20:14:27 <DavidWHodgins> #startmeeting
20:14:27 <Inigo_Montoya> Meeting started Thu Mar 26 20:14:27 2015 UTC.  The chair is DavidWHodgins. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:14:27 <Inigo_Montoya> Useful Commands: #action #agreed #help #info #idea #link #topic.
20:14:36 <DavidWHodgins> #chair wilcal MrsB
20:14:36 <Inigo_Montoya> Current chairs: DavidWHodgins MrsB wilcal
20:14:53 <DavidWHodgins> #topic Who's new? If you are new then come and say hello!
20:14:56 <wilcal> your in charge David :-)
20:15:10 <DavidWHodgins> Anyone here who hasn't been to a qa team irc meeting before?
20:15:28 <Stormi> hi I'm new
20:15:41 <DavidWHodgins> :-)
20:15:59 <DavidWHodgins> #topic Mageia 5
20:16:06 <tjandrews> Hi, Stormi
20:16:32 <wilcal> ok lots of testing here this morning
20:16:33 <DavidWHodgins> I've just finished syncing latest iso images.  Will start testing, after the meeting. How is everyone else making out?
20:17:12 <wilcal> we've still got the oh no thing with Gnome. 1st restart clears it
20:17:23 <lewyssmith> Little has changed for me for ages, still no real go.
20:17:35 <wilcal> Yipeee. I can Vbox client UEFI install now
20:17:49 <wilcal> that's huge ad thank you tmb
20:18:03 <olivier_cc> I just saw that the partitioning tool was functional in UEFI now
20:18:07 <wilcal> I put a LOT of time into two wrinkles
20:18:07 <DavidWHodgins> wilcal: Is that with today's live iso images?
20:18:11 <wilcal> yes
20:18:44 <olivier_cc> except it creates a second /boot/EFI partition if one from windows already exists
20:18:49 <wilcal> So the slow down of a Vbox client is real and I say it's everywhere
20:19:05 <wilcal> I am only testing in Vbox to a blank drive and that works well
20:19:08 <DavidWHodgins> tmb: As a workaround, what do you think about the idea of forcing a reboot, after the root/user passwords are set?
20:20:13 <DavidWHodgins> I guess tmb is not here right now.
20:20:26 <Luigi12_work> probably needs a nap
20:20:48 <lewyssmith> One can imagine so.
20:21:05 <wilcal> I think we have a serious problem booting media from a USB drive. Ran into that on real hardware and the boot.iso
20:21:14 <tmb> meh, I was just testing stuff
20:21:44 <tmb> I think I can force a reboot... it's not nice, but it should work
20:21:45 <DavidWHodgins> wilcal: In uefi mode, or bios mode?
20:22:05 <wilcal> I have no UEFI hardware here
20:22:18 <wilcal> well I do i just don't use that mode
20:23:08 <tmb> the testfix in round 4 is not working... (I'm doing a display-manager restart, but it's not enough)
20:23:19 <wilcal> So I attemted to use boot.iso on my i7 system and it just sat there on the initial "boot" screen
20:23:31 <wilcal> sat there forever
20:24:00 <wilcal> I burned the same iso to a CD and it worked just fine
20:24:23 <wilcal> i've seen others comment about booting from USB drive
20:24:44 <DavidWHodgins> wilcal: See https://bugs.mageia.org/show_bug.cgi?id=15422#c6
20:24:45 <[mbot> Bug 15422: normal, Normal, qa-bugs, NEW , Update Isodumper to v0.41, isodumper
20:25:04 <DavidWHodgins> I suspect the usb stick hadn't finished being written.
20:25:55 <lewyssmith> Best to 'sync' & wait a couple of times.
20:25:58 <wilcal> the way I write an ISO to a usb tells me exactly that the writting process is completed
20:26:15 <tjandrews> I've only done it twice, but so far have had no problems booting from usb flash media.
20:26:24 <DavidWHodgins> sync doesn't work for copying an iso image to a device, as there is no file system involved, to sync.
20:26:36 <lewyssmith> Oh.
20:26:45 <Kernewes> I use dd and it works for me
20:26:45 <Luigi12_work> eject works to sync it though
20:27:05 <Luigi12_work> I use dd + eject, never have a problem
20:27:15 <wilcal> I only looked at the boot.iso this morning and did not get a chance to look at the Live media on USB drive
20:27:24 <tjandrews> I used isodumper with the i586 classical iso.
20:27:37 <wilcal> my problem may only be on the boot.iso
20:28:05 <DavidWHodgins> It varies depending on the usb stick. I have one old 16GB stick, where the command prompt doesn't return after dd, until it's finished, and several newer 8GB sticks that return almost immediatly. That's why I suggest using htop to see if the device waits are gone for the kernel threads.
20:28:30 <Luigi12_work> or just use eject :o)
20:29:44 <Kernewes> Luigi12_work: is that it, literally type "eject"?
20:29:52 <Luigi12_work> Kernewes: + the device
20:29:57 <Luigi12_work> so like eject sdb, if it's /dev/sdb
20:30:05 <Kernewes> Luigi12_work: useful tip, thank you
20:30:07 <DavidWHodgins> #info During iso testing, please (in addition to looking for any new bugs), concentrate on testing the bugs in http://lstu.fr/Mga5ReleaseBlockers
20:30:09 <[mbot> [ Log in to Mageia Bugzilla ]
20:30:33 <DavidWHodgins> Luigi12_work: Again, that may not work for some usb sticks, as there is no file system to umount.
20:30:38 <Luigi12_work> don't worry if eject gives an error message, mga4 has a bug
20:30:45 <Luigi12_work> DavidWHodgins: no it works just fine
20:30:49 <Luigi12_work> eject doesn't care about file systems
20:31:20 <DavidWHodgins> I'll have to experiment with that, with a stick that I know can cause problems.
20:31:44 <wilcal> more testing here too
20:32:22 <DavidWHodgins> Anyone have any questions or comments on Mageia 5 iso testing, before we move on?
20:32:29 <Luigi12_work> I have an older USB HDD that I used to have sync problems with and I used to fsck -a after umounting something, but then I figured out that eject always fixes it
20:32:44 <Luigi12_work> it also works with writing ISOs to sticks
20:34:09 <DavidWHodgins> #topic Testing updates
20:34:36 <wilcal> we're really falling behind
20:34:44 <DavidWHodgins> We have a lot of updates to test, most of which are security updates. In addition to iso testing, we need some people to work on testing Mageia 4 updates.
20:34:53 <DavidWHodgins> Any volunteers?
20:35:13 <DavidWHodgins> I'll test the krb5 update, after the meeting.
20:36:08 <olivier_cc> I can make some tests tomorrow
20:36:13 <tmb> I'd like to get the virtualbox one validated if possible so I can start pushing some new kernels too
20:37:02 <tmb> (as I have to push new kmods for vbox, it will replace the ones currently in testing)
20:37:04 <wilcal> i'm pretty happy with the Vbox update
20:37:12 <Luigi12_work> yeah we can validate that one
20:37:58 <DavidWHodgins> I'll get the advisory added to svn after the meeting.
20:38:17 <DavidWHodgins> Then validate it, as I've been using it on x86_64.
20:39:24 <DavidWHodgins> It would help if everyone could pick one update to test. Add a comment indicating you are testing it, then test it.
20:39:44 <wilcal> There's some in there I can poke at
20:40:03 <DavidWHodgins> I won't dwell on the issue, but everyone please help with testing Mageia 4 updates.
20:40:11 <Luigi12_work> most have test procedures, many have the beginnings of one at least
20:40:21 <olivier_cc> if anyone knows suricata and can review my testing on 64 bits and tell me if I can validate it, I can test it on 32 bits
20:40:38 <lewyssmith> I am bogged down writing an article, but it should be done over the weekend. Then I can look in again.
20:41:13 <wilcal> I got my new Raspberry Pi 2 last week  :-) a little bit of a time eater
20:41:48 <DavidWHodgins> #topic Luigi's Roundup
20:42:31 <DavidWHodgins> Hopefully, not to many more security updates in the near future. Wishful thinking, I know. :-)
20:42:47 <wilcal> we can use a break here
20:42:49 <Luigi12_work> yeah tomorrow better not be like last Friday, since I have a class starting in the afternoon
20:43:39 <Luigi12_work> I may have mentioned a quassel DoS issue last week.  Still hasn't received a CVE.  I will update quassel after the mga5 release, so that's not imminent.
20:44:07 <DavidWHodgins> #info quassel update expected after Mageia 5 release.
20:44:10 <Luigi12_work> shibboleth-sp update got pushed in Cauldron, so just waiting on guillomovitch to push it for mga4.  It uses xerces-c so they could be tested together if it comes soon enough.
20:44:44 <DavidWHodgins> #info shibboleth-sp and xerces-c updates expected soon.
20:44:48 <Luigi12_work> ruby-rest-client has a similar security issue to what python-requests had that we just fixed, so hopefully pterjan will be able to look at that soon
20:44:55 <Luigi12_work> xerces-c update is already assigned to QA
20:45:15 <DavidWHodgins> #undo
20:45:15 <Inigo_Montoya> Removing item from minutes: <MeetBot.items.Info object at 0xb697028c>
20:45:37 <Luigi12_work> qemu has some DoS issues where a user (maybe just root) in a guest could cause unbounded resource consumption in the host.  Waiting for Fedora to backport patches.
20:45:41 <DavidWHodgins> #info shibboleth-sp update expected soon, which can be tested along with xerces-c.
20:46:29 <DavidWHodgins> #info ruby-rest-client waiting for packager to update, and qemu dos fix expected.
20:47:03 <Luigi12_work> there's a minor issue for cups where ANSI escape sequences in printer names could be executed in a terminal in some cases.  We'll probably deal with that after mga5.  RH accidentally embargoed it for almost a year and just opened the bug recently :o)
20:47:16 <DavidWHodgins> lol
20:47:51 <DavidWHodgins> #info cups update expected after Mageia 5 release.
20:47:55 <Luigi12_work> batik (java package) has some security issue, waiting on someone (pterjan or david_david maybe) to patch that.  Maybe I'll do it, we'll see.
20:48:39 <DavidWHodgins> #info batik (java package) update for security issue(s) waiting for packager(s).
20:48:58 <Luigi12_work> selinux-policy has a funny issue, where if you have selinux enabled if it finds funny business going on in a file it'll query the rpm database to see which package owns it.  Unfortunately that allows an attacker-controlled filename to inject arbitrary shell commands into a daemon running as root.
20:49:18 <Luigi12_work> that was just announced, no fix available yet
20:49:31 <DavidWHodgins> Saw that on oss security list earler today.
20:49:46 <Luigi12_work> and qt-creator has a built-in SSH client which wasn't verifying SSH host keys.  RedHat has a patch for that.  Hoping neoclust will have a look at it.
20:50:16 <tmb> does that even effect us ? we dont have selinux support enabled anywhere ?
20:50:19 <Luigi12_work> so really not much to say this week, not much seemingly imminent but we'll see
20:50:30 <DavidWHodgins> #info selinux-policy security fix waiting for upstream to fix
20:50:33 <Luigi12_work> tmb: I suppose it could if you went out of your way to enable it
20:50:38 <Luigi12_work> tmb: at the very least it's not a big deal for us
20:50:43 <DavidWHodgins> Agreed
20:51:22 <DavidWHodgins> #info qt-creator fix for ssh key checking waiting for packager.
20:51:46 <Luigi12_work> any questions for me?
20:51:52 <wilcal> not from me
20:51:57 <DavidWHodgins> Not here.
20:52:05 <DavidWHodgins> Thanks for the great work Luigi12_work!
20:52:14 <lewyssmith> As always.
20:52:16 <Luigi12_work> yep
20:52:32 <DavidWHodgins> That just leaves ...
20:52:35 <DavidWHodgins> #topic Anything else?
20:52:52 <olivier_cc> nothing for me
20:53:00 <lewyssmith> Nor me.
20:53:08 <wilcal> Sorry to go back to uefi. Any idea when we get through it all
20:53:36 <lewyssmith> What problems remain?
20:53:37 <DavidWHodgins> wilcal: My understanding is that the main problem right now, is uefi install in a dual boot with windows.
20:53:53 <lewyssmith> And?
20:53:55 <wilcal> yup works just fine i think to a blank drive
20:54:07 <wilcal> at least in Vbox
20:54:11 * Luigi12_work wonders why anyone would want to keep windows
20:54:11 <tmb> yeah, that's the part missing from efi support in installer
20:54:41 <DavidWHodgins> Also, deciding how to handle esp partitions. Multiple, or re-use existing one.
20:54:54 <wilcal> works on blank drive does not work on a dual boot drive
20:55:01 <wilcal> hmmmm
20:55:05 <lewyssmith> DavidWHodgins: Better the latter.
20:55:27 <DavidWHodgins> tmb: I'm wondering if vb failing to store the nvram between boots is enough of a regression to hold the current vb update?
20:56:04 <lewyssmith> No other EFI-capable distribution has a problem with installing alingside Win8.
20:56:26 <tmb> hm, the detecting of existing esp was supposed to be fixed in 16.71
20:57:03 <DavidWHodgins> I'll retest with today's live iso image, to confirm.
20:57:39 <tmb> lewyssmith, well it atleast works when you do manual partitioning, so we are getting there :)
20:57:40 <DavidWHodgins> Any idea what's causing the problem(s) with windows dual boot?
20:58:13 <lewyssmith> tmb: It did for M4... (thanks to you).
20:58:27 <olivier_cc> I could dual boot with latest live DVD
20:58:36 <olivier_cc> from grub menu
20:59:13 <olivier_cc> install from live DVD I mean
20:59:41 <tmb> I will setup a win system here so I can debug it... I will try to test 7/8.1/10
21:00:07 <wilcal> All the new live media  booted to a working desktop for me
21:00:07 <DavidWHodgins> Have fun! ☺
21:00:36 <lewyssmith> tmb: I have Win8.1 & can test what you like with pre-partitioning.
21:00:57 <DavidWHodgins> Anyone testing upgrading from M4 with latest classical iso images?
21:01:42 <wilcal> I tested the last CI's but not the one today
21:01:49 <wilcal> looked pretty good
21:01:55 <DavidWHodgins> phone. brb
21:02:25 <DavidWHodgins> Back. (telemarketer).
21:03:31 <Kernewes> DavidWHodgins: didn't you hang on and try and persuade him or her to use mageia?
21:03:37 <tjandrews> I get lots of calls from that nice Rachel to lower the interest on my credit card. So far, I've resisted.
21:03:56 <DavidWHodgins> #info In addition to testing clean installs, please don't forget to test upgrading from Mageia 4.1 with the classical installer i586 and x86_64 dvd iso images, both with/without uefi.
21:04:09 <DavidWHodgins> Kernewes: Recorded message, so just hung up.
21:04:15 <Kernewes> :(
21:04:35 <DavidWHodgins> This one was for "duct cleaning".
21:04:56 <tmb> lewyssmith, well, capture fdisk -l and gdisk -l before and after install to see if we screw up partition tables
21:05:03 <tjandrews> Kernews, I should try that the next time that guy from "Microsoft Security" calls. ;-)
21:05:19 <Kernewes> tjandrews: :)
21:05:21 <wilcal> I got one of those a couple days ago
21:05:33 <wilcal> talked to him for about 15min
21:05:36 <DavidWHodgins> tmb: fdisk doesn't work with gpt, correct"
21:05:58 <lewyssmith> No.
21:06:07 <lewyssmith> It does, sort of.
21:06:28 <tmb> DavidWHodgins, nope, but I want to see if it picks up any hybrid partition table
21:06:38 <DavidWHodgins> Ok.
21:06:58 <lewyssmith> fdisk  is a dialog-driven program for creation and manipulation of par‐
21:06:58 <lewyssmith> tition tables.  It understands GPT, MBR, Sun,  SGI  and  BSD  partition
21:06:58 <lewyssmith> tables.
21:08:04 <lewyssmith> But still better to use gdsik.
21:08:05 <DavidWHodgins> Ah. Been a while since I've looked at it's man page. It used to (IIRC) have a warning that it didn't work with gpt.
21:08:49 <lewyssmith> Can we finish?
21:09:08 <wilcal> i'm done
21:09:10 <DavidWHodgins> Anyone have anything else before we close the meeting?
21:09:16 <Kernewes> no
21:09:20 <DavidWHodgins> 5
21:09:25 <DavidWHodgins> 4
21:09:28 <tmb> on a nice note, the Intel Broadwell NUC I have here our current Mga5 RC isos seem to work nicely :)
21:09:28 <DavidWHodgins> 3
21:09:36 <lewyssmith> Goodbye all, thanks Dave.
21:09:37 <DavidWHodgins> Good
21:09:39 <DavidWHodgins> 2
21:09:43 <olivier_cc> bye
21:09:44 <DavidWHodgins> 1
21:09:49 <DavidWHodgins> #endmeeting