20:00:59 <DavidWHodgins> #startmeeting
20:00:59 <Inigo_Montoya> Meeting started Thu Nov  2 20:00:59 2017 UTC.  The chair is DavidWHodgins. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:00:59 <Inigo_Montoya> Useful Commands: #action #agreed #help #info #idea #link #topic.
20:01:09 <DavidWHodgins> #chair wilcal tarazed_ lewyssmith
20:01:09 <Inigo_Montoya> Current chairs: DavidWHodgins lewyssmith tarazed_ wilcal
20:01:39 <DavidWHodgins> Sorry everyone for messing up things with the time change again. I'll add a page to the wiki later, for future reference
20:01:46 <DavidWHodgins> #topic * Who's new? - If you are then come and say hello.
20:02:05 <DavidWHodgins> Anyone here who hasn't participated in a qa team irc meeting before?
20:02:38 <Benmc> no new hands raised
20:02:48 <DavidWHodgins> #topic * Mageia 5 EOL extended to Dec. 31st.
20:03:06 <wilcal> Big discussion at the Council meeting
20:03:11 <kekePower> That's nice. I think it needs to be kept for a litle longer
20:03:35 <wilcal> I'll let David give his take on what the Council decided
20:03:41 <DavidWHodgins> Yes. It's been extended to give packagers more time to fix the problems with upgrading from 5 to 6 using mgaapplet
20:04:01 <lewyssmith> And we should fix the 'nobody' login thing beforehand.
20:04:26 <DavidWHodgins> It's only a 2 month extension, so it'll still be a tight timeline for people to upgrade once mgaapplet upgrading is enabled again
20:04:39 <wilcal> Critical security fixes will be worked
20:04:53 <wilcal> 31 Dec 2017 is a brick wall stop
20:05:24 <DavidWHodgins> Yes, during the extended support period, only critical updates and those that fix problems with upgrading will be allowed
20:05:52 <wilcal> Vbox will not be updated to Ver 5.2
20:05:58 <kekePower> will there be _no_ updates after this date?
20:06:06 <DavidWHodgins> Docteam is working on info to help people do clean installs, keeping /home, etc.
20:06:07 <kekePower> whatsoever?
20:06:12 <DavidWHodgins> kekePower: Correct
20:06:15 <kekePower> ok
20:06:29 <wilcal> If an update breaks Vbox the fix is to upgrade to M6
20:06:44 <DavidWHodgins> Any questions about Mageia 5 updates?
20:07:01 <DavidWHodgins> #topic * Mageia 6.1 ISO images will be coming soon
20:07:17 <DavidWHodgins> tmb: Care to comment on the new hardware that will be supported?
20:07:46 <wilcal> I understand this is mostly or exclusively to handle new hardware
20:07:58 <kekePower> kernel 4.4.92 fixes CVE-2017-12193
20:08:18 <DavidWHodgins> Yes. Though all updates that have been released prior to the iso images being built, will be included in the new iso images
20:08:53 <wilcal> Is this mostly for the new Intel processors?
20:09:29 <DavidWHodgins> That's my guess. It was during the council meeting that I first learned that a 6.1 iso was being considered
20:09:44 <wilcal> Me too
20:09:46 <kekePower> https://bugzilla.redhat.com/show_bug.cgi?id=1501215
20:09:48 <[mbot> [ Bug 1501215 – CVE-2017-12193 kernel: Null pointer dereference due to incorrect node-splitting in assoc_array implementation ]
20:10:10 <DavidWHodgins> Yes, some security fixes as well as the new hardware support.
20:10:22 <wilcal> So most of us cannot test for the new processor operation
20:10:26 <wilcal> If any of us
20:10:27 <tmb> yeah, Intel Coffee Lake platform + Proper AMD Polaris support + various new wireless and other hw
20:10:29 <DavidWHodgins> Correct
20:10:52 <tjandrews> I can't afford new hardware. I'll just check to make sure it still works on the older stuff I already have.
20:10:58 <tmb> + better fixes on Amd Ryzen platform
20:11:03 <wilcal> That's all we can do for now
20:11:04 <DavidWHodgins> I'm really out of date on what hardware is becoming available
20:11:06 <lewyssmith> That applies to almost everyone.
20:11:40 <lewyssmith> That was meant for TJ!
20:11:53 <DavidWHodgins> So the iso testing should require a lot less time than a normal release. It will only be testing for regressions compared to the Mageia 6 iso images
20:12:36 <DavidWHodgins> We still need to test all of the iso images on real hardware (efi and bios firmware), and under virtualbox.
20:12:38 <tjandrews> But, of course this means that we will have a greater need than ever for some people on QA that CAN afford new hardware.
20:12:59 <lewyssmith> And want it...
20:13:08 <kekePower> oh, I want it
20:13:23 <DavidWHodgins> As always, with packages or updates for hardware no one on the qa team has, all we can test is that everything works ok on older hardware
20:13:23 <kekePower> it's just the price of $4000 that's stopping me
20:13:43 <wilcal> Newest Intel platform here is an I5-4460
20:13:45 <DavidWHodgins> That, and figuring out what to get. :-)
20:14:57 <DavidWHodgins> I'll send a msg to everyone on the Mageia 6 iso testing list, informing them that 6.1 is coming, asking them to also help with the 6.1 testing.
20:15:22 <tjandrews> Will we be using the pretest system on this one, or hope that it isn't needed?
20:15:25 <neoclust> there is a release date for mageia 6.1 ?
20:15:37 <lewyssmith> tjandrews, TJ:I think not.
20:16:26 <DavidWHodgins> No pretests. Just a reminder that anyone who is likely to become upset if time spent downloading an iso turns out to be for nothing, that they should wait for prelimiary reports from others before investing the time downloading
20:16:30 <tmb> neoclust, we try to target mid december 2017 for Xmas (or when ready)
20:16:42 <neoclust> tmb: ok :)
20:16:54 <neoclust> tmb: i ask to finish my qt 5.6.3 update + plasma 5.8.8 :)
20:16:56 <tmb> We should be able to build test isos in ~2 weeks
20:17:33 <neoclust> ok i see with david to finish this
20:17:35 <DavidWHodgins> Hopefully including any fixes needed for upgrading
20:17:36 <neoclust> tmb: tks :)
20:17:39 <tmb> neoclust, yes please try to get those done asap so they can go out on isos
20:18:17 <tjandrews> Funny - it doesn't seem like we've been using Mageia 6 for very long, even though I switched my production install to it back in January.
20:18:30 <DavidWHodgins> Any questions about Mageia 6.1 iso images before we move on?
20:18:41 <wilcal> Well Intel Coffe Lake came along
20:18:48 <wilcal> not from me
20:18:51 <tmb> and any 6.1 installer related fixes are limited to rebuild for know new hw ids + possible upgrade fixes
20:19:42 <DavidWHodgins> Yes. By keeping the changes as minimal as possible, it reduces the testing needed to be primarly looking for regressions
20:19:54 <DavidWHodgins> #topic * Mageia 7 planning.
20:20:40 <lewyssmith> I rejoice that this is thought og.
20:20:43 <lewyssmith> of
20:20:44 <DavidWHodgins> The planning for Mageia 7 is basically on hold pending a packagers meeting, to get agreement within the dev team on who is willing to work on which features
20:20:53 <wilcal> https://wiki.mageia.org/en/Category:ProposedFeatureMageia7
20:22:19 <DavidWHodgins> Given the delays getting the features approved, and the impact some of them will have on the installers etc., if they are approved, I doubt we will even try to target a release date of 8 months after Mageia 6.
20:22:45 <wilcal> Sept 18 maybbe
20:22:47 <DavidWHodgins> That will have to be decided by the full council though, and is very dependant on what features are approved.
20:23:33 <ennael> (hello sorry back from vacation.)
20:23:44 <lewyssmith> Bienvenue!
20:23:49 <DavidWHodgins> The biggiest and most controversial is mana tools replacing mcc and other drak tools
20:23:53 <DavidWHodgins> HiYa ennael
20:24:05 <ennael> hi :)
20:24:08 <DavidWHodgins> Hope you enjoyed the vacation!
20:24:15 <rindolf> ennael: hey
20:25:03 <DavidWHodgins> ennael: I think we may need you to really push for a dev team meeting to get the proposed features discussed
20:25:25 <lewyssmith> The first step.
20:25:36 <DavidWHodgins> Yep
20:26:12 <ennael> ok the pb here is more about some features not all
20:26:31 <ennael> especially drak* versus mana*
20:26:45 <DavidWHodgins> Yes. manatools and dnf as defaults are the two most controversial
20:27:01 <lewyssmith> The important thing is to agree what is wanted. Then go for it.
20:27:11 <ennael> ok can you list the guys implied in it for mana*
20:27:26 <ennael> so that we organize this
20:27:39 <lewyssmith> You = ?
20:28:16 <DavidWHodgins> That's the problem. We don't know which packagers would be working on the feature, if it is approved. So far only 2.
20:28:32 <DavidWHodgins> Neal and one more, who I don't remember off hand.
20:29:08 <lewyssmith> OK, so what is agreed is constrained by what can be done.
20:29:58 <neoclust> i think people wanting to default to dnf don't see the huge work, and for not a real gain
20:30:02 <ennael> the thing is we need ot organize this meeting or at least a mail discussion with those guys only
20:30:03 <neoclust> #my2cts :)
20:30:13 <neoclust> ennael: + tvignaud
20:30:20 <ennael> not to hide anything but to make discussion easier and clearer
20:30:29 <ennael> neoclust: I know the guys on drak* :)
20:30:38 <DavidWHodgins> Some people will clearly be annoyed if we replace the drak tools, but at the same time, that is not a valid reason to avoid considering mana tools and dnf
20:30:51 <neoclust> ennael: i plan to work on drak*
20:31:38 <kekePower> drak is and has been the heart of mga forever
20:31:51 <lewyssmith> Is this not more a development matter?
20:32:17 * ennael just wants name of guys currently working on manatools
20:32:25 <ennael> not to take a decision tonight :)
20:32:26 <kekePower> urpm is, imho, too what makes Mageia the awesome distro it is
20:32:30 <ennael> just to organize meeting
20:32:39 <lewyssmith> This is QA, Anne. How do we know?
20:32:54 <ennael> I thought you knew it :)
20:33:11 <lewyssmith> Désolé.
20:33:15 <ennael> no pb :)
20:33:31 <neoclust> ennael: i see alien and anaselli
20:33:35 <neoclust> ennael: only those 2
20:34:19 <ennael> ok I will ask them
20:34:22 <DavidWHodgins> Neal Gompa is the primary person trying to get this through
20:34:35 <neoclust> DavidWHodgins: i don't see hil commited on the code
20:34:51 <DavidWHodgins> He's been the most vocal about it
20:35:10 <neoclust> Pharaoh_Atem: speaks a lot :)
20:35:11 <DavidWHodgins> I haven't looked at any of the packages actually involved in it.
20:35:36 <DavidWHodgins> Pharaoh_Atem is Neal Gompa
20:35:44 <ennael> yep ok
20:35:50 <ennael> sorry for the noise I will mail :)
20:35:56 <DavidWHodgins> Thanks
20:36:25 <DavidWHodgins> Ok. Back to regular topics. :-)
20:36:35 <DavidWHodgins> #topic * Testing updates - Any difficulties, problems, issues?
20:36:39 <lewyssmith> https://bugs.mageia.org/show_bug.cgi?id=21774
20:36:41 <[mbot> [ 21774 – Git cvsserver OS Command Injection (CVE-2017-14867) ]
20:36:51 <lewyssmith> This could be for you, Dave.
20:36:57 <neoclust> ennael: you can add my on the loop :)
20:36:59 <wilcal> I have an over all issue we need to discuss
20:37:01 <neoclust> ennael: you can add me on the loop :)
20:37:27 <lewyssmith> wilcal, Over to you, Bill.
20:37:29 <DavidWHodgins> Yes. Pretty sure I've tested it before. Been a while though
20:37:37 <wilcal> Critical issue we need to decide. I no longer have any real 32-bit hardware. All went to the recycler last month. So trying to test 32-bit Nvidia, Vbox or x11 updates is a non starter here. tmb brought this up in a recent bug. No 32-bit in M7? We've got a couple outstanding bugs that maybe can't be tested in 32-bit.
20:38:14 <DavidWHodgins> We can still test a 32 bit install on 64 bit hardware
20:38:21 <lewyssmith> I think M7 is good for dropping 32-bit. I agree, it is getting impossible to test - Herman is our hero.
20:38:28 <wilcal> Does not work here for some reason
20:38:45 <rindolf> x/j
20:38:56 <wilcal> Vbox testing is fine for me
20:39:12 <DavidWHodgins> I have 32 bit and 64 bit test installs here, both on real hardware, and under vb.
20:39:16 <wilcal> But you can't test Vbox in a Vbox client
20:39:29 <DavidWHodgins> My video hardware is all ati
20:39:39 <lewyssmith> We are agreed not to mess with VB 32-bit host.
20:39:39 <wilcal> No testing nvidia in a Vbox client
20:39:58 <DavidWHodgins> For vbox, we only test it on real hardware, and only in a 64 bit system
20:39:58 <wilcal> so we can still do some 32-bit testing but it is becoming limited
20:40:09 <lewyssmith> Very.
20:40:20 <DavidWHodgins> Yes, vb, xen, etc. we only test on 64 bit installs
20:40:32 <wilcal> So maybe a feature in M7 is no 32-bit?
20:40:44 <lewyssmith> I think so.
20:40:56 <tmb> and I'm inclined to state nvidia-current as x86_64 only too
20:40:58 <wilcal> Is there a 32-bit RedHat
20:41:10 <wilcal> Thank you Thank you tmb
20:41:28 <wilcal> so we can push that one update
20:41:31 <lewyssmith> There *are* many 32-bit distros at present. Our problem is coping with it, not the technology.
20:41:38 <DavidWHodgins> No. Given the interest in downloading the i586 iso images, I'm not in favour of dropping 32 bit support, even if real 32 bit hardware is becoming very rare
20:41:42 <tmb> if someone can test it on 32bit, then good... but I dont see it as a show-stopper
20:41:53 <kekePower> I've an issue with XFS not getting included in the initramfs when upgrading a kernel
20:42:09 <DavidWHodgins> For most things, testing 32 bit software on 64 bit hardware is fine
20:42:09 <wilcal> something to consider as we head into 2018 and M7
20:42:19 <kekePower> had to use dracut to be able to boot
20:42:50 <kekePower> was for 4.13 and kernel-tmb-4.9.59
20:43:26 <brian_> XFS - for / I assume
20:43:27 <DavidWHodgins> For efi, we only test 64 bit, for for bios firmware, we still need to test both 32 bit and 64 bit installs, even if only tested on 64 bit hardware.
20:43:32 <kekePower> xfs is enabled in the kernel, just not in intramfs
20:43:34 <kekePower> brian_: yes
20:43:39 <Pharaoh_Atem> DavidWHodgins: hey?
20:44:23 <DavidWHodgins> Pharaoh_Atem: We were discussing the Mageia 7 planning, just giving the qa team a overview of where the council discussions are at.
20:44:34 <Pharaoh_Atem> ah
20:44:37 <tjandrews> Some third party software, particularly from Google, often works better in the 32-bit versions. Google Earth is a good example.
20:44:41 <DavidWHodgins> Not trying to make any decisions here.
20:45:32 <tmb> tjandrews, that's why we still add 32bit medias on 64bit installs to make it easy to install 32bit stuff when needed
20:45:34 <brian_> tmb:  thoughts on XFS isue?
20:45:35 <Pharaoh_Atem> DavidWHodgins: ah okay, also, JFYI, manatools developers are myself, anaselli, AL13N, papoteur, with associated contributors (besser82, etc.)
20:45:48 <DavidWHodgins> tjandrews: True. Even though I use it in a 64 bit Mageia install, I use the 32 bit google earth software.
20:46:44 <tjandrews> tmb: Right. But if we drop 32-bit support, won't we be pressured to drop support for those medias, too?
20:46:46 <lewyssmith> Does this imply that 32-bit is important for some applications, rather than the OS itself?
20:47:13 <kekePower> reported it in bug 21950
20:47:14 <DavidWHodgins> lewyssmith: Yes
20:47:14 <tmb> brian_, not yet as its the first issue heard... it could have been a temporary glitch or the big xfs stuff that got backported to 4.13.x and 4.9.x stable branches that gives some issue
20:47:15 <[mbot> Bug: ['kernel 4.13.10 is missing support for XFS filesystem', 'NEW', 'Thomas Backlund'] https://bugs.mageia.org/show_bug.cgi?id=21950
20:47:20 <Pharaoh_Atem> lewyssmith: for a while, I suspect as much
20:47:38 <Pharaoh_Atem> there are plenty of applications that function purely through multiarch
20:48:49 <tmb> tjandrews, no, we can simply state no 32bit isos, and 32bit os installs only supported on best effort...
20:49:11 <DavidWHodgins> While it would reduce our workload, if we dropped 32 bit software completly, as long as it works on 64 bit hardware, I see no reason to drop it.
20:49:38 <tmb> so the bs can still build 32bit stuff for now... and we only fix 32bit issues if bugreports arise
20:49:38 <lewyssmith> It would 1/2 the workload, roughly.
20:50:04 <lewyssmith> Say 1/3.
20:50:27 <lewyssmith> i.e. - 1/3
20:50:42 <DavidWHodgins> I know. That's why I was originally in favour of it. But the Mageia 6 torrent stats I've been seeing show there is still a much stronger interest in 32 bit iso images than I expected
20:50:50 <wilcal> still 32-bit boot.iso's
20:50:57 <wilcal> net install
20:51:17 <kekePower> is it time, now, to drop 32-bit support? I've seen quite a few distros doing it because of the work and effort and how not many people are using it these days
20:51:30 <lewyssmith> i WOULD ACCEPT USER INTEREST IN 32-BIT AS A GOOD REASON FOR KEEPING IT.
20:51:44 <lewyssmith> Sorry about caps...
20:51:53 <DavidWHodgins> That may just be people who like to collect all of the iso images, or mistakenly think 32 bit software will use less space, but it may be people who really do still have 32 bit hardware
20:51:58 <wilcal> so by the time we get to M7 functions final decision we should also decide on where we are on 32-bit
20:52:00 <kekePower> lewyssmith: you shocked me there a bit :)
20:52:07 <DavidWHodgins> :-)
20:52:49 <tjandrews> I thought he was just being emphatic.
20:52:52 <wilcal> IMO operation of the Raspberry Pi may out weigh 32-bit operation
20:53:05 <lewyssmith> tjandrews, Words alone suffice!
20:53:06 <tmb> yeah, I think we'll keep 32bit for mga7 with current planned release schedule
20:53:09 <DavidWHodgins> We intentionally do not try to keep download stats, partly to avoid tracking users, and also as it wouldn't be reliable due to no control over mirrors, etc.
20:53:57 <kekePower> would a voluntary "ping" from the Welcome screen work?
20:54:22 <DavidWHodgins> At some point we should ask all users speak up if they are really using 32 bit hardware, but until that's done we have to assume it's actually being used
20:54:39 <tmb> but maybe state 32bit is "secondary arch" so it wont block updates needed for x86_64...
20:54:44 <DavidWHodgins> kekePower: Yes. Something like that, or a request for feedback in the post-install phase
20:55:05 <DavidWHodgins> tmb: That's a good idea. Same as current policy for arm.
20:55:27 <kekePower> tmb: that'll make testing a lot easier and faster
20:55:42 <tmb> Well, technically we will get an "phone home" feature if I ever get to finish up the mirrorbrain stuff :/
20:55:42 <wilcal> I'll gently bring this up in the Forum
20:55:45 <lewyssmith> tmb, We already sometimes use 64-bit tests alone as sufficient.
20:56:27 <kekePower> I see Tails posting stats about usage once a month, so there's something they do to keep "track" of boots or other stuff
20:56:34 <DavidWHodgins> In most cases, that is enough. It's only hardware packages like the kernels thate really need both.
20:56:44 <tmb> lewyssmith, yeah, but I meant if a bug at release time it would not block release
20:56:57 <DavidWHodgins> Ah. Gotcha.
20:58:23 <lewyssmith> Can we take it that the 'current updates' topic is sorted?
20:58:40 <tmb> I myself switched all my systems to 64bit in 2006, so I'm just waiting for 32bit to die :)
20:59:00 <wilcal> :-)
20:59:04 <lewyssmith> 11y!
20:59:06 <DavidWHodgins> My 32 bit system died a few years ago now.
20:59:19 <DavidWHodgins> It was roughly 15 years old then.
20:59:39 <DavidWHodgins> Anyway, back to updates. :-)
20:59:43 <kekePower> there is something nostalgic about "i386"
20:59:53 <DavidWHodgins> Anyone here have a system that uses nvidia-current?
20:59:59 <kekePower> like an old friend that we don't want to let go
21:00:08 <brian_> well -  use the 32-bit on two of my old lower memory boxes
21:00:23 <wilcal> Intel 8080? Did you know that Intel had a 4040 product
21:00:29 <brian_> with with 512mb and the other with 1 GB of ram, 32-bit "seems" to use less memory
21:00:36 <lewyssmith> You are our other real 32-it tester.
21:01:00 <brian_> :wilcal - lol - yeah
21:01:39 <DavidWHodgins> If we can't find anyone to test nvidia current, bug 21855, we'll approve it based on the update installing cleanly over the prior version
21:01:41 <[mbot> Bug: ['Update request: nvidia-current 384.90', 'NEW', 'QA Team'] https://bugs.mageia.org/show_bug.cgi?id=21855
21:01:42 <kekePower> wilcal: when you say it like that, it seems a logical conclusion. 4040 bumps to 8080
21:01:44 <tjandrews> I use nvidia 340 on one set of hardware. Everything else is Intel graphics.
21:01:49 <wilcal> https://en.wikipedia.org/wiki/Apollo_Guidance_Computer
21:01:52 <wilcal> 16-bit
21:01:57 <lewyssmith> kekePower, I tried Debian recently, and it seems to be compiled for 386...
21:01:58 <brian_> mine are older than that
21:02:13 <brian_> nvidia 340
21:02:28 <DavidWHodgins> Any questions about remaining updates? They look pretty straight forward.
21:02:29 <kekePower> lewyssmith: yeah, they've kept the 386 moniker
21:02:50 <kekePower> ever since I found Mandrake, I've seen i586
21:02:52 <brian_> LOL - I guess just text for that box
21:02:56 <DavidWHodgins> Ok. Final topic.
21:03:00 <DavidWHodgins> #topic * Anything else?
21:03:08 <wilcal> Not from me
21:03:13 <lewyssmith> Not here.
21:03:22 <brian_> bye all
21:03:32 <kekePower> thanks to wally for pushing neovim. it's an awesome addition, imho
21:03:33 <DavidWHodgins> Thanks everyone! Countdown time then
21:03:38 <DavidWHodgins> t - 5
21:03:41 <lewyssmith> Goodbye everyone, thanks for coming.
21:03:43 <DavidWHodgins> 4
21:03:46 <DavidWHodgins> 3
21:03:50 <DavidWHodgins> 2
21:03:54 <DavidWHodgins> 1
21:03:59 <wilcal> bye all
21:04:01 <DavidWHodgins> Thanks for coming everyone!
21:04:05 <DavidWHodgins> #endmeeting