handsome_pirate> roshi:  I like btrfs :)
16:10:32 <danofsatx-work> Oh wow...I must have missed that. Too much data for me to parse right now, but I'll get back to y'all.
16:10:33 <roshi> same here adamw
16:10:33 <adamw> yeah, it's interesting
16:10:45 <cmurf> It's tentative
16:10:52 * roshi hasn't messed with it much
16:10:59 <drago01_> does not mean default for f21 though
16:11:10 * handsome_pirate wonders if this means he needs to dust off the btrfs test cases he wrote a few years ago ...
16:11:22 <adamw> but i think the question of whether it's ready or not is a technical one not a QA one; i'm sure they'll take input from domain experts. might be worth checking the storage folks know about it, though.
16:11:30 <handsome_pirate> drago01_:  Default for the Workstation product, not necessarily other products
16:11:53 <cmurf> Open question if anaconda can support different product defaults
16:11:55 <handsome_pirate> adamw:  From using it everyday for going on two-three years, I think it's ready
16:12:22 <cmurf> I'd say Server folks probably will want XFS, and if we can't have a different default with products then XFS is fine for Workstation too, not a big deal.
16:12:30 <handsome_pirate> cmurf:  Two ways to do that, I think:  Ansible (like what nirik proposed a couple years ago) or kickstart
16:12:52 <nirik> hum?
16:13:18 <handsome_pirate> nirik:  Was it you that proposed using ansible playbooks in Fedora a couple of Fudcons ago?
16:13:19 <nirik> oh, yeah, I would think we could work something out...
16:13:33 <nirik> yes, but I don't think it's a good fit for this case really. ;)
16:13:53 <handsome_pirate> Ah, okay
16:13:53 <adamw> yeah, still wiggle room there.
16:13:53 <adamw> hawkey backend for updates has been approved by fesco, i believe, though it's something we need to take note of
16:13:53 <adamw> i say this all the time, but it means we ought to be careful to test graphical updates and not just use yum all the time
16:13:54 <danofsatx-work> ok, not as much data as I thought.
16:13:57 <handsome_pirate> But, that's largely a question for Anaconda folks
16:14:04 <nirik> also, I note that kernel folks have reservations for btrfs... but the product folks can decide if they want.
16:15:51 <handsome_pirate> All right, any more thoughts?
16:15:58 <danofsatx-work> handsome_pirate: yeah, I would say so. I haven't quite figured out btrfs even after reading all of cmurf's emails on the subject
16:15:59 <danofsatx-work> pardon my naivete - but don't all anaconda instances use a kickstart file? Wouldn't that be a simple change in the kickstart?
16:16:42 <adamw> "The workstation will ship with a single theme, which will have support for the included toolkits: gtk3, qt and gtk2. Applications are expected to work well with this theme, as well as with the high-contrast theme that is used for accessibility" might be slightly trickier to implement than expected
16:16:43 <adamw> it's fine as a high level goal, it's an area i'd want to inject caution into for f21 release requirements though
16:16:44 <nirik> yes, depending on if each of the products have a seperate kickstart and don't share one... for example the install dvd
16:17:05 <danofsatx-work> wow...I'm lagging bigtime
16:17:19 <nirik> I think freenode is lagging. ;(
16:17:38 <adamw> handsome_pirate: i believe it'd be reasonably trivial to do anaconda side
16:17:39 <handsome_pirate> It got hit hard this weekend
16:17:58 <handsome_pirate> adamw:  Indeed
16:18:12 <cmurf> Personally I think if the Workstation WG wants Btrfs for either F21 or F22 that Rawhide should go to Btrfs sooner than later. We can always change it at branch or after alpha if it's not going to be used, but it might give it broader coverage.
16:18:22 <adamw> nirik: right, that's why i noted probably a good idea just to make sure the relevant folks who assess btrfs' 'readiness' have advisory input into the desktop WG's decision
16:18:34 <danofsatx> lemme try this window - different server
16:18:45 <adamw> not really qa's job to do that, but hey
16:19:04 <adamw> handsome_pirate: anaconda already has mechanisms for varying its defaults depending on what product it's installing. rhel uses different defaults from fedora.
16:19:11 <cmurf> So we need some people to assess that readiness and report to the Workstation WG.
16:19:16 <adamw> handsome_pirate: shouldn't be difficult to hook different fedora products into that mechanism.
16:19:29 <adamw> danofsatx-work: that's...um. your understanding is wrong, let's say.
16:19:37 <adamw> ultimately, the release images are built from kickstarts. yes.
16:19:46 <adamw> but not in the way that you could specify the default filesystem to be used when installing from that image, in the kickstart from which the image is built.
16:19:46 <adamw> if you see what I mean.
16:19:53 <adamw> or being DoSed. again.
16:19:54 <adamw> so yeah, the spec defines a " Core Package list " and says "This package list will be the priority focus for QA and bug fixing. "
16:19:59 <adamw> others have commented on the desktop@ list about whether a core package list is really appropriate for a tech spec
16:20:06 <adamw> but insofar as the basic idea goes, i think it's reasonable - obviously there'll be a 'core' set of  functionality which is higher testing priority so far as that product goes. so, i don't see a problem there.
16:20:21 <adamw> so far the spec doesn't seem to be locking in any wild expectations for f21, which is good.
16:20:21 <adamw> cmurf: yeah. again, not really a QA responsibility, but no harm in us dropping a line to the relevant kernel devs or whatever to ensure they're aware of it, if desktop isn't already planning to consult them - drago01?
16:20:48 <adamw> guess he's lagging too :) or i am
16:21:12 * roshi reads you fine
16:21:29 <adamw> oh, well.
16:22:19 <handsome_pirate> adamw:  Roger, and understood
16:23:17 <handsome_pirate> Anyway, probably time to move along
16:23:17 <handsome_pirate> #topic installer discussion: partitioning simplification?
16:23:21 <cmurf> adamw: well apparently there's some variation on what constitutes ready as some fs devs say its ready within certain constraints and others have reservations, and while I know the specifics of those constraints I don't know the specifics of the reservations.
16:23:54 <handsome_pirate> danofsatx:  Was this your topic?
16:24:13 <adamw> so i think what we really need at this point is some kind of solid proposal (even if it's just a 'run-it-up-the-flagpole' type thing) in front of the relevant people
16:24:16 <danofsatx-work> no, I didn't set any topics (that I know of)
16:24:26 <adamw> it's cmurf's RFC
16:24:30 <handsome_pirate> Ah
16:24:33 <handsome_pirate> cmurf:   Ahoy!
16:24:51 <danofsatx-work> but I did respond to cmurf's RFC ;)
16:25:01 <adamw> kicking it around on test@ is all well and good, but i think we're all more or less agreed that we want at least some kind of simplification of the guided path...but we don't get to decide that, so we need to get it in front of those who do, in a concrete way.
16:25:12 <adamw> i'm willing to try and start that, or leave it to cmurf - either way
16:25:59 <cmurf> Right. Well is there any concensus/agreement on dropping Partition Scheme pop-up menu as essentially useless?
16:26:32 * satellit 2 day old 21-24-1 boot.iso and koji lives do do custom installs for me
16:26:42 <bochecha_> join #fedora-meeting-2
16:26:42 <bochecha_> oops, sorry about that :)
16:27:06 <handsome_pirate> cmurf:  Aye, I don't see much point in it
16:27:21 <danofsatx-work> it is useless - I like to consider myself an above average intelligence user, and I find it very confusing and hard to use
16:27:29 <handsome_pirate> I either don't care about partition layout for throw-away vms and go with default, or I want to set up my own.
16:28:02 <cmurf> My thought is, if it's going to be useful, it should list use cases or workloads to help people get a better layout; otherwise it's just gobbdygook terminology which isn't helpful for the Guided/Automatic target market.
16:28:50 <roshi> I concur - if there was something saying "If you want to do to X choose fs-Y" or just have one default and go from there
16:29:54 <handsome_pirate> +1
16:30:12 <adamw> cmurf: I like the idea of dropping it entirely *in theory*. i suspect we won't be able to get there in practice.
16:30:27 <adamw> or, do you mean the specific UI design rather than the choice itself?
16:30:48 <cmurf> adamw: Both
16:31:04 <adamw> i think starting the discussion off with a very detailed issue about the implementation might be a tactical mistake and lead to the discussion getting bogged down
16:31:11 <danofsatx-work> adamw: the UI is wonky, but yeah, the choice shouldn't be there.
16:31:15 <adamw> i'd want to start it off at a more general level, but that may just be my preference
16:31:42 * cmurf is lost
16:31:51 * danofsatx-work is too
16:32:01 <danofsatx-work> lag is killing us
16:32:08 <cmurf> let's talk about it in fedora-qa later
16:32:28 <roshi> I think adamw means do you just want to use radio buttons instead of a dropdown - or no choice at all making the UI tool not matter
16:32:28 <danofsatx-work> might have a better synch on @test today
16:32:33 <handsome_pirate> I would say go with two choices:  Default or Custom
16:32:43 <handsome_pirate> Anything else is not necessary
16:32:51 <satellit> +1 radio buttons
16:33:06 <danofsatx-work> handsome_pirate: as adamw pointed out on the list, that choice already exists.
16:33:09 <adamw> again, i like that idea in concept, i'm worried especially the "don't want LVM" crowd would make it impractical in practice. but it might make a good trial balloon
16:33:18 <roshi> (or I just devolved the discussion more :) whee for bike shedding
16:33:31 <adamw> then we could 'compromise' on still having LVM/non-LVM choice available, but ditching "thinp" and "btrfs" at least
16:33:50 * adamw realizes discussing your Cunning Plan in a logged meeting may be a mistake :P
16:34:23 <danofsatx-work> well, if btrfs is default, isn't the LVM/non LVM choice a  moot point?
16:34:42 <cmurf> danofsatx-work: Yes
16:35:10 * danofsatx-work doesn't know enough about XFS yet
16:35:16 <adamw> cmurf: um, no?
16:35:27 <cmurf> Btrs on LVM makes zero sense as a default.
16:35:32 * handsome_pirate doesn't know how popular xfs would be
16:35:40 <adamw> cmurf: the question of what's the default and the question of how many non-default options we offer in non-custom path appears to be a separate one.
16:35:47 * handsome_pirate actually has not touched it outside of a rhel7 install
16:35:53 <adamw> making btrfs the default, right now, would simply consist of changing the default selection in the drop-down.
16:36:06 <roshi> are we still talking about Workstation or anaconda in general?
16:36:07 <adamw> it doesn't imply any of the other three choices goes away, *necessarily*. the two questions seem to be separate.
16:36:16 <danofsatx-work> handsome_pirate: that's where I first saw it too. RHEL has other issues, I haven't even delved into the FS yet
16:36:16 <adamw> i was on anaconda in general
16:36:16 <roshi> point adamw
16:36:18 <handsome_pirate> roshi:  I think anaconda in general
16:36:23 <cmurf> adamw: yes i understand that, ok look my basic premise has now been lost
16:36:48 <roshi> btrfs is default only (tentatively) on Workstation, AIUI - which is where my confusion came from
16:37:07 <cmurf> Either we think the Partition Scheme pop-up menu that uses terminology is incompatible with "newbies" who don't understand such terms, or we think large numbers of users who will use the default path actually understand those options.
16:37:07 <adamw> cmurf: OK, so let's get back to it. what's your default premise?
16:37:10 <adamw> sigh. basic premise.
16:37:31 <cmurf> I think 10% have any idea what the distinction is between LVM, Standard Partitions, Btrfs, and LVM Thin Provisioning.
16:37:36 * satellit I have trouble with formatting when I use anaconda f21  need to format HD in gparted then install..? otherwise fails
16:37:53 <cmurf> I think maybe 25% understand the distinction between LVM and Standard Partitioning
16:38:12 <synchris> i think the problem is between mbr and gpt sometimes
16:38:19 * handsome_pirate understands!  I am the 25% :)
16:38:23 <adamw> i'm wary of throwing numbers around without any data at all, but sure, as a theoretical premise.
16:38:26 <synchris> i retain hybrid format for my disks
16:38:46 <adamw> synchris: i don't see that it has anything to do with the current question.
16:38:56 <cmurf> The numbers are subjectively based off various user formums.
16:39:21 <adamw> cmurf: reasonable.
16:39:28 <synchris> adamw: maby it has to do with of what satellit days
16:39:30 <synchris> says
16:39:31 <cmurf> It has everything to do with it. Either that pop-up menu is useful or it's not useful. And I'm arguing it's compeltely useless because it's using terms almost no one understands.
16:39:44 <synchris> adamw: i doesnt has to do with urs
16:39:57 <adamw> cmurf: synchris brought up mbr vs. gpt. i was trying to avoid us chasing THAT yak around with a razor.
16:40:09 <danofsatx-work> cmurf: it's not only the terms - when I can't make it do what I want, it is useless.
16:40:13 <handsome_pirate> lol
16:40:27 <cmurf> it is a default layout that's it
16:40:33 <danofsatx-work> as I pointed out on the list, there was no way to tell it what PVs I wanted in what LV - at least not that I could find.
16:40:39 <adamw> cmurf: so, are you working from "many people don't understand the choice offered" to "the choice should be pushed down the custom install path"?
16:40:41 <cmurf> It's not going to work for a lot of people.
16:41:03 <adamw> danofsatx: again, that's really not on topic for the current question. sorry to be a topic nazi, but discussing the installer design is a really complex area and we have to stay focused
16:41:28 <danofsatx-work> I know, I know....I'll sit down and shut up.
16:41:30 <cmurf> adamw: I'm saying *IF* you want a pop-up menu of choices in the Guided path, *make them useful*. And right now they are useless.
16:41:35 <synchris> adamw: sorry for being out of topic
16:41:40 <adamw> we're discussing the guided path, where you don't get to make decisions that detailed and it's not likely that you will; if the custom path doesn't work well for you that's obviously an issue, but it's not what we're taling about ATM
16:42:03 <cmurf> adamw: So that suggests two paths: Get rid of the pop-up menu and go with one default layout. Or change the options to be use case or workload driven rather than technobabble driven.
16:42:15 <adamw> cmurf: OK. i'd just be a bit worried that broadening out the scope like that is an invitation to bikeshedding, which results in 'well, everyone has a different idea what should be in the dropdown so let's just leave it how it is'...
16:42:28 <adamw> but yeah, i think you're on track with the basic idea
16:42:38 <cmurf> adamw: This is why monarchy can be much more efficient.
16:42:45 <adamw> =)
16:42:51 <adamw> so long as I get to be king!
16:43:09 <cmurf> I don't care who is king or queen, just someone make a decision and execute the naysayers.
16:43:11 * tflink shudders at the thought of adamw as king
16:43:20 <adamw> cmurf: so, are you interested in taking it forward with a concrete proposal?
16:43:26 <adamw> execution? CAN DO
16:43:31 <nirik> so, could this be a place products could want specific things for their product?
16:43:33 <handsome_pirate> adamw:  What, you're not firing tflink over that?
16:43:34 <adamw> someone make me a .guillotine macro
16:43:40 <drago01_> tflink: prepare the pitforks just in case ;)
16:44:10 <adamw> nirik: i think that might wind up being a detailed question...
16:44:11 <danofsatx-work> pifork? is that what they use for BBQ pits?
16:44:18 <tflink> drago01_: pitchforks and torches - can't have a good mob without torches :)
16:44:30 <drago01_> ;)
16:44:35 <adamw> nirik: there's quite a difference between "product X and Y have different defaults and different choices in the drop-down" and "product X has the drop-down but product Y doesn't", for instance...
16:44:36 * nirik isn't fully sure which popup you are talking about, would have to go do an install to recall.
16:44:56 <adamw> nirik: in current F20, there's an Installation Options popup after you pick disks
16:45:05 <handsome_pirate> tflink:  adamw doesn't eat bacon, as I recall; we could toss him in a pit of yummy, crispy bacon
16:45:15 <cmurf> QA is already doing a lot of testing in Manual/Custom because of the limitations of Guided which already has four options in it. So it's kinda like, I don't see our workload increases by dropping four options in Guided. At least we have a one shot default.
16:45:22 <cmurf> … that works.
16:45:35 <adamw> it has a dropdown where you can pick LVM, LVM thinp, ext(4), or btrfs partitioning (IIRC) strategies
16:46:07 <handsome_pirate> Anyway, probalby about time to move on; we've kicked this horse around some now
16:46:08 <adamw> in F21 the IO popup has been basically merged down into the disk selection screen, but the partitioning strategy choice box is still there.
16:46:42 <adamw> if you pick one and then go for guided partitioning, that's the one that gets used: if you pick one and go for manual partitioning, it becomes your default for the manual partitioning screen.
16:46:50 * cmurf is surprised no one is arguing "if you don't understand the other options in that pop-up, then why are you changing the setting?"
16:47:00 <nirik> alright.
16:47:36 <handsome_pirate> #topic installer discussion: variant installers?
16:47:52 <handsome_pirate> cmurf:  Is this one yours, as well?
16:48:17 <cmurf> handsome_pirate: I'm contrarian so pinning me down is not easy.
16:48:19 <danofsatx-work> yup, the horse is dead.
16:48:20 <adamw> handsome_pirate: no, i'd rather have a concrete plan of action here
16:48:20 <adamw> i.e. someone gets an action item to take this forward outside of QA
16:48:20 <adamw> we've kicked it around enough, like you said, so now we need to *do* something :)
16:48:20 <adamw> if cmurf or anyone else wants the action item, great: if not i'll take it. but we shouldn't delay it any longer
16:48:45 <handsome_pirate> cmurf:  Up for it?
16:49:20 <cmurf> adamw: if you want to do a draft that includes what's in and out of scope, then I'll fill in the blanks with some arguments supporting the position and against the predicted opposition :-P
16:49:32 <tflink> handsome_pirate: that's ... disgusting and fits the definition of cruel and unusal :-P
16:50:07 <handsome_pirate> #action adamw to draft up a proposal to the Anaconda devs about guided partitioning screen
16:50:18 <handsome_pirate> #action cmurf to fill in blanks on the proposal
16:50:29 <cmurf> yay action
16:50:30 <handsome_pirate> tflink:  I wouldn't find it cruel :)
16:50:37 <handsome_pirate> Okay, moving on
16:50:43 <handsome_pirate> Alternate installers?
16:51:06 <cmurf> sounds like that's answered as a likely yes, but I don't know how it manifests
16:51:21 <cmurf> Will there be different Live's for each product?
16:53:28 <nirik> we don't know yet. ;) but I think likely would be:
16:53:48 <nirik> cloud just ships an image, workstation ships a live, server ships an install iso.
16:55:25 <cmurf> Or can we combine products on one LiveDVD and make it an installer choice?
16:55:26 <cmurf> netsplit time?
16:55:26 <handsome_pirate> cmurf:  That's a question that I don't think has been touched yet
16:55:26 <handsome_pirate> cmurf:  Want to find out?
16:56:41 <nirik> cmurf: dunno. That might be nice for ambassadors... especially if you tossed on the other desktop spins.
16:56:57 <nirik> but could get pretty complex.
16:57:01 <cmurf> OK so we need to hear back from the WG's on this question I think. I'd like to see consolidation of media for Server and Workstation.
16:57:13 <cmurf> A LiveDVD that contains both and choose which one at install time.
16:57:17 <cmurf> And the same happens for Netinst
16:57:22 * satellit end up with a 8 GB USB installer?
16:58:02 * satellit or Double Layer DVD...
16:58:14 <mjg59> cmurf: That sounds pretty awful
16:58:40 <mjg59> cmurf: Separate products benefit from being, well, separate products
17:00:33 <satellit> netsplit...
17:00:58 <danofsatx-dt> yup, split
17:02:43 <danofsatx-dt> we're wrapping the meeting up in the "other" meeting channel...
17:03:06 <danofsatx-dt> to continue in #fedora-qa if Freenode ever catches up to us
17:04:03 <satellit> other?
17:05:13 <danofsatx-dt> I have two clients up on two servers - most of us stayed connected to the original channel.
17:05:19 <danofsatx-dt> this is the split one :(
17:06:47 <satellit> I have 2 PC's and both are split here....
17:07:00 <handsome_pirate> danofsatx:  No, it has to be in this channel
17:07:00 <handsome_pirate> hrm
17:07:06 <danofsatx> looks like we're all back together
17:07:11 <handsome_pirate> #info netsplits causing meeting issues
17:07:24 <handsome_pirate> #topic taskotron update
17:07:33 <handsome_pirate> tflink:  Ping
17:08:11 <handsome_pirate> tflink:  Want to talk about this, or shall we go ahead and end the meeting due to freenode borkage?
17:08:33 * satellit both working here
17:08:40 <danofsatx-dt> I say end it. pick up later in qa or blockerbug channel
17:08:48 * handsome_pirate doesn't see a response, so
17:09:02 <handsome_pirate> #info Ending meeting due to irc issues
17:09:06 <handsome_pirate> #endmeeting