20:07:25 <MrsB> #startmeeting
20:07:25 <Inigo_Montoya> Meeting started Thu Feb 13 20:07:25 2014 UTC.  The chair is MrsB. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:07:25 <Inigo_Montoya> Useful Commands: #action #agreed #help #info #idea #link #topic.
20:07:53 <MrsB> Hi everybody welcome to another meeting
20:08:08 * MrsB looks around, who's going to get it
20:08:19 <MrsB> #chair wilcal
20:08:19 <Inigo_Montoya> Current chairs: MrsB wilcal
20:08:21 * Kernewes hasn't seen the topics
20:08:24 <MrsB> :D
20:08:31 <wrw105> Hi, My name is Bill and I'm a Mageiaholic.
20:08:32 <wilcal> QA a bit over worked I'd say
20:08:53 <wilcal> No David today?
20:09:06 <lewyssmith> He must be bad.
20:09:12 <MrsB> not heard from him this week, I'll drop him an email tomorrow to see how he's doing
20:09:27 <lewyssmith> Send our sympathies.
20:09:31 <MrsB> #topic Who's new?
20:09:34 <MrsB> will do :)
20:09:41 <MrsB> Is there anybody new here today?
20:09:51 <MrsB> anyone joining us for the first time
20:10:18 <MrsB> doesn't look like it
20:10:32 <wilcal> For the record I said, and others too, good things about QA at the General meeting at FOSDEM.
20:10:40 <MrsB> #topic Testing Updates
20:10:45 <MrsB> that's good to hear wilcal
20:10:51 <wilcal> Wow testing updates WOW
20:11:01 <MrsB> surprise eh?
20:11:07 <MrsB> morning akien
20:11:25 <wilcal> Difficult to sort the really important ones from the not so important
20:11:39 <wilcal> I felt bad about filing another BUG today
20:11:41 <Kernewes> no it isn't, the important ones are the ones I can do
20:11:46 <Kernewes> i.e. hardly any
20:11:59 <lewyssmith> Doesn't the bold type help?
20:12:05 <wilcal> Yes
20:12:21 <MrsB> #info most of the updates have a procedure now
20:12:45 <MrsB> #info So priority, the major one right now is pkcs11
20:13:07 <lewyssmith> Just done my little bit!
20:13:10 <MrsB> that's the one which should cure all the issues for nvidia people with applications not starting
20:13:24 <lewyssmith> It' made for Bill.
20:13:35 <wilcal> Yes, I am deep into the Nvidia thing
20:13:36 <MrsB> i saw lewis thanks. We need anybody with nvidia and mga4 to test that please
20:13:52 <wilcal> https://bugs.mageia.org/show_bug.cgi?id=12696
20:13:53 <wilcal> empty /etc/pkcs11/pkcs11.conf breaks some apps with drivers nvidia
20:13:54 <Kernewes> I've only got Nvidia with Mga3
20:13:55 <[mbot> Bug 12696: major, High, qa-bugs, NEW , empty /etc/pkcs11/pkcs11.conf breaks some apps with drivers nvidia, p11-kit-0.20.1-3.mga4.src.rpm
20:14:10 <MrsB> it'll be obvious if it has worked, everything that didn't start before should now start
20:14:11 <wilcal> It's great to see other people jumping on that one
20:14:35 <wrw105> I can take a look at that one...just tried kwrite and it didn't start.
20:14:53 <wilcal> Fix is easy but it's a big problem
20:15:02 <MrsB> There are some security updates too tntnet and cxxtools can be tested together. Info on the tntnet bug how to do so.
20:15:19 <wilcal> Why does that one conf file do what it's ( not ) doing? It's an empty file
20:15:25 <lewyssmith> MrsB: Have just looked. Appreciate your notes.
20:15:52 <MrsB> Most of them have a test procedure listed now, you can always see which have on the madb page
20:15:56 <lewyssmith> wilcal: After the update?
20:15:58 <RemyServices> I'll be trying the tntnet set tonight
20:16:16 <wilcal> I was DEEPLY involved in the ffmpeg thing this morning
20:16:27 <wilcal> ffmpeg update to 2.0.3
20:16:29 <wilcal> https://bugs.mageia.org/show_bug.cgi?id=12698
20:16:32 <[mbot> Bug 12698: normal, Normal, qa-bugs, NEW , ffmpeg update to 2.0.3, ffmpeg-2.0.2-2.mga4.src.rpm
20:16:36 <MrsB> great, the gadu one is dead easy to test too so please have a go at it while you're doing stuff
20:16:37 <tmb> wilcal, thats the bug... it should ignore an empty file, but instead it crashes / hangs on it
20:16:46 <RemyServices> I keep looking at libpng to set it, but it is only i586, is there a way to mark 64 testing as completed or unneeded?
20:17:27 <MrsB> RemyServices: it's always both architectures
20:17:47 <MrsB> 64bit libs are actually lib64 but thats the only didfference
20:18:06 <tmb> (unless its a "noarch" package)
20:18:30 <MrsB> technically yes, but we still at least ensure it's built in both medias
20:18:37 <wilcal> The ffmpeg thing indicates some type of MCC listing problem??
20:18:48 <RemyServices> MrsB: I'll take this offline then, I was having issues finding it, ill have another look tonight
20:19:07 <tmb> MrsB, true
20:19:10 <MrsB> Hard to say wilcal, I didn't follow what you were trying to say there
20:20:17 <MrsB> what exactly happened with it?
20:20:39 <wilcal> M4 i586 libavformat works just fine but M4 x86_64 lib64format does not get selected in the MCC
20:21:04 <wilcal> If I install it manually it works fine. Anyway me an luigi are work'n it
20:21:10 <MrsB> it will only be selected for update if it is actually installed first
20:21:53 <wilcal> I've another difficult BUG I'm working
20:22:03 <wilcal> https://bugs.mageia.org/show_bug.cgi?id=12137
20:22:05 <wilcal> Remote CUPS server not being detected automatically
20:22:06 <[mbot> Bug 12137: normal, release_blocker, mageia, ASSIGNED , Remote CUPS server not being detected automatically, only when searching for hostname, system-config-printer
20:22:07 * MrsB needs to get mga4 installed
20:22:34 <MrsB> I'm CC'd on that one. It's not ready for QA yet though.
20:22:40 <wilcal> then I found a problem with the flash-player-installer this morning
20:22:57 <MrsB> what kind of problem?
20:23:11 <wilcal> Anywa
20:23:13 <wilcal> https://bugs.mageia.org/show_bug.cgi?id=12759
20:23:14 <wilcal> Incompatible architecture for flash installer rpm
20:23:16 <[mbot> Bug 12759: normal, Normal, bugsquad, NEW , Incompatible architecture for flash installer rpm, flash-player-plugin-11.2.202.336-1.mga4.nonfree.x86_64.rpm
20:23:46 <MrsB> I suspect your medias are not configured correctly. I'll help you later to check if you like
20:24:27 <wilcal> They work fine with maybe 50 other rpm installs
20:24:33 <wilcal> this morning
20:24:54 <wilcal> next step is to resync and try again
20:25:06 <wilcal> Lotta stuff going on
20:25:07 <tmb> wilcal, you reported it against i586, but try to install an x86_64 ??
20:25:13 <MrsB> ok, worth a check though maybe anyway. Later though, let's keep to the updates we already have :D
20:25:18 <wilcal> will do tmb
20:25:35 <Stormi> arch is easy to overlook in bugzilla :)
20:25:35 <wilcal> Ya I hated to open that BUG
20:25:51 <MrsB> If anybody has forgotten..
20:25:53 <MrsB> http://mageia.madb.org/tools/updates
20:25:58 <MrsB> This is where to find them
20:26:01 <[mbot> [ Mageia App Db - Current Update candidates ]
20:26:01 <Stormi> what's that page MrsB?
20:26:15 <MrsB> the longest page anywhere on the internet Stormi
20:26:19 <wilcal> What are the plans of the developers to help with backlog
20:26:32 <MrsB> morning :)
20:26:40 <lewyssmith> Very +ve
20:27:11 <MrsB> Well akien and stormi are helping out, colin has a bit too but made us another update in the progress :D
20:27:12 <tmb> wilcal, as stated we'll try to help flush them by the end of the week (meaning Sunday, not Friday)
20:27:41 <wilcal> Many many thanks
20:27:50 <Stormi> even Luigi12 tested an update
20:27:55 <MrsB> It doesn't mean we can sit back and wait for packagers to do our job for us though
20:27:55 <wilcal> Kool
20:27:56 <tmb> as many of us are swamped with $dayjob during weekdays
20:28:01 <dvg_l> i am about to try and start helping, if i can understand what is needed
20:28:06 <MrsB> thats true Stormi!
20:28:12 <Luigi12> Stormi: yeah :o)  I wanted to test more of them, but then I realized I packaged almost all of them :o(
20:28:21 <Stormi> and neoclust all the mate updates a few days ago
20:28:27 <MrsB> dvg_l: if in any doubt please just ask, don't sit and get frustrated
20:28:36 <dvg_l> will do
20:28:38 <MrsB> yes neoclust did too, true
20:29:01 <Stormi> Luigi12: MrsB: I'd like to raise the "the packager can't test his own update" part of the policy actually, today or another time
20:29:11 <MrsB> It's still our burden to carry so please everybody dig in and lets get on top of these again
20:29:22 <lewyssmith> Extra help has been very impressive.
20:29:49 <MrsB> Stormi: you're pro that aren't IIRC
20:29:56 <MrsB> +you
20:29:58 <wilcal> I think the list is getting shorter
20:30:21 <Stormi> MrsB: yes, but I can detail my view later in a specific topic if we have time
20:30:27 <MrsB> it is, but only through alot of hard work, please can we all make sure we do our part too
20:30:35 <MrsB> sure Stormi
20:30:54 <wilcal> I'd like to get that ffmpeg bug outta there
20:31:02 <MrsB> #info most of the remaining updates have a procedure on the bug now
20:31:20 <lewyssmith> This is really helpful, claire.
20:31:42 <MrsB> They're mostly found through clicking the Bugzilla link on madb though lewis
20:32:02 <MrsB> if you see one without a procedure that's the best place to look
20:32:07 <Stormi> yes, that's what I did a few days ago for a bunch of updates, I knew that "yes" in the table is kind of magical
20:32:16 <MrsB> if you find one then add a link on the update bug :)
20:32:25 <lewyssmith> A ruse new to me.
20:32:41 <MrsB> you know which link i mean lewis?
20:32:44 <Luigi12> I looked at cxxtools/tntnet at work yesterday but didn't understand the test procedure.  I got the project compliled, but I don't see what's supposed to listen on port 8000 or whatever it was
20:33:03 <Kernewes> MrsB: you mean the procedures aren't on a wiki page yet
20:33:28 <lewyssmith> MrsB: Proc available or Quick search?
20:33:32 <MrsB> not many Kernewes, feel free to add them there though :)
20:33:52 <Kernewes> MrsB: yes, I will when I run out of updates I can do
20:33:55 <MrsB> lewis, next to that, the end column, where it says Bugzilla / Wiki
20:34:06 <MrsB> click the word Bugzilla
20:34:13 <MrsB> it searches for previous updates
20:34:23 <lewyssmith> You live & learn.
20:34:39 <Stormi> that page is a swiss army knife
20:34:54 * Stormi humbly congratulates self
20:34:57 <MrsB> lol
20:35:05 * MrsB pats Stormi on th eback
20:35:22 <Stormi> thanks, my "eback" needed that
20:35:22 <Kernewes> I've had a busy week or so with real life, hope to do some testing again now
20:35:28 <MrsB> typo night again
20:35:46 <lewyssmith> In the cicumstances, eback seems legit.
20:35:51 <MrsB> great, if we all pull our weight we'll bring the list down in no time
20:36:04 * Kernewes slides a gin and tonic down to Stormi
20:36:04 <wilcal> Sounds like something printed on a t-shirt
20:36:14 <MrsB> many are quite easy to test, even puppet isn't too bad
20:37:07 <MrsB> testing things you've never heard of a a fantastic way to learn about stuff you would never otherwise realise existed
20:37:24 <MrsB> is a*
20:37:47 <MrsB> Are there any updates there we don't like the look of?
20:37:47 <wilcal> At least one new thing a week
20:38:17 <Stormi> pacemaker
20:38:36 <Stormi> this one has scared me for days
20:38:43 <Stormi> I avoid it and hope MrsB takes it :)
20:38:47 <MrsB> yes, that's about the only one now which is difficult. And a bit of a search i did earlier showed it might not even be that hard
20:38:50 <Kernewes> just hope we don't get power cuts
20:38:57 <wilcal> The /etc/pkcs11/pkcs11.conf could be a can of worms
20:39:14 <MrsB> Stormi: i'll have a crack at it in th emorning
20:39:20 <MrsB> the morning too
20:39:55 <lewyssmith> How to test 9878 generic MariaDB update?
20:40:00 <MrsB> pkcs11 is just basically ensuring the apps which didn't work before now do work and making sure you can still log in and out ok and things like kwallet work still
20:40:45 <wilcal> why does it effect only nvidia kwrite
20:40:58 <MrsB> mariadb is used by kde (akonadi) and also easily tested with webapps like phpmyadmin and anything which uses a database
20:41:18 <Luigi12> mariadb's proably the most important one that's not grayed out right now
20:41:36 <Luigi12> as far as security updates I mean
20:41:51 <Luigi12> obviously the pkcs11 thing is more important than all of them
20:42:02 <MrsB> yep, that's top priority at the moment
20:42:09 <MrsB> it's affecting alot of people
20:42:09 <tmb> wilcal, it triggers bad signal handling in nvidia driver, something that nvidia is working on fixing / improving
20:42:11 <wilcal> My concern on that is that it effects lots of things we don't know about
20:42:52 <wilcal> And I reported the Kwrite thing first over a month ago. Toss me out   :-))
20:43:11 <lewyssmith> That is why I said it was your baby.
20:43:19 <MrsB> you report it, you get to test it, that's just how it works ;)
20:43:30 <tmb> wilcal, nope, the fix is "simple" so the fallout is "0" (famous last words)
20:43:40 <Stormi> MrsB: tmb: what kind of issue could we see
20:43:46 <Stormi> *if* the package was broken
20:44:02 <Stormi> The whole system seems to depend on it so that's quite critical
20:44:04 <MrsB> You can always check what uses the libs using urpmq --whatrequires and then check those things still work ok with the update applied
20:44:11 <wilcal> I don't think there's anything wrong with Kwrite
20:44:30 <lewyssmith> It is not accused.
20:44:31 <Stormi> MrsB: urpmq --whatrequires-recursive p11-kit gives an MEGA list
20:44:45 <MrsB> yeah best to stick to --whatrequires
20:45:17 <Stormi> MrsB: which gives other libs
20:45:22 <tmb> Stormi, yes, but you need a buggy nvidia driver to trigger it, for the rest of the gpus/drivers they seem to cope nicely with it,
20:45:46 <Stormi> tmb: I have the nvidia problem and it is fixed by the update, but I'm looking for a way to test for regression
20:46:00 <Stormi> since it's so deep in the dependencies it could affect a lot of things if broken
20:46:27 <lewyssmith> Surely the fix is playing safe?
20:46:38 <MrsB> #info after checking the apps which didn't previously start, generally check nothing new seems broken on your system after installing pkcs11 updates. It has potential to affetc alot of things.
20:47:08 <Stormi> lewyssmith: I think so and it was reviewed by several packagers
20:47:41 <tmb> well, to put the fix simple: before the fix it tried to mmap an empty file, fail at it and then try to unmap an non-existing reference
20:47:47 <MrsB> Can everybody with mga4 please give it a go and particularly if you have nvidia
20:48:03 <MrsB> you said simple :P
20:48:08 <wilcal> Will do
20:48:09 <lewyssmith> Have done without nvidia though.
20:48:30 <RemyServices> will do
20:48:33 <MrsB> if you don't notice any new breakage it's probably fine lewis
20:48:47 <tmb> after the fix: try to mmap file, if empty / non-exstant, dont mmap, and dont try to unmap
20:48:57 <MrsB> we need to confirm the bug is fixed though as far as possible for as many as possible
20:49:49 <lewyssmith> It just did what it said it would: replaced an empty config file with a valid one.
20:50:06 <MrsB> good, and nothing new seems broken?
20:50:13 <lewyssmith> Not so far!
20:50:14 <Stormi> lewyssmith: not only, there's patch of code, that tmb just described
20:50:23 <Stormi> so the binaries are different
20:50:38 <Stormi> (which rpmdiff shows)
20:50:39 <lewyssmith> OK, I see the visible side.
20:50:41 <wrw105> I just got it installed here (after a boneheaded move) seems to be behaving kdewallet, kwrite.
20:50:58 <lewyssmith> See Bill?
20:50:59 <MrsB> So we need to put all this on the bug please
20:51:28 <wrw105> Once I finish here, I'll do a logout and login, test a bit more and add it to the bug.
20:51:39 <MrsB> are you nvidia wrw105?
20:51:54 <wrw105> nvidia-304
20:52:09 <MrsB> does it fix kwrite, kdenlive etc for you too?
20:52:17 <wrw105> That it does.
20:52:19 <MrsB> \o/
20:52:40 <MrsB> Are there any other updates which look scary?
20:52:56 <MrsB> (they're not scary btw even if they look it)
20:53:01 <wrw105> I had to go back and play with kdewrite--I've started using kate for most of my text editing in kde.
20:53:13 <MrsB> kate was affected too i think
20:53:36 <RemyServices> fyi, I need to step out for a few, be back
20:53:44 <MrsB> see you in a bit RemyServices
20:53:45 <wilcal> see ya soon
20:53:47 <wrw105> Shows how much day-to-day text editing I have to do then!
20:54:16 <MrsB> So any other's or shall we see what Luigi12 has in store for us :)
20:54:36 <MrsB> -'
20:54:36 <wilcal> Luigi is deeply involved in all this
20:56:02 <Stormi> (digikam was affected too)
20:56:06 <MrsB> Well, try not to be scared by any of them. They're only packages. Nobody knows about all of them and they're new to everybody at least once.
20:57:03 <MrsB> remeber about the Bugzilla link on the madb page, that's the best thing to find testing procedures
20:57:25 <MrsB> and remember too to ask if there is anything you're strugglign with :)
20:57:29 <lewyssmith> We might think of formalising this aspect.
20:57:33 <MrsB> excluding spelling
20:57:47 <Stormi> maybe I should remove the wiki link, I don't think we're going to use the wiki for procedures actually
20:57:58 <MrsB> we can do it again yes, has it been helpful?
20:58:16 <MrsB> Stormi: kernewes may add some for us :)
20:59:09 <MrsB> when we get on top of things again we can review our documentation and see where it falls short
20:59:38 <MrsB> IS there anything else before we move on?
20:59:58 <wilcal> Lets move on
21:00:08 <lewyssmith> Seconded.
21:00:11 <MrsB> you do it then wilcal :P
21:00:26 <wrw105> I've got to run...strange not being square for a change!
21:00:37 <MrsB> Thanks for being here wrw105!
21:01:03 <wrw105> The day job was open what with all the snow.. :-)
21:01:04 <lewyssmith> How many people today know what 'square' means?
21:01:12 <MrsB> old people
21:01:21 <RemyServices> lol
21:01:28 <lewyssmith> Count me out.
21:01:32 <MrsB> ppffff
21:01:41 <RemyServices> does that make me old? It came too early
21:01:57 <MrsB> wilcal, you know how to do it?
21:02:17 <Kernewes> MrsB: give him a chance, he's not going to be as fast as Dave yet :)
21:02:32 <lewyssmith> He did once OK.
21:02:42 <wilcal> I'm think'n
21:02:45 <MrsB> :D
21:02:53 <MrsB> use #topic
21:03:04 <wilcal> Sorry confused here
21:03:20 <MrsB> i'll do it then, you can do the next one
21:03:24 <wilcal> new topic thanks
21:03:31 <MrsB> #topic Luigi's Roundup
21:03:40 <wilcal> Oh ya now I know
21:03:51 <MrsB> Luigi12_work_: ping!
21:03:57 <Luigi12> he's not there
21:04:00 <MrsB> ohh
21:04:08 <MrsB> how come?
21:04:11 <lewyssmith> He has already spoken.
21:04:13 <Luigi12> snow and ice
21:04:29 <wilcal> East Coast is Snowed out
21:04:31 <Luigi12> I marked him as away when I left yesterday, guess it doesn't hold sometimes if you get disconnected :o(
21:04:33 <MrsB> free holiday \o/
21:04:42 <Luigi12> free as in I don't get paid :o)
21:04:43 <wilcal> Time to crash a lot of cars
21:05:15 <MrsB> Over to you then David
21:05:17 <Luigi12> ok
21:05:39 <Luigi12> so QA as usual has been doing a great job of keeping the really important updates cleared
21:05:53 <Luigi12> not marked as such yet, but probably the most important one now is mariadb
21:06:15 <MrsB> #info mariadb is next after pkcs11
21:06:20 <Luigi12> it's a shame that tomcat6 is such a pain, but without help from the maintainer or someone who understand the packaging for it, can't do much with it
21:06:28 <wilcal> You hear of a "CRITICAL" Flash compromise identified in the last days? It effects all platforms.
21:06:40 <Luigi12> wilcal: the sun also came up yesterday
21:06:44 <MrsB> I'll have a look at that one Luigi12
21:06:46 <wilcal> LOL
21:07:30 <Luigi12> I don't have any other current updates to highlight.  As far as what's coming soon, we should have an ImageMagick update soon as some confusion with the CVE assignments gets sorted out.
21:07:32 <MrsB> #info help needed with tomcat6
21:07:43 <Luigi12> other than that, there are handful waiting on action from their maintainers
21:07:53 <Luigi12> including virtualbox on mga3 (hint tmb :o)
21:08:10 <MrsB> #info imagemagick coming soon and maybe virtualbox
21:08:27 <lewyssmith> Maintainers are all busy testing updates, perhaps.
21:08:29 <tmb> I guess I could try to figure out the tomcat6 one as I have played some with it
21:08:46 <MrsB> we're in alot better place than we were at the beginning of the week but still a long way to go
21:08:57 <Luigi12> thanks, that'd be great.  It's heavily exploited in the wild according to one of my coworkers who knows about such things
21:09:02 <MrsB> that would be good tmb please, it's long overdue now
21:09:59 <wilcal> Sorry to burst in with this, the was published 12-hours ago:
21:10:00 <wilcal> Adobe Systems Inc. is urging users of its Flash Player software to upgrade to a newer
21:10:02 <wilcal> version released today. The company warns that an exploit targeting a previously unknown and
21:10:03 <wilcal> critical Flash security vulnerability exists in the wild, and that this flaw allows attackers
21:10:05 <wilcal> to take complete control over affected systems.
21:10:28 <Luigi12> for mediawiki, there are some issues that might not be apparent...adding the requested README.urpmi would just be putting lipstick on a pig and I don't want to do that.  It should be simpler and automatable, but some broken, undocumented cruft added by the previous maintainer is preventing that.
21:10:46 <MrsB> No update for linux yet wilcal
21:10:54 <MrsB> http://www.adobe.com/software/flash/about/
21:10:55 <[mbot> [ Adobe - Flash Player ]
21:10:56 <Luigi12> so we can either push them (as they work fine) or I'll need help either fixing or documenting the cruft in the package, or I'll just rip it all out
21:11:10 <wilcal> Updated Flash for Linux is 11.2.202.336
21:11:21 <Luigi12> wilcal: ping Anssi about it in #mageia-dev
21:11:52 <MrsB> We can't knowingly push an update which will break people's installations though, so I think it's worth doing right
21:12:08 <MrsB> #info help needed for mediawiki packaging
21:12:11 <Luigi12> it doesn't break anything, you just have to run one command
21:12:16 <Luigi12> which is documented in the advisory
21:12:18 <MrsB> yes, but only when you know about it
21:12:32 <MrsB> advisories aren't shown when updates are installed
21:12:49 <MrsB> that's why the request for a readme.update.urpmi
21:12:50 <Luigi12> I'll mail on -dev about it then
21:12:53 <MrsB> k
21:13:11 <Luigi12> any questions for me on anything else?
21:13:17 <lewyssmith> but only when you know about it: so often is it thus.
21:13:28 <MrsB> None here, thanks Luigi12
21:13:33 <Luigi12> well it's documented upstream too
21:13:45 <Luigi12> but it's stupid too, Moodle does it automatically for you, why can't MediaWiki?
21:13:52 <MrsB> yeah
21:14:06 <Kernewes> lewyssmith: tell me about it, the problems I thought I was having with Mga3 turned out months later to be due to a bug in the Errata which I didn't know would affect me
21:14:13 <MrsB> simple answer is really the readme.update.urpmi if you don't get any help with it though
21:14:25 <Luigi12> like I said, there's some reasons that doesn't make sense
21:14:37 <MrsB> share them
21:14:48 <Luigi12> so if I get no help, I'll rip Oliver's cruft out of it, and get it to run the command for you automatically
21:14:58 <Luigi12> has to do with the "instances" feature that he added
21:14:59 <MrsB> lol ok that would be better :)
21:15:24 <MrsB> i thought kharec had done some work on that
21:15:37 <Luigi12> he updated the package once, but didn't mess with the "instances" stuff
21:15:56 <MrsB> i thought he'd removed it.
21:15:58 <Luigi12> the idea apparently is to allow the package to serve more than one wiki, but it's broken and undocumented
21:16:03 <Luigi12> MrsB: I wish :o(
21:16:04 <MrsB> yeah
21:16:18 <Luigi12> yeah I'd like to get rid of it honestly
21:16:28 <Luigi12> only worry is if anyone actually knows how it works and is actively using it
21:16:39 <MrsB> yep, it's a bad time to do it really
21:16:53 <Luigi12> yeah, it sucks that we couldn't get 1.22 packaged before mga4 came out
21:17:01 <Luigi12> my fault, I had forgotten about it
21:17:21 <MrsB> mail -dev then and see what help you get, that's going to decide what can and can't be done
21:17:29 <Luigi12> ok
21:17:38 <Luigi12> guess that's it then :o)  next topic
21:17:43 <MrsB> thanks Luigi12
21:17:49 <MrsB> wilcal, over to you!
21:18:31 <wilcal> #topic Anything Else?
21:18:39 <MrsB> well done, thanks :)
21:18:46 <MrsB> Is there anything else?
21:18:51 <wilcal> One from me
21:18:52 <tmb> yes
21:19:02 <MrsB> lots then, wilcal was first..
21:19:08 <wilcal> The videos I shot at FOSDEM are up on YouTube
21:19:09 <wilcal> Mageia at FOSDEM 2014: http://youtu.be/AeQH_eMGjm4
21:19:10 <[mbot> [ Mageia at FOSDEM 2014 - YouTube ]
21:19:11 <wilcal> FOSDEM Beerfest: http://youtu.be/L6MoZ95wfR0
21:19:12 <[mbot> [ Beerfest at FOSDEM 2014 - YouTube ]
21:19:12 <wilcal> Enjoy
21:19:14 <MrsB> oh cool
21:19:16 <MrsB> thanks!
21:19:20 <Kernewes> great
21:19:29 <lewyssmith> Bill, could you do a mailgroup msg summarising your time at Fosdem?
21:19:38 <MrsB> #info the videos wilcal shot at FOSDEM are up on youtube
21:19:50 <wilcal> Not quite sure how to do that lew
21:19:53 <MrsB> http://youtu.be/AeQH_eMGjm4 Mageia at FOSDEM
21:19:54 <[mbot> [ Mageia at FOSDEM 2014 - YouTube ]
21:20:04 <MrsB> http://youtu.be/L6MoZ95wfR0 FOSDEM Beearfest
21:20:05 <[mbot> [ Beerfest at FOSDEM 2014 - YouTube ]
21:20:07 <MrsB> oops
21:20:14 <lewyssmith> Just send a msg to the group, like any other.
21:20:18 <wilcal> OK
21:20:19 <MrsB> #info beerfest, no bears
21:20:24 <lewyssmith> TIA
21:20:47 <Luigi12> wilcal: Anssi checked and doesn't see a Flash update available
21:20:48 <wilcal> FWIW the Ustream live video went way better then I thought it would
21:21:14 <wilcal> I think the that the fix is the last update we did in flash
21:21:17 <MrsB> didn't know you were doing that
21:21:27 <wilcal> which was just a few days ago
21:22:08 <wilcal> We may have it covered
21:22:24 <MrsB> OK, so anything else else?
21:22:28 <MrsB> tmb..
21:22:31 <wilcal> I'm done
21:22:44 <tmb> yes, a paiful thing to report :/
21:22:51 <MrsB> :\
21:22:55 <tmb> painful...
21:23:33 <Stormi> new ISOs for the AHCI issue?
21:23:41 <Luigi12> ahh yes
21:23:49 <tmb> We will have to do a re-release of Mageia 4 isos due to syslinux bug: https://bugs.mageia.org/show_bug.cgi?id=12549
21:23:51 <tmb> yep
21:23:52 <[mbot> Bug 12549: critical, Normal, thierry.vignaud, NEW , Can not install Mageia from DVD, but good from USB flash
21:23:54 <MrsB> i noticed you building drakx and syslinux stuff
21:23:55 * Stormi wins
21:24:06 <MrsB> O. M. G.
21:24:09 * Luigi12 is just glad tmb is OK
21:24:14 <RemyServices> I have an issue for after the meeting, unsure if it is a bug or my setup, but I have a question for those smarter then me :)
21:24:29 <wilcal> I was expecting re-issue of the M4 ISO's
21:24:48 <MrsB> Well perhaps we can include the pkcs11 fix too then
21:25:05 <Stormi> the more fixes we include, the more risks of regression
21:25:11 <MrsB> #info Bombshell: <tmb> We will have to do a re-release of Mageia 4 isos due to syslinux bug: https://bugs.mageia.org/show_bug.cgi?id=12549
21:25:12 <RemyServices> and mediawiki possibly?
21:25:15 <[mbot> Bug 12549: critical, Normal, thierry.vignaud, NEW , Can not install Mageia from DVD, but good from USB flash
21:25:48 <MrsB> at least testing now should be quick and relatively easy
21:25:53 <tmb> So I talked to Anne briefly, and we'll wait out atleat the pkcs11
21:25:58 <MrsB> We can't do this until we're on top of updates though tmb
21:26:03 <wilcal> I've been using the Live-DVD's
21:26:24 <Stormi> (although I would not be against having apper disabled, so that people installing don't get that buggy notification of update until they actually manage to update)
21:26:31 <tmb> MrsB, yoah, we thought of starting to build them next week
21:26:48 <MrsB> let's see how we get on with reducing the list
21:27:22 <Stormi> list, we will kill you. Nothing personal, it's you or us.
21:27:32 <tmb> and we were discussing about maybe pulling in all validated updates too
21:27:32 <Luigi12> indeed
21:27:38 <Luigi12> that'd be nice
21:27:44 <MrsB> yes that's worth doing
21:27:54 <MrsB> a job worth doing, is worth doing well :)
21:28:09 <tmb> that would get all nice translations, less drakx crashes and so on
21:28:15 <Kernewes> a bit of practice for when we do later ISOs as we were discussing before
21:28:38 <MrsB> yes. While it's still fresh in everybodies mind how to do it
21:28:50 <tmb> and for you MrsB the bug 6077 is squashed :)
21:28:53 <[mbot> Bug https://bugs.mageia.org/show_bug.cgi?id=6077 critical, High, tmb, RESOLVED FIXED, Some Centrino laptops forced to use vesa driver & No ACPI, kernel
21:29:14 <MrsB> yes, thanks for that. My laptop is on it's last legs now though :D
21:29:41 <tmb> noooo.... ou have to keep it going now when the bug is fixed :)
21:29:51 <MrsB> i know. I'd feel a bit guilty
21:30:13 <MrsB> the power connection is failing
21:30:25 <MrsB> i can sympathise with it
21:30:43 <tmb> but seriosly, that fix made some othert systems boot properly too, so all good :)
21:31:06 <MrsB> any ideas what is causing tonyb's problems?
21:31:29 <Stormi> he got a cursed intel driver, I'd say
21:31:48 <MrsB> strange it's failing on several different systems
21:31:56 <lewyssmith> tmb: I can kill 2 bugs because the machine showing them is henceforth out of service.
21:32:07 <tmb> nope, not yet.... the problem is that the intel x driver has still not gotten to final release stage
21:32:44 <MrsB> so waiting on upstream
21:32:46 <tmb> so he might just be an unlucky one for now
21:32:49 <wilcal> FWIW I've fixed a LOT of old laptops with Puppy Linux and the users are very happy
21:33:06 <Luigi12> I thought Puppy was a rescue distro
21:33:19 <MrsB> Are there any workaround he could try which haven't been mentioned?
21:33:24 <wilcal> It is, it rescues old laptops in the closet
21:33:29 <Luigi12> oh lol
21:33:56 <tmb> MrsB, I'll re-read the messages and see if I can suggest something
21:34:10 <MrsB> thanks tmb, he's a qa team member
21:34:31 <MrsB> so...
21:34:36 <Stormi> o/
21:34:38 <MrsB> Is there anything else else else?
21:34:43 <MrsB> stormi..
21:34:44 <tmb> as for iso testing I was thinking of putting them on public mirrors too, so the affected ones can help test
21:34:45 <wilcal> Not from me
21:34:47 <Stormi> I'll keep it quick
21:34:57 <diogenese> :)
21:35:00 <MrsB> sure, why not tmb, we can rsync from there just as easily
21:35:34 <tmb> since I doo hope we need only one build :)
21:35:58 * MrsB not going to tempt fate
21:36:08 <wilcal> we're outta here
21:36:13 <MrsB> no wait
21:36:18 <MrsB> Stormi: your go
21:36:24 <Stormi> our policy currently forbids packagers to test their own updates. It makes sense in theory because you can't judge your own work. However, one of our problems is to understand how the package works and packagers often know it better.
21:36:49 <Luigi12> yeah, relaxing that rule might actually help encourage packagers to post testing procedures
21:36:52 <Stormi> So when the packager tests one arch + tells the procedure used for it, I would favor counting it as valid
21:37:19 <Stormi> "I tested it quickly" doesn't count. "I tested it on mga4 32 bits, here is how..." counts
21:37:20 <wilcal> Procedure MUST be simple, and repeatable, to follow
21:37:25 <lewyssmith> If one cannot judge one's own work, one can certainly *test* it.
21:37:29 <MrsB> I can see positives and negative to his idea
21:37:39 <MrsB> this*
21:37:48 <Stormi> of course the packager couldn't test both archs
21:37:58 <MrsB> positives, it would actually happen
21:38:36 <Stormi> I don't think we're talking about a lot of occurences actually.
21:38:52 <lewyssmith> I think it is sensible & natural to do.
21:38:59 <Luigi12> so I guess someone else would still have to add the OK tag once it's determined to be valid
21:39:58 <MrsB> negative, we've heard numerous times 'it works for me' and then found issues, which could then create conflict.
21:40:21 <Stormi> MrsB: I would agree with the negative part if we did comprehensive tests
21:40:23 <Luigi12> that's true, and QA is great at sussing out other issues people don't always find themselves
21:40:27 <Luigi12> checks and balances are good
21:40:52 <lewyssmith> Even if it got a 2nd test, it would have a better chance of succeeding.
21:41:01 <MrsB> IMHO i think it's best to keep the roles separate but i can see the positives in it too
21:41:09 <Stormi> It's true that having to discover the hard way how to use packages helps to find issues
21:41:15 <Stormi> but then we shouldn't ask for procedures :)
21:41:40 <Stormi> when we have procedures we follow that path and maybe find less issues
21:41:40 <Kernewes> perhaps if the packager does a test first, it might help prevent some of the issues QA find with updates
21:41:42 <Luigi12> maybe encourage packagers to test their stuff and post procedures, but say the QA team still has final discretion about when it's OK to validate things
21:41:49 <Kernewes> we can still test it as well
21:41:58 <Luigi12> Kernewes: in some cases yes, you're absolutely right
21:42:04 <MrsB> IIRC we talked about this a long long time ago and I was out voted, we can have a vote again now if you liike?
21:42:08 <tmb> I guess the best thing to do here is to try and see how it works out, and discuss the outcome in a few weeks / months / something...
21:42:11 <lewyssmith> Kernewes: What I tried to say above.
21:42:25 <Stormi> if it doesn't speed up the process, why would packagers test more tomorrow than today? :)
21:42:26 <Luigi12> MrsB: maybe best to hold votes on the mailing list so all interested parties can participate
21:42:50 <MrsB> well votes would be from qa team
21:42:51 <Kernewes> Stormi: it would save QA some wasted time
21:43:00 <lewyssmith> Seconded.
21:43:01 <Luigi12> yeah, they're just not all here now :o)
21:43:01 <Stormi> we already invite them at testing and providing procedures
21:43:11 <Stormi> sometimes we get some, sometimes we don't get any
21:43:24 <Luigi12> Stormi: sort of, but with the restriction in place which discourages it
21:43:28 <MrsB> Ok So lets have a vote on the ML. Can we get a quick idea here now though please?
21:43:33 <Stormi> Luigi12: indeed
21:43:43 <MrsB> Who would like to see this trialed?
21:43:44 <Luigi12> dunno if it'd help, but if they know their testing might help speed up the process, it might help encourage it
21:44:00 <Kernewes> can't do any harm to try
21:44:01 <Luigi12> I'd like to see it trialed.  I'm OK with either resolution, it's just an interesting idea.
21:44:08 <lewyssmith> It should speed things up.
21:44:16 <Luigi12> I don't want broken updates anymore than anyone else :o)
21:44:36 <Stormi> I'm for raising the restriction
21:44:47 <MrsB> Anybody think it's not a good idea?
21:44:48 <Stormi> for one arch only
21:45:00 <Stormi> and provided they tell what they have tested
21:45:07 <wilcal> Gotta have those proceedures in there
21:45:14 <wilcal> simple ones
21:45:15 <MrsB> yes, it would need boundaries like that Stormi
21:45:19 <Kernewes> we might decide it's a bad idea but at least we'll have tried it
21:45:26 <Luigi12> Stormi: well you might as well let them both arches if they want to, but someone else should have to also test for it to be validated anyway
21:45:30 <tmb> Stormi, well, better if they test both arches, but they can only validate one :)
21:45:45 <Stormi> tmb: Luigi12: yes, that's idea, better phrased
21:45:47 <MrsB> well we do have procedures for most things. There isn't much we haven't updated at least once. Pacemaker is one of those though :\
21:45:51 <Luigi12> yeah, QA still holds discretion on when it's enough testing for validation
21:46:25 <Stormi> I would say the following:
21:46:37 <Stormi> packagers are encouraged to test and give testing steps
21:46:46 <Stormi> QA decides then if it's sufficient or not
21:47:01 <Luigi12> and QA adds the OK and validation tags
21:47:04 <Stormi> yes
21:47:23 <Stormi> sometimes it will be ok, sometimes we will still want to test
21:47:39 <MrsB> If they provide steps to show what they've done to test it would give some assurance that they had actually tested
21:47:46 <lewyssmith> I think QA should always test anyway.
21:47:49 <MrsB> that's would have to be a stipulation
21:48:09 <Kernewes> QA should test at least one arch
21:48:32 <Stormi> sometimes packager will fix bugs and test only that the bug is fixed. Then it's not sufficient because it lacks the basic regression testing
21:48:51 <Stormi> a QA team member judges from what's written by the packager
21:49:04 <Kernewes> although when QA is overstretched, we can't do any more than that ourselves
21:49:16 <Stormi> and of course "I tested in cauldron" doesn't work :)
21:49:41 <lewyssmith> Don't we all basically check (if we can) that a bug is fixed?
21:49:52 <Stormi> that + basic regression testing
21:50:07 <Stormi> you can break the whole package and still fix the bug
21:50:08 <MrsB> Not only that. take for example a couple of instances i can think of off the top of my head. Packager tested the update and it worked fine for them but on a system where the package was installed new it failed
21:50:33 <MrsB> Ok. So lets' trial this with restrictions.
21:50:36 <Kernewes> perhaps they could test it in a VM
21:50:54 <MrsB> Packager can test one arch and must give the procedure used to test with
21:51:03 <MrsB> tested ok, isn't enough
21:51:03 <Stormi> MrsB: that happens with QA team members too, one arch is marked OK, then the second tester finds an issue
21:51:03 <lewyssmith> Hence the need to do it twice; but quicker the 2nd time.
21:51:27 <Stormi> A last argument in favor of my proposal: when packagers find bugs while testing... They fix their package directly
21:51:38 <Stormi> no "feedback" sitting for days
21:51:46 <MrsB> Stormi: I'm agreeing with you, lets just set some rules
21:51:57 <lewyssmith> The idea should improve the quality of update candidates.
21:52:53 <MrsB> #info we agree to  atrial of allowing packagers to test one arch of their updates only providing they give the procedure used to perform the test
21:53:15 <wilcal> Sounds good to me
21:53:31 <Stormi> yes, that would be a good atrial (whatever that means) :P
21:53:31 <MrsB> #info QA must test at last one arch and will have final say on when the package is properly tested
21:53:44 <Kernewes> fine
21:54:05 <lewyssmith> I still think both arch's.
21:54:22 <MrsB> we may aswell not be here then
21:54:54 <lewyssmith> Not what I was saying: QA to do both arch's.
21:55:05 <lewyssmith> Sorry.
21:55:09 <MrsB> ohh, yes well we can do
21:55:49 <MrsB> #info nothing stops QA still from testing both arches, just packager is free to test one as long as they give the procedure they used to do the testing.
21:55:54 <lewyssmith> But each party doing just 1 arc. would be slick!
21:56:01 <lewyssmith> arch.
21:56:11 <Luigi12> who's gonna test ARM?
21:56:13 * Luigi12 hides
21:56:15 <MrsB> So how long shall we trial this for?
21:56:25 <MrsB> is this ok stormi?
21:56:26 <Kernewes> Luigi12: I've got a Raspberry Pi
21:56:30 <Stormi> yes
21:56:42 <wilcal> I use a Pi for my back up server
21:56:51 <Luigi12> nice
21:56:53 <RemyServices> I'll have a BeagleBone in two months to test with if needed
21:56:56 <lewyssmith> Kernewes: So you are down for ARM!
21:57:25 * tmb have one rpi and one wandboard
21:57:39 <lewyssmith> To answer the question: 3 months?
21:57:52 <MrsB> #info Let's allow a month or so to trial this, not forgetting we'll be doing ISOs again soon. If any conflicts or conflicts of interest arise we will review it sooner
21:58:32 <Kernewes> Luigi12: is ARM nearly ready?
21:59:01 <Luigi12> Kernewes: not sure, but it'll be discussed again in the next packager meeting I think
21:59:08 <MrsB> Is everybody OK with what's been said?
21:59:12 <tmb> I'm waiting on a status update from rtp
21:59:13 <Kernewes> MrsB: yes
21:59:24 <lewyssmith> Yes.
21:59:37 * MrsB outvoted again, drat!
22:00:04 <MrsB> Soo.. is there anything else else else else?
22:00:09 <Stormi> yes
22:00:11 <Kernewes> no no no no
22:00:14 <lewyssmith> Goodnight.
22:00:16 <Stormi> "let's end that meeting"
22:00:18 <Luigi12> more snow
22:00:28 <Kernewes> more rain and wind
22:00:38 <tmb> nothing else but big thanks for all your great work
22:00:39 <wilcal> I'm done
22:00:46 <MrsB> Well thankyou all for coming and not being square (that's boring for the young'uns)
22:00:47 <Luigi12> yep thanks
22:00:53 <MrsB> t-5
22:00:55 <Kernewes> MrsB: thank you
22:00:55 <MrsB> 4
22:00:57 <MrsB> 3
22:00:58 <MrsB> 2
22:01:00 <MrsB> 1
22:01:02 <MrsB> #endmeeting