15:00:43 <adamw> #startmeeting Fedora QA meeting
15:00:43 <zodbot> Meeting started Mon Apr 14 15:00:43 2014 UTC.  The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:43 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:48 <adamw> #meetingname fedora-qa
15:00:48 <zodbot> The meeting name has been set to 'fedora-qa'
15:00:48 <spstarr_work> Hello :)
15:00:51 <adamw> #Topic Roll call
15:00:54 <adamw> ahoyhoy
15:00:58 <spstarr_work> finally :)
15:01:40 <kparal> http://kamikazeegirl.files.wordpress.com/2011/02/hold-a-meeting.gif
15:01:40 * roshi is here
15:01:48 * jreznik is around if needed
15:01:54 * pschindl is here
15:02:02 * mkrizek lurks
15:02:29 * spstarr_work is here
15:02:47 * adamw is barely here
15:03:07 <adamw> kparal: i expect you brought the donuts, then, smartypants? :)
15:03:42 * pingou joins for the Donuts
15:03:43 * roshi doesn't see any people - just words he thinks might have people behind them...
15:03:55 <mpduty> .fas mohanprakash
15:03:55 <zodbot> mpduty: mohanprakash 'mohan prakash' <mpduty@gmail.com>
15:03:57 <pingou> roshi: the cake is a lie
15:04:11 <roshi> it's so true
15:04:16 <spstarr_work> I like danishes, please, make sure it's apple cinnamon with a little icing sugar.
15:04:22 <spstarr_work> thanks :)
15:04:33 <adamw> it's true that the cake is a lie?
15:04:58 <roshi> yup
15:04:58 <kparal> that's a lie
15:05:24 <pingou> there is no spoon to eat the cake anyway
15:05:27 <adamw> i don't believe you
15:05:38 <adamw> bah, who eats cake with a spoon?
15:05:44 <adamw> dessert forks is where it's at, yo.
15:05:45 <roshi> you never catch the dragon adamw
15:06:02 <spstarr_work> adamw: if its an ice cream cake....
15:06:07 <spstarr_work> a spoon is acceptable :P
15:06:11 * tflink is present
15:06:15 <roshi> I always eat my fictional cake with chop sticks - just to be *that guy*
15:06:17 * pwhalen is here
15:06:30 <adamw> spstarr_work: in that case you should use a knife
15:06:34 <adamw> ahoyhoy pwhalen
15:06:35 <spstarr_work> lol :)
15:06:44 <adamw> okely dokely
15:06:50 <adamw> #topic Previous meeting follow-up
15:07:01 <pwhalen> adamw, happy monday :)
15:07:21 <adamw> #info "adamw to put out the initial rawhide validation testing matrices and announcement mails" - obviously, I did that, it's the next topic. https://lists.fedoraproject.org/pipermail/test/2014-April/120931.html
15:07:27 <adamw> #topic Rawhide validation testing
15:07:39 <adamw> so, just wanted to throw in a topic for this, see if there are any kinks to work out
15:08:27 <adamw> from the list, it looks like we should edit the Rawhide page and the matrix introduction to explain the 'appliance builds' better
15:09:10 <spstarr_work> erm Storage device tests ... missing x86_64?
15:09:39 <adamw> spstarr_work: 'x86' means i686 or x86_64
15:09:47 <adamw> you're the second person to ask that, so i guess it needs explaining too...
15:10:11 <spstarr_work> x86 used to mean i386/486/586/
15:10:40 <spstarr_work> ia32 is now i386/486/586... x64 is... 64bit... i guess x86 is both now
15:10:48 * spstarr_work shakes fist
15:11:30 <spstarr_work> adamw: I am officially, unconfused
15:11:59 <adamw> i would just like everyone to say x86_32, x86_64 and x86 (generic) from now on
15:12:00 <adamw> thank you
15:12:07 <adamw> :P
15:12:17 <adamw> looks like i need to sort out some kinks with using download.fedoraproject.org for links too
15:12:29 <spstarr_work> proclaimed, and therefore it is :)
15:12:45 <kparal> everyone uses it a bit differently, unfortunately. microsoft uses x86 and x64
15:12:59 <adamw> #action adamw to explain 'appliance images' and 'x86' for the new matrices (and Rawhide page, for appliance images), and look into issue with download.fp.o mirrors sometimes missing Rawhide boot.iso
15:13:21 <adamw> so this meeting is super productive for me, i hope everyone else is enjoying the donuts ;)
15:13:55 * kparal is enjoying wiener schnitzel
15:14:04 <spstarr_work> lol
15:14:12 <spstarr_work> adamw: the test cases look good
15:14:15 <adamw> kparal: and beer?
15:14:18 <adamw> spstarr_work: thanks!
15:14:31 <kparal> adamw: no! I'm at work, of course
15:14:41 <kparal> that's forbidden, you know
15:14:42 <tflink> kparal: donuts == weiner schnitzel where you're from? :)
15:14:43 <adamw> kparal: oh man, you said that with a straight face. i'm impressed.
15:14:51 <spstarr_work> i will begin beating on installation ones
15:14:51 <adamw> (although I suppose it's easier over IRC.)
15:15:02 <roshi> true
15:15:12 * adamw imagines brno office breaking down in tears of laughter
15:15:50 <spstarr_work> question
15:16:18 <spstarr_work> and i think you elude to this in the email but once fedora.next comes,do we know what test cases these will add to rawhide testing?
15:16:42 <adamw> we didn't work it out entirely yet
15:16:47 <adamw> although now would probably be an excellent time to start...
15:16:55 <adamw> the products are fairly defined at this point
15:17:17 <adamw> maybe we can talk about it in open floor, shouldn't take long to get there
15:17:17 * FranciscoD is here and will quietly observe
15:17:28 <roshi> it's not totally apparent to me what rawhide is going to look like with fedora.next
15:17:50 <adamw> i don't see why it'd look any different
15:17:56 <adamw> well, maybe it'll have a playground repo
15:18:04 <adamw> FranciscoD: have a schnitzel
15:18:09 <roshi> will it roughly equate to being base on branch?
15:18:22 <adamw> ?
15:18:47 * FranciscoD noms
15:18:50 <roshi> meaning when we branch from rawhide for 21, will the branch be the BaseWG product?
15:19:04 <adamw> aiui the products are built from the repos.
15:19:19 * roshi could also be speaking nonsense - it's still the morning for me and I haven't had coffee
15:19:22 <adamw> i don't think we're planning for them to have their own unique repos at this point.
15:19:37 <roshi> ah, ok
15:19:37 * spstarr_work pours roshi coffee
15:19:45 <roshi> thanks :)
15:19:57 <adamw> might wanna check with mattdm or someone, but htat's my understanding
15:20:02 <adamw> so rawhide would still be rawhide, pretty much
15:20:11 <adamw> OK, let's blow through the agenda so we can get to open floor fun :)
15:20:14 <adamw> #topic Heartbleed status
15:20:21 <spstarr_work> isn't this all just comps.xml voodoo though its just addons to the overall structure?
15:20:30 <adamw> so just thought i'd throw this in since it was the big topic last week
15:20:41 <roshi> it was the topic last week
15:20:46 <FranciscoD> Is there anything left to do with heartbleed?
15:20:47 <adamw> spstarr_work: it may wind up involving other mechanisms too, but conceptually, that's a decent way to think about it.
15:20:53 * FranciscoD thought the storm had passed
15:21:21 <adamw> FranciscoD: in fedora QA terms nothing huge
15:21:34 <adamw> but we did do updated images for Fedora 20, with the new openssl included
15:21:43 <adamw> we probably ought to double check that those all work as expected
15:22:14 <adamw> dgilmore: around? which images wound up getting respun in the end?'
15:22:16 <spstarr_work> the RPMs from koji seem ok
15:22:23 <jreznik> and as a result of it, cloud wg submitted this change - https://fedoraproject.org/wiki/Changes/%28A%29Periodic_Updates_to_Images - and it would require attention from QA
15:22:24 <adamw> #chair roshi spstarr_work
15:22:24 <zodbot> Current chairs: adamw roshi spstarr_work
15:22:28 * adamw throws chairs around
15:22:35 * spstarr_work ducks
15:22:53 <adamw> jreznik: wow, that's kind of huge. thanks.\
15:23:09 * jreznik is still laughing after reading backlog from 17:15 adamw's comment :)
15:23:21 <adamw> hehe
15:23:36 <spstarr_work> those cloud folks always make things.. cloudy ;p
15:23:38 <spstarr_work> :)
15:23:46 <FranciscoD> I expect "emergency updates" like the openssl one would require all media to be respun, like this time.
15:23:58 <jreznik> FranciscoD: all?
15:24:27 <adamw> FranciscoD: I don't think we did respin all images
15:24:40 <FranciscoD> ah, I thought we did
15:24:42 <adamw> dgilmore and i were holding out for only doing the cloud image, but i think we wound up doing live too.
15:24:51 <adamw> (client side vulnerability, blah blah.)
15:24:54 <FranciscoD> aye
15:25:48 <FranciscoD> I think I'd glanced over a trac ticket notification that required updates to the image checksums on fp.o. I just assumed it was all images.
15:26:20 <adamw> so yeah, anyhow, we should check the respun images
15:26:39 <adamw> anyone want an action item to figure out which got respun and throw up a sanity check matrix, or should I do it?
15:27:06 * roshi can
15:27:11 <FranciscoD> I can check which ones got respun, but don't know anything about the sanity check part
15:27:44 <pwhalen> not seeing updated images, are they expected to be posted now?
15:28:19 <adamw> #action roshi to set up mini validation matrix for heartbleed-respun Fedora 20 images
15:28:22 <adamw> thanks roshi!
15:28:27 <adamw> pwhalen: I thought they were, yeah
15:29:54 <adamw> did anyone have any other heartbleed-y concerns, or shall we move on?
15:30:24 <roshi> in case anyone hasn't/isn't - change all your passwords
15:30:37 <roshi> </PSA>
15:31:29 <jreznik> roshi: it's probably useless doing it now for many sites now, even for fedora webs
15:32:01 <jreznik> and yes, it's my password(s), I don't want to change it(them) ;-)
15:32:12 <adamw> #info everyone change your passwords, and mail a copy of the new ones to the NSA. just so everyone's clear on where they stand.
15:32:17 <roshi> well, if they got your password from memory on an affected service - then the updated openssl won't protect you from that
15:32:35 <roshi> haha
15:32:47 <spstarr_work> heh
15:33:26 <adamw> #topic Open floor
15:33:34 <adamw> okey dokey, so, what were we talking about
15:33:39 <spstarr_work> well, that sure didn't take long to get to Open Floor :)
15:33:39 <adamw> ah, yeah, .next
15:33:57 <spstarr_work> yes .next
15:34:19 <adamw> my vague idea for how to approach this was to try and quantify what ought to be done in theory in a perfect universe to validate all the .next products, and then get down to the icky business of deciding a) how much of it we can do and b) who gets to do it
15:34:33 <spstarr_work> I think from what I know there's only a limited set of 'roles' being created for Fedora 21?
15:34:41 <adamw> roles are a Server product thing
15:34:58 <spstarr_work> well, we have to test that too, no?
15:34:59 <spstarr_work> :)
15:35:03 <adamw> sure
15:35:14 <adamw> just saying, it's not something that applies to Workstation or Cloud
15:35:23 <spstarr_work> right
15:35:25 <spstarr_work> well
15:35:31 <spstarr_work> CLoud might have roles?
15:36:40 <adamw> possibly? i'm still not sure exactly how cloud and server are supposed to interact.
15:36:41 <adamw> anyhow
15:36:57 <spstarr_work> hmm
15:37:02 <adamw> i'm thinking maybe we should detail someone to work with each WG to come up with a rough test plan
15:37:14 <adamw> i can do server, which leaves workstation and cloud
15:38:02 <adamw> we had someone liaising with each group, anyone want to own up to it being them? :)
15:38:10 <roshi> I don't have a ton of experience with cloudy things - but I've been following their work thus far
15:38:19 <spstarr_work> I can help
15:38:25 <roshi> so I can work with cloud for the test plan
15:38:30 <spstarr_work> OpenStack I know nothing of yet though
15:38:54 <adamw> thanks
15:39:09 <adamw> #action adamw to draft a test plan for Server product
15:39:17 <adamw> #action roshi to draft a test plan for Cloud product
15:39:29 <FranciscoD> I've been following the workstation SIG. I can give it a try. Expect a lot of queries, though. I have almost no actual QA experience.
15:39:39 <adamw> FranciscoD: that's fine, neither do we :P
15:39:47 * FranciscoD generally just sticks to updates-testing and other *easy* stuff
15:39:47 <spstarr_work> lol
15:39:48 <adamw> #action FranciscoD to draft a test plan for Workstation product
15:39:51 <FranciscoD> adamw: yeah right :P
15:40:02 <adamw> oh, i have no clue what i'm doing
15:40:08 <adamw> i live my life wondering when y'all are going to notice
15:40:13 <adamw> well, except viking-ice, he figured it out years ago
15:40:20 <FranciscoD> XD
15:40:28 <spstarr_work> i guess we just need to understand 'what' the tests will be vs just installation of them, but running though the typical 'types' of configurations
15:40:52 <roshi> I'm sure we'll go through some iterations for each
15:40:52 <adamw> FranciscoD: don't sweat it - the idea is really simply just to figure out, roughly, the scope of what needs testing before the release of the product
15:41:05 <FranciscoD> aye
15:41:18 <adamw> i'm not expecting anything in super-formal-QA-ese, don't worry too much about the 'official definition' of a test plan or anything, write it out however's most convenient to you
15:41:32 <FranciscoD> I'll look over the stuff we used before
15:41:47 <FranciscoD> Oh, and obviously, I'll steal from whatever you folks write up ;)
15:42:02 <roshi> this is open source - *that's what we do!*
15:42:05 <spstarr_work> adamw: I guess a good question is, what limit of testing?
15:42:07 <adamw> we used to draw up 'test plans' for installation each cycle, but i kinda stopped doing it lately as it never seemed very useful, but in case you do find it useful as a template: https://fedoraproject.org/wiki/Category:Install_Test_Plans
15:42:18 * FranciscoD notes links
15:42:20 <adamw> i tend to think of the matrices as the real 'test plans', for current Fedora
15:42:22 <spstarr_work> adamw: installation, sure, but actual running of apps and testing if we can crash the GUI? (for Workstation)
15:42:28 <spstarr_work> what is QA's scope for testing
15:42:40 <spstarr_work> say, LibreOffice
15:42:45 <adamw> so for Fedora.current (as opposed to Fedora.next) the 'test plan' is just what's in the Installation, Base and Desktop matrices
15:42:45 <spstarr_work> you could write a book on it's test cases!
15:42:49 <adamw> spstarr_work: sure
15:42:50 <adamw> exactly
15:43:04 <adamw> spstarr_work: so where i'd start is to look at the PRD
15:43:06 <adamw> each product has one
15:43:23 <roshi> aiui it would just be a mapping between our current matrices and which release criteria apply to which WG product
15:43:30 <adamw> look at each element of the PRD as a box that needs checking, and just document what we'd need to do to check it
15:43:51 <adamw> roshi: i was kinda expecting we'll need to extend coverage
15:43:53 <spstarr_work> so we're talking more higher-level testing
15:44:11 <adamw> the actual testing isn't all necessarily going to be on QA, but we need to figure out what (at least in an ideal world) needs to be tested by *someone* to 'validate' a Fedora.next release
15:44:17 <spstarr_work> it installs, it runs, it can be stopped/started, it doesnt crash doing basic X tasks
15:44:41 <adamw> so, you know, for Server, we need to check that the 'supported' roles deploy and upgrade cleanly, we need to  test the role manipulation tools, maybe we test the firewall works as described in the PRD, that kinda thing
15:44:48 <spstarr_work> (that sounds more sanity testing though vs indepth)
15:44:51 <roshi> true - but the baseline would be scrounging the relevant things frmo the matrices and release criteria - then extend from there, right?
15:45:02 <adamw> spstarr_work: https://fedoraproject.org/wiki/Workstation/Workstation_PRD
15:45:05 <spstarr_work> adamw: ok so those are more in-depth
15:45:10 <adamw> roshi: that'd be one way to start, sure
15:45:31 <adamw> roshi: like i said i'd probably aim to start from the PRD; that defines what the product's supposed to be, so that's what we have to test
15:45:54 <roshi> makes sense
15:45:55 <spstarr_work> adamw: ok so we have the basic cases, now we can create test cases based on those
15:45:59 <adamw> roshi: it would indeed be useful to note where stuff maps to an existing criterion / test case
15:46:08 <spstarr_work> 'Desktop Apps: Up to date desktop with email client, browser, productivity suite, messaging, and a complete set of desktop apps and utilities. Desktop apps should be sufficient to make this system the developer's only computer. '
15:46:23 <roshi> I was just thinking of a good starting point - so it wouldn't feel like starting from scratch
15:46:28 <spstarr_work> so validate evolution (or kmail) work fine, firefox works fine, LibreOffice works fine...
15:46:35 <adamw> spstarr_work: *in an ideal world*, yep
15:46:38 <spstarr_work> but then we have to define the basic tasks
15:46:43 <adamw> of course, the next step is to figure out how much of it we can do :P
15:46:58 <spstarr_work> :)
15:47:17 <adamw> i'll try and get a basic rough draft of the server one in soonish, so you can get an idea of what i'm blithering about
15:47:26 <adamw> and then either follow that, or explain why i'm an idiot and we should do it differently
15:47:39 <roshi> wfm
15:47:44 <adamw> thanks for volunteering, you fools^H^H^H^H^Hresponsible citizens
15:48:04 <roshi> :)
15:48:16 <roshi> for certain definitions of responsible - I guess :p
15:48:18 <adamw> anything else for open floor? was there something else we left hanging?
15:48:33 <kparal> wouldn't want to start any flamewar here, but schnitzels are maybe even better than... bacon
15:48:50 <roshi> I was wondering if anyone but adamw and kparal checked out the test maps demo I had
15:48:50 <adamw> .fire kparal heresy
15:48:51 <zodbot> adamw fires kparal heresy
15:49:02 <adamw> roshi: ooh yeah.
15:49:03 <spstarr_work> ;)
15:49:09 <roshi> relevant email here: https://lists.fedoraproject.org/pipermail/test/2014-April/120811.html
15:49:26 <adamw> #info roshi's 'test maps' demo is at http://188.226.194.38/ , check it out and give him feedback
15:49:41 <adamw> that sketchy-looking IP address is absolutely not a client-side heartbleed attack, honest
15:50:04 <roshi> it's one of my digital ocean droplets - it won't be there forever :)
15:50:29 * roshi didn't want to buy a domain if it was an idea nobody liked or cared for
15:51:14 <adamw> the idea is to help you run logical sets of validation test cases together, or in sequence
15:51:26 <adamw> for the greater good of efficient testing of the workers' product! comrades!
15:51:48 <roshi> and eventually be able to have a CRUD for testcases and stats for measuring results
15:52:09 <tflink> roshi: we could setup a dns entry for it, though
15:52:18 <tflink> so it isn't just a bare ip address
15:52:27 <roshi> true - hadn't thought that far :)
15:52:28 * FranciscoD feels his brain shutting down.
15:52:37 <adamw> FranciscoD: QA meetings will do that to you.
15:52:40 <FranciscoD> I should head to bed. Good night folks. Happy testing :)
15:52:41 <roshi> then it wouldn't feel like as much like a nasty hack
15:52:42 <adamw> after a while, you learn not to turn it on.
15:52:45 <adamw> night!
15:52:55 <FranciscoD> adamw: oh, today it's just the flu and a wet miserable night :/
15:52:59 * FranciscoD dozes off
15:53:45 <adamw> welp, sounds like we're all checking out
15:53:56 <adamw> that was a good hard work week folks, have a beer, we earned it
15:54:16 <spstarr_work> :)
15:54:18 <spstarr_work> mine is going to be fun
15:54:28 <adamw> heartbleed 2: electric boogaloo?
15:55:04 * adamw sets quantum fuse
15:56:11 <adamw> thanks for coming, everyone!
15:56:16 <adamw> #endmeeting