20:04:27 <wilcal> #startmeeting 20:04:27 <Inigo_Montoya> Meeting started Thu Dec 8 20:04:27 2016 UTC. The chair is wilcal. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:04:27 <Inigo_Montoya> Useful Commands: #action #agreed #help #info #idea #link #topic. 20:04:47 <wilcal> #chair lewyssmith 20:04:47 <Inigo_Montoya> Current chairs: lewyssmith wilcal 20:05:04 <wilcal> #topic Who's new? 20:05:11 <wilcal> anyone here new? 20:06:09 <wilcal> nobody raised their hand 20:06:17 <tjandrews> I could try re-inventing myself, but at my age it's too much work. 20:06:22 <wilcal> so on to 20:06:34 <wilcal> #topic M5.1 20:06:42 <brian__> hurray 20:06:46 <wilcal> it's done and released 20:06:54 <wilcal> is it in the repos yet 20:07:21 <lewyssmith> I should imagine so by now. 20:07:26 <azziam> I checked the download page last night and it was there 20:07:27 <ennael> it is 20:07:39 <lewyssmith> Any news about it yet? 20:07:52 <ennael> no dead computer for now :) 20:07:59 <wilcal> it's on the mirrors.kernel.org so that makes it super official 20:08:31 <wilcal> it's fine 20:08:46 <lewyssmith> Anne had her work cut out over it. 20:09:04 <wilcal> the boot.iso will install from the repo 20:09:14 <wilcal> so really M5 is behind us 20:09:17 <DavidWHodgins> Back 20:09:24 <wilcal> Welcome back 20:09:49 <wilcal> #chair DavidWHodgins 20:09:49 <Inigo_Montoya> Current chairs: DavidWHodgins lewyssmith wilcal 20:09:50 <tjandrews> So Dave can be Dave this week? 20:09:56 <DavidWHodgins> :-) 20:10:15 <wilcal> we were just saying M5.1 is released and on the mirrors 20:10:22 <lewyssmith> wilcal: You are typing fast tonight; beat me twice. 20:10:23 <DavidWHodgins> Thanks everyone for all of the hard work getting Mageia 5.1 out 20:10:24 <ennael> could be a nice song title 20:10:35 <DavidWHodgins> lol 20:10:42 <wilcal> It built a really good iso team 20:10:50 <DavidWHodgins> #topic * Mageia 6 - Here we go again :-)) 20:11:03 * ennael runs 20:11:35 <DavidWHodgins> I forgot to send out the msg to the qa iso testers group, but it looks like most people figured out what's needed 20:11:44 <wilcal> there's still some ( major ) wrinkles in M6 but for the most part it's in pretty good shape 20:11:55 <tjandrews> One minor niggle, just discovered, hardly worth mentioning, but... 20:12:21 <wilcal> lots of work to clear the "Good Luck" bug 20:12:21 <DavidWHodgins> Are we expecting the live iso images before sta2 release? 20:12:27 <tjandrews> The login and logout sounds in Plasma have been disabled by default. 20:12:42 <hviaene> ?????Minor???? Like not booting???? 20:12:45 <wilcal> lots of notification sounds are not working 20:13:02 <tjandrews> No, no. Just the sounds not working. 20:13:11 <DavidWHodgins> Those are minor bugs that won't stop sta2 from being released 20:13:11 <wilcal> Ahhhh 20:13:21 <wilcal> Auto login still does not work 20:13:21 <azziam> I have 9 M6 64-bit installs acting pretty nice. Haven't heard notification sounds and forgot they might exist 20:13:50 <tjandrews> Easy to activate, if you know whay to do. Like most things. 20:14:00 <tjandrews> whay=what 20:14:10 <azziam> Auto-login is working for me except in one instance. 20:14:39 <tjandrews> I'm on my laptop, in Mga6. Not used to the keyboard even yet. 20:15:20 <lewyssmith> How good is sta2 at present? Much to fix? 20:15:22 <DavidWHodgins> ennael: Any eta for next buils? 20:15:32 <DavidWHodgins> s /buils/builds/ 20:15:33 <lewyssmith> We should just mention the pre-test initiative. Not sure whether it will happen before next ISO rounds, but it would be nice if it could. 20:16:23 <Bequimao> what about uefi support? I can't even start without uefi support. 20:16:32 <azziam> I have a few instances of not being able to open MCC. Anyone else? 20:17:07 <wilcal> Can we agree, or not, that the EFI on Vbox is a valid test? 20:17:19 <wilcal> a Vbox client 20:17:36 <tjandrews> Any idea when the new nvidia drivers are due tol float here from upstream? 20:17:45 <DavidWHodgins> Not really. If it works great, but if it doesn't, that isn't a blocker type bug. 20:17:58 <lewyssmith> wilcal: Bill: forget EFI in VBox. A fatuous situation. 20:18:01 <tjandrews> tol=to Sigh. 20:18:08 <tmb> wilcal, efi support in vbox sucks 20:18:25 <lewyssmith> It is simply not important. 20:18:59 <tmb> tjandrews, no, no info yet... Aaron told me he cant give any ETA yet. 20:19:13 <tjandrews> :( 20:19:26 <DavidWHodgins> Bequimao: It's not a good idea to mix uefi and non-uefi installs on the same system. On some systems booting into non-uefi mode will wipe the nvram data 20:19:36 <wilcal> rats I knew you were work'n that pretty hard when we first started with EFI 20:19:48 <tmb> tjandrews, but you can test on nvidia hardware ... it will use nouveau, or you can try to boot with "xdriver=modesetting" 20:20:32 <Bequimao> DavidWHodgins: I system does not have an CSM mode. Only uefi or legacy mode. 20:20:32 <lewyssmith> Bequimao: The EFI boot fix is already in place for the next ISOs; I think. 20:20:51 <lewyssmith> Join the crowd! 20:21:40 <azziam> Seems the next thing is to get isos with UEFI. The installer seems to be working nicely on my legacy machine. 20:21:42 <DavidWHodgins> One of the problems with uefi, is that each manufacturer has their own interpretation of how to do it. It isn't as standardized as legacy bios booting 20:22:19 <lewyssmith> Such differences are not visible to our installers. 20:22:21 <wilcal> it's one setting on one of my test systems MB 20:22:47 <wilcal> so now that M6 has kinda settled down I'll tinker with it 20:22:53 <azziam> I'll be trying a Gigabyte "hybrid EFI" board and holding my breath. 20:23:11 <wilcal> I have one of those too 20:23:27 <wilcal> I fear setting it and not being able to unset it 20:23:39 <lewyssmith> Setting what? 20:23:44 <azziam> I don't care about unsetting it. 20:23:46 <wilcal> killdisk should wipe the drive 20:23:58 <wilcal> setting efi option on the mobo 20:24:12 <DavidWHodgins> It's the nvram that has to be set/unset to control booting into efi or non-efi mode 20:24:48 <DavidWHodgins> Think of the nvram as a really small, slow usb stick that's always present, but not a standard /dev/sd device. 20:25:22 <azziam> nvram is part of CMOS? 20:25:54 <lewyssmith> MTAIM M6 is poised to move ahead in good shape. 20:25:59 <lewyssmith> BTAIM 20:26:02 <DavidWHodgins> It's similar to cmos, but doesn't have the limited number of writes that cmos has 20:26:11 <tjandrews> This laptop is my only EFI hardware, carefully left in legacy mode the way it was when I got it. Fewer headaches, I think. 20:26:34 <azziam> Headaches need testing too. 20:26:55 <tjandrews> By someone other than me. ;) 20:27:26 <azziam> :) 20:27:37 <wilcal> I ran into this a couple days ago: 20:27:38 <wilcal> https://bugs.mageia.org/show_bug.cgi?id=19904 20:27:40 <[mbot> Bug 19904: normal, Normal, shlomif, NEW , VLC segmentation faults on launch from terminal, vlc-2.2.4-8.2.mga6.tainted.x86_64.rpm 20:27:47 <ennael> (sorry connection failed...) 20:27:52 <wilcal> not fully sure what's going on 20:28:15 <lewyssmith> Accepting that new M6 ISOs are near, shall we move on? 20:28:26 <wilcal> I think that's the best 20:28:35 <DavidWHodgins> legacy hardware has the boot software (bios) stored in rom with drive parameters etc., stored in cmos. uefi hardware has the boot software stored in rom, with the list of os's etc, stored in nvram 20:28:48 <DavidWHodgins> ennael: Any eta for next sta2 builds? 20:29:15 <ennael> well after speaking with lewyssmith we decided to have pretests to avoid crappy isos uploaded 20:29:23 <lewyssmith> ennael: Anne, just sent you an e-mail. 20:29:24 <azziam> I thought Anne had actually run away! 20:29:31 <ennael> no :) 20:29:40 <ennael> I will release new stage2 just after 20:29:48 <ennael> and then build again isos 20:30:12 <DavidWHodgins> So, hopefully, some time tomorrow? 20:30:22 <wilcal> for the weekend for sure 20:30:31 <ennael> yep 20:30:38 <wilcal> FWIW I have a very busy weekend 20:30:46 <azziam> So, someone will pretest on UEFI? Is that the only change besides newer kernel? 20:30:50 <wilcal> but should be able to put some cycles in it 20:30:52 <ennael> lewyssmith: can you explain about pre tests for isos ? 20:31:27 <tjandrews> Reminds me of my grandmother. Whenever life was giving her a hard time she threatened to "run down the road with a blanket over her head," 20:31:29 <tmb> ennael: I think I should enable rabbit as buildnode temporarily to flush build queue, any objections ? 20:31:37 <ennael> nope 20:31:42 <ennael> no rush for now 20:32:28 <lewyssmith> As you wish. We are setting up a procees of minimally pre-testing new ISOs just to ensure they basically work, *before* announcing them for general testing. Involving a few people 'behind the scenes'. 20:32:33 <DavidWHodgins> We've done pre-tests in the past. One or two people download/test iso images before they are made available to all of the qa iso testers. Means bootable iso images, but delays getting them to testers group 20:32:57 <lewyssmith> But they should be testable the. 20:33:04 <wilcal> dead isos do no one any good 20:33:46 <lewyssmith> The organisation is happening, and when it is in place will be explained. 20:34:04 <tjandrews> Why not delay the bad isos until they aren't bad? 20:34:22 <lewyssmith> Exactly what pre-testing is about. 20:34:27 <DavidWHodgins> My opinion is that people should watch the discuss ml. If initial reports are that it won't boot, abort the sync. Who ever is first to test should update the list asap. 20:34:29 <ennael> tjandrews: because sometimes for any reasons we do not test properly isos before asking QA to test ir 20:34:34 <wilcal> IMO just making sure they boot to a desktop 20:35:17 <ennael> yep 20:35:23 <lewyssmith> The idea is to obviate the need for ISO bulders to test them at all. QA do. 20:35:45 <lewyssmith> We can make *their* life easier. 20:36:08 <lewyssmith> And that of the host of ISO testers. 20:36:22 <DavidWHodgins> Having just a few qa iso testers do the initial testing creates another bottleneck/delay. Not everyone is available 24/7. 20:36:36 <DavidWHodgins> I'm not in favour of it. 20:36:52 <lewyssmith> It is being sorted, Dave. And major errors should be fixed sooner. 20:37:01 <hviaene> I am in favour 20:37:27 <lewyssmith> DavidWHodgins: Dave, you will not notice. 20:37:28 <DavidWHodgins> No faster than having the iso images available to everyone, and whoever finds the problem first reports it 20:37:48 <hviaene> It's one of the most annoying things to distribute a dead ISO to a bunch of paople 20:37:58 <hviaene> people 20:38:03 <DavidWHodgins> Claire and I have done pre iso testing in the past, but only in very select cases 20:38:04 <lewyssmith> Please do not let us argue about it; it can be started or dropped icognito. 20:38:12 <DavidWHodgins> :-) 20:38:13 <lewyssmith> incognito 20:38:14 <wilcal> Only takes me 15-min to find out if a CI installs Vbox client and boots to a working desktop 20:38:27 <lewyssmith> Takes me 2h! 20:38:32 <wilcal> less for a Live-DVD to boot to a desktop 20:38:44 <hviaene> Hm, downloading takes more than the test 20:38:50 <DavidWHodgins> wilcal: Depends on the connection speed, for syncing the iso images 20:39:27 <lewyssmith> Anyhow, things are moving in a good direction for the future. 20:39:33 <tjandrews> That's my only potential problem. My connection can be erratic. 20:39:34 <wilcal> ya the latest set up takes considerably longer but I start it just before I hit the sack 20:39:44 <wilcal> and test it the next morning 20:40:00 <DavidWHodgins> So Mageia 6 testing is on hold till next iso builds are available. Correct? 20:40:17 <wilcal> Ya I would say so 20:40:31 <lewyssmith> Then it will fly. 20:40:33 <DavidWHodgins> #info Mageia 6 testing is on hold till next iso builds are available. Likely tommorow, or this weekend 20:40:38 <wilcal> start from scratch with the next isos 20:40:43 <DavidWHodgins> #topic * Testing updates - Any difficulties, problems, issues? 20:41:29 <DavidWHodgins> I really dislike testing xen. :-) I have too many installs on my system as is. 20:41:56 <lewyssmith> People have done a marvellous job getting the list down. It is very reasonable at the moment - xen apart... 20:42:06 <Luigi12> there are new XSAs since the xen build anyway :o( 20:42:19 <DavidWHodgins> Luigi12: lol 20:42:20 <lewyssmith> Helo David! 20:42:29 <Luigi12> late lunch today :o) 20:42:41 <DavidWHodgins> So hold off xen testing? 20:42:47 <Luigi12> tmb should answer that 20:42:53 <lewyssmith> DavidWHodgins: If you know xen, you have just volunteered to test the update. Thanks. 20:43:15 <DavidWHodgins> Yeah, I'll do it. :-) 20:43:40 <tmb> Luigi12, yeah, but xsa-200 is extremly hard to abuse (and mostly a minor info leak), xsa-201 is arm, and others are a couple of weeks before dropping out of embargo 20:43:47 <wilcal> How do you test it David? 20:43:54 <Luigi12> tmb: ok 20:44:15 <DavidWHodgins> Have to boot to an install that starts xen, which then starts the kernel etc. 20:44:31 <hviaene> Have to go, good night all 20:44:36 <wilcal> nite 20:44:47 <DavidWHodgins> I think it's in the wik. If not, in older xen bug reports 20:44:58 <DavidWHodgins> s /wik/wiki/ 20:45:06 <wilcal> thks 20:45:52 <wilcal> BTW huge team effort on the Kernel yesterday 20:45:59 <DavidWHodgins> subversion is easy to test. lewis and I use if for adding the advisories to svn 20:46:05 <wilcal> really outstanding 20:46:13 <lewyssmith> QA people have been great. 20:46:15 <brian__> yeah - I'm guessing my one issue is a VB setup issue on my part 20:46:29 <Luigi12> the svn client works fine, I use it for packaging too, but security issues are usually in the Apache module 20:46:31 <brian__> saw it on 32 as well 20:46:43 <DavidWHodgins> Note that xen can only be tested on real hardware, not under vb 20:47:03 <brian__> yeah - I did it once, it was one heck of a challenge 20:47:16 <tjandrews> I went to check the kernel out on my brother's computer this morning, but it had already been pushed. 20:47:26 <Luigi12> apache-mod_dav_svn is what really needs to be tested 20:47:35 <DavidWHodgins> I've done the server testing before, but it's been a long time. Have to look at the old bug reports to see what I did 20:47:41 <wilcal> Ya a hyper critical bug 20:47:52 <Luigi12> yeah I linked the procedure in the bug 20:48:40 <wilcal> thks 20:49:21 <DavidWHodgins> mcabber, I don't remember ever testing before. (Haven't looked at it yet) 20:49:35 <DavidWHodgins> tomcat we have procedures for 20:49:59 <lewyssmith> It is surprisingly easy. I will do x64 tomorrow. 20:50:48 <lewyssmith> DavidWHodgins: Shall I do the phpmyadmin advisory, or do you want to? 20:50:58 <DavidWHodgins> I'll do it after this meeting 20:51:17 <lewyssmith> OK. 20:51:54 <DavidWHodgins> python-tornado we have old bug reports for, to look at 20:52:10 <DavidWHodgins> Same with openafs 20:52:35 <Benmc> (Made it) 20:52:38 <DavidWHodgins> The backport for flightgear is easy to test. Just a big download though 20:52:42 <DavidWHodgins> Benmc: HiYa 20:52:44 <lewyssmith> Welcome Ben. 20:53:06 <DavidWHodgins> #topic * Luigi's roundup - Security news & expected updates 20:53:27 <DavidWHodgins> http://madb/mageia.org/tools/security 20:53:54 <lewyssmith> It is disheartening. 20:54:01 <DavidWHodgins> Oops http://madb.mageia.org/tools/security 20:54:03 <[mbot> [ Mageia App Db - Security issues ] 20:55:04 <lewyssmith> Luigi12: Anything to add, David? 20:55:11 <Luigi12> not really 20:55:20 <lewyssmith> Fine. 20:55:20 <DavidWHodgins> Such as any expected updates that are not security updates? 20:55:38 <DavidWHodgins> That only leaves ... 20:55:39 <Luigi12> not that I can think of 20:55:41 <DavidWHodgins> #topic * Anything else? 20:55:54 <wilcal> I'm ready for some new M6 isos 20:56:01 <lewyssmith> A quickie from me, please. 20:56:03 <brian__> Anybody else run into problems in the VM i586 client when installing the updates? 20:56:04 <DavidWHodgins> :-) 20:56:26 <brian__> lewysmith: go ahead 20:57:11 <lewyssmith> There was NO dispute about the idea of removing the checksums from the PAD. Can we confirm & do that? A bit less work again. 20:57:34 <DavidWHodgins> Fine with me, provided the date/time are included 20:57:49 <lewyssmith> That is very necessary. 20:57:50 <brian__> ok with me 20:58:03 <DavidWHodgins> We have had multiple iso images in the same day before 20:58:25 <DavidWHodgins> Rare, bug does happen 20:58:33 <DavidWHodgins> s /bug/but/ 20:58:43 <DavidWHodgins> Need to clean my keyboard. :-) 20:58:51 <brian__> granted - on second you could do MD5 20:59:35 <lewyssmith> The timestamp is *the* key identifier. Why even MD5 on the PAD? 20:59:37 <brian__> but a good point 20:59:53 <DavidWHodgins> I think date/time is enough for qa testing. The hashes should all be checked before releasing any iso images though. 21:00:07 <brian__> so timestamp liked 20161208150125 21:00:40 <wilcal> I've never seen a MD5 sum error 21:00:40 <brian__> less seconds I suppose 21:00:46 <DavidWHodgins> The contents of the DATE.txt file 21:01:22 <tjandrews> I have, but it was always a human error. 21:01:32 <brian__> ok 21:01:36 <lewyssmith> I am done. Thanks to the many people who showed up tonight, & all your good work. 21:01:55 <DavidWHodgins> There have been times when the DATE.txt file was not updated, but that's been rare and quickly spotted. 21:01:59 <wilcal> I'm done 21:02:03 <brian__> ditto 21:02:08 <DavidWHodgins> Countdown time then 21:02:12 <DavidWHodgins> t - 5 21:02:13 <wilcal> bye all 21:02:16 <DavidWHodgins> 4 21:02:19 <DavidWHodgins> 3 21:02:20 <azziam> bye 21:02:22 <lewyssmith> Goodbye everyone. 21:02:23 <DavidWHodgins> 2 21:02:23 <tjandrews> Bye. 21:02:26 <DavidWHodgins> 1 21:02:31 <DavidWHodgins> Thanks for coming everyone 21:02:39 <DavidWHodgins> #endmeeting