19:09:26 <ennael> #startmeeting
19:09:26 <Inigo_Montoya> Meeting started Mon Apr  6 19:09:26 2015 UTC.  The chair is ennael. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:09:26 <Inigo_Montoya> Useful Commands: #action #agreed #help #info #idea #link #topic.
19:09:36 <ennael> hi all thanks for attending this meeting
19:09:53 <marja> thanks for leading it :-)
19:10:53 <ennael> ok so proposal is to review last critical bugs and speak about RC release
19:13:23 <Stormi> can you remind us with the link to the list?
19:14:21 <marja> https://bugs.mageia.org/buglist.cgi?priority=release_blocker&order=Importance&query_format=advanced&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED
19:14:30 <DavidWHodgins> Stormi: http://lstu.fr/Mga5ReleaseBlockers
19:14:42 <marja> DavidWHodgins: your's is nicer :-)
19:14:56 <DavidWHodgins> :-)
19:16:10 <ennael> oups sorry internet cxion is making some joke tonight
19:16:25 <Stormi> should https://bugs.mageia.org/show_bug.cgi?id=13901 be release blocker?
19:17:03 <ennael> there is already a fix proposed
19:17:07 <ennael> needs to be tested
19:17:47 <Stormi> already tested
19:17:55 <Stormi> it fixes an issue but fails afterwards
19:18:00 <Stormi> tv asked me to open a bug
19:18:06 <Stormi> but maybe I should just reuse this one
19:18:13 <ennael> yep
19:19:15 <ennael> ok let start review
19:19:16 <ennael> https://bugs.mageia.org/show_bug.cgi?id=3910
19:19:30 <ennael> that one is an old one and for now looks frozen
19:20:23 <tmb> yeah, I dont consider that one a release blocker...
19:20:36 <marja> when bugs are very old, tv says they can't be release blockers because it's always been like that
19:20:58 <ennael> release blocker is a bug that cannot fixed after isos are out
19:22:04 <wilcal> IMO any bug existing in M3 or earlier should not be an M5 release blocker
19:22:10 <marja> ennael: true, but he un-release-blocked a bug once, saying it has been like that since 10 yrs (admitted, that is longer than 3 or 4)
19:22:57 <ennael> marja: I'm not speaking about this but about the specific bug listed above :)
19:23:16 <tmb> just move it to mga6 and we debate it again then :)
19:23:34 <DavidWHodgins> :-)
19:23:38 <marja> ennael: if few people are affected, then I'm fine with moving to Mga6
19:23:48 <DavidWHodgins> Agreed
19:24:07 <ennael> Akien: can you remind me the tracker for mageia 6 ?
19:24:36 <tmb> well, since the simple workaround for 3910 is "dont close the lid" :)
19:25:07 <DavidWHodgins> https://bugs.mageia.org/show_bug.cgi?id=15527
19:25:37 <ennael> yep thanks
19:25:53 <ennael> https://bugs.mageia.org/show_bug.cgi?id=10024
19:26:24 <ennael> bootloader does not run grub2-mkconfig when grub2 is not installed to MBR
19:28:36 <DavidWHodgins> I'd be tempted to suggest using grub legacy, unless installing on a uefi system.
19:28:38 <marja> there'll be few people installing grub2 legacy someplace else than in the MBR
19:28:50 <marja> I mean Grub2-non-efi
19:29:10 <ennael> so could be mentionned in errata
19:29:16 <ennael> if you are all ok
19:29:23 <marja> yep, fine
19:29:23 <DavidWHodgins> I'm ok with that.
19:29:36 <ennael> volunteer ? :)
19:31:04 <ennael> they are all gone to look for eggs :)
19:31:15 <DavidWHodgins> It's already there, under multi-booting
19:31:21 <ennael> ok then
19:32:00 <marja> ennael: this one should maybe be added to the errata, too ... it is only valid on non-efi grub2 systems (didn't test very recently, though) https://bugs.mageia.org/show_bug.cgi?id=11827
19:32:19 <ennael> https://bugs.mageia.org/show_bug.cgi?id=11105
19:32:26 <ennael> RAID ( ddf / isw ) broken for new install
19:34:35 <tmb> this one should be somewhat fixed as long as we force isw* to be handled by dmraid
19:34:47 <ennael> ok so the fix in isolinux.cfg?
19:35:11 <tmb> ennael, the "noiswmd" yes
19:35:20 <ennael> ok I will do it just after meeting
19:36:06 <Akien> Hi there, sorry for running late :)
19:36:21 <ennael> https://bugs.mageia.org/show_bug.cgi?id=12305
19:36:22 <tmb> but we have one issue... it works oob on legacy installs, but switch on efi and we crash :)
19:36:37 <ennael> System crashed after upgrading from Mga3 to Cauldron. Stops after splash screen with message, /dev/resume does not exist.
19:36:41 <ennael> argh damned
19:37:19 <tmb> but we are reworking the drakx code with tv, so it should hopefully work out nicely...
19:37:41 <ennael> ok any time it will be available?
19:40:04 <tmb> that really depends on tv .. I have debugged the falure the parts of drakx /dracut/dmraid / isw* stuff... but to get all perl code correct....
19:40:30 <ennael> I mean is it realistic to think we will have it for mga5 ?
19:41:04 <tmb> but we can release RC without the fix if needed and see if/how it works out for final
19:41:13 <ennael> ok
19:41:16 <DavidWHodgins> Agreed
19:41:43 <ennael> ok so
19:41:43 <ennael> https://bugs.mageia.org/show_bug.cgi?id=12305
19:41:48 <ennael> System crashed after upgrading from Mga3 to Cauldron. Stops after splash screen with message, /dev/resume does not exist.
19:43:02 <tmb> hm, that should have been fixed in dracut-038-13.mga5, but I see we never pushed it to release, only to testing
19:43:32 <Akien> Time to get fixes in release IMO, there aren't many people testing updates_testing
19:43:45 <Akien> (especially nobody from QA since they test mostly ISOs)
19:43:54 <DavidWHodgins> And then the iso images rebuilt
19:44:08 <marja> Martin whitaker tested it, but it failed, too
19:44:56 <ennael> the one in testing ?
19:45:29 <marja> ennael: yes, he says "installing dracut-038-13.mga5"
19:45:48 <wilcal> repeated changes to M5 isos require testing and take away cycles in QA for testing updates
19:45:53 <marja> after "tested this by"
19:48:40 <ennael> shall we first move it from testing to release ?
19:49:55 <DavidWHodgins> I think so, so it will be in next iso images. I don't think the testing repos should still be in use for cauldron.
19:51:40 <ennael> ok
19:51:57 <ennael> https://bugs.mageia.org/show_bug.cgi?id=14520
19:52:06 <ennael> Outdated ldconfig cache resulting in "Oh no! Something has gone wrong" in GNOME (liveDVD with xdriver=...)
19:52:19 <Stormi> big hammer
19:52:26 <DavidWHodgins> Isn't that fixed by the forced reboot?
19:52:43 <wilcal> I've not seen this issue lately
19:52:48 <ennael> but bug is waiting for feedbacks :)
19:52:54 <ennael> if fixed then it has to be closed
19:54:15 <tmb> well it didn't show up on my testsystem atleast 10+ installs in a row so ... :)
19:54:41 <tmb> (after the "big hammer" fix that is...)
19:54:46 <marja> :-)
19:55:30 <DavidWHodgins> Just to be clear, is this bug is only about installed systems, not live mode?
19:59:45 <wilcal> brb
19:59:47 <spturtle> this bug was only about live mode
20:00:46 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15060
20:00:53 <ennael> network in virtualbox is horribly slow (NAT)
20:01:42 <marja> Stormi: were you the one with the missing cursor...... Paul Dufresne reports that, too, in comment 33 of bug 14520
20:01:48 <ennael> so I've seen errata proposal and fix to be included
20:03:12 <tmb> for 15060, suggest errata with usning virtio as nic, and the "ethtool ..." change as an other option...
20:03:55 <ennael> what about Luigi12_work comment?
20:04:22 <Stormi> marja: about missing cursor I've tried to reproduce with LiveDVD... But it hangs before X even starts :)
20:04:34 <Stormi> probably an nvidia optimus issue
20:05:02 <marja> Stormi: :-/
20:06:07 <tmb> well, it's mostly "underpowered vms" that hit the GRO issue, ... and adding automatic disabling of GRO in kernel depending on if we run in vbox or not is not a fun thing to do...
20:06:21 <ennael> ok
20:06:32 <DavidWHodgins> tmb: Could it be tied into bug 44?
20:06:36 <ennael> can you add it in errata? unless it"s already done
20:07:13 <tmb> DavidWHodgins, could be
20:07:47 <tmb> on my systems I dont hit either 44 or 15060
20:08:07 <DavidWHodgins> Luigi12_work: Could you test with the kernel option "divider=10"?
20:08:53 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15235
20:09:04 <ennael> Unable to boot, from Live media or after install
20:10:39 <tmb> so that seems to be fglrx closed source issue... not much to do about that :/
20:10:54 <wilcal> I ran into a problem this morning. Installed with the i586 latest boot.iso and it exhibited this bug. Did it twice. More testing here to confirm.
20:11:13 <wilcal> Did it all in Vbox
20:11:45 <tmb> wilcal, wich bug ?
20:12:07 <wilcal> https://bugs.mageia.org/show_bug.cgi?id=15235
20:12:35 <wilcal> i'll put more time into it after this meeting
20:13:18 <ennael> at least a mention in errata?
20:16:10 <Stormi> is that bug likely to hit many users? I mean, is it worth looking for a workaround for final?
20:16:14 <tmb> seems Martin just suggested a simple fix/hack for that... I guess we should try it out on new set of isos
20:17:52 <ennael> yep indeed
20:18:09 <ennael> will add it then
20:18:57 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15253
20:19:04 <ennael> 5b3: Display corruption once the cursor moves (64bit is KO, 32bit is OK) with 16bpp & 24bpp (OK with 15bpp)
20:22:08 <tmb> I think that goes to errata for now
20:22:09 <DavidWHodgins> tmb: As a workaround, could 15bpp be forced?
20:22:28 <DavidWHodgins> Or at least defaulted to?
20:23:18 <tmb> nope, there are also hw out there not liking 15bpp at all, so we would shift the problem
20:24:12 <DavidWHodgins> Ouch
20:26:05 <tmb> anyway, in the 3.19.3 kernel I also have backported fbdev "sanity" checks from 4.0 so it will be interestng to see if they make any difference
20:29:44 <marja> DavidWHodgins: wilcal: isn't there a tracker for Mga 4 -> 5 upgrade issues?
20:30:18 <wilcal> yes there is
20:30:36 <marja> wilcal: do you mind giving me the link?
20:31:21 <DavidWHodgins> I don't see one
20:31:37 <marja> DavidWHodgins: maybe there are nearly no issues :-)
20:31:47 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15257
20:31:54 <ennael> Live installer (not live mode) partitioner: Available disk space is not computed correctly (at least in small VM disks)
20:32:17 <ennael> so that one also deals with errata
20:32:22 <wilcal> Hmm I've filed individual bugs
20:34:06 <wilcal> does not look like there's an overall
20:34:08 <tmb> yeah, errata for 15257, as I haven't found yet where it misses the fact that it knows about the needed disk size
20:35:51 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15350
20:35:58 <ennael> Upgrade failed mga4 to mga5 when adding online media to DVD - 195 transactions failed
20:36:19 <ennael> we are waiting for tv on that one
20:37:06 <DavidWHodgins> I just completed a test with online media added, and selecting everything except individual package selection. Didn't see any problems during install, just checking the log files.
20:37:33 <Stormi> I had an issue during network install but it was probably mirrors syncing
20:37:53 <Stormi> it brought me back to package selection, I proceeded again and it worked
20:37:55 <marja> wilcal: I opened a tracker, please add any upgrade issues in it https://bugs.mageia.org/show_bug.cgi?id=15637
20:38:04 <tmb> and in worst case errata: try offline upgrade first, then update the rest
20:38:14 <ennael> DavidWHodgins: you mean upgrade?
20:38:35 <Stormi> oops was install for me
20:39:19 <DavidWHodgins> I'm going to close 15350, as no transactions failed here
20:39:38 <marja> lol, just added it to the tracker
20:39:59 <ennael> DavidWHodgins: ok
20:40:18 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15374
20:40:25 <ennael> btrfs / and /home on sda7 and sda8 correctly forces grub2 but is unable to install it
20:40:37 <ennael> tmb: is the last commit fixing this bug?
20:40:57 <tmb> ennael, yep, tested and all :)
20:42:03 <ennael> great you can close it then :)
20:42:35 <tmb> yeah, I just need to release new stage2 & drakxtools
20:42:41 <ennael> ok
20:43:10 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15400
20:43:19 <ennael> Problem installer when creating raid partitions (INTERNAL ERROR: unknown device md0)
20:43:55 <tmb> should be fixed with the commit that is listed last n that report
20:44:35 <DavidWHodgins> Is that fix in the latest iso images?
20:46:15 <ennael> should be given the date
20:46:20 <tmb> should hve been, as the fix was in stage2 16.73
20:46:57 <DavidWHodgins> Argh. Reopening 15350, as I tested a clean install, not an upgrade. Will retest with an upgrade.
20:47:06 <ennael> yerk
20:48:58 <DavidWHodgins> I'll test that after the meeting.
20:49:51 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15468
20:50:07 <ennael> 5RC: Gnome Live install "Oops something has gone wrong" ("Gjs-WARNING **: JS ERROR: Error: Argument 'string' (type utf8) may not be null")
20:50:22 <ennael> we need feedback on that one
20:50:28 <ennael> it "should" be fixed
20:50:34 <Luigi12_work> sorry, I was teaching.  divider=10 is irrelevant to the virtualbox issue
20:50:42 <DavidWHodgins> Luigi12_work: Ok, thanks.
20:50:49 <wilcal> I've not seen that error in awhile
20:51:35 <Luigi12_work> I don't think you can run ethtool in stage1, so at least for loading stage2 that stinks, but you probably can run it in stage2, I'm not sure if ethtool is built into stage2 so that it would be available immediately
20:51:58 <tmb> 15468 definately should be fixed by the "big hammer" fix
20:52:07 <ennael> :)
20:52:11 <Luigi12_work> I was hoping the fact that the ethtool command worked would help indicate what the problem is.  Maybe the command is a workaround but maybe there's a better solution that can be found
20:52:11 <ennael> so we can close it?
20:52:38 <wilcal> I think so
20:52:42 <tmb> ennael, yeah
20:53:55 <Luigi12_work> as for bug 3910, vbetool needs to be fixed.  The other aspect doesn't need to be a blocker (the original bug), but vbetool needs to be fixed
20:54:02 <tmb> Luigi12_work, well, the help from disabling GRO is because it disables buffering since the virtual cpu chokes on it...
20:54:48 <Luigi12_work> that sounds yucky
20:55:08 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15486
20:55:15 <ennael> Upgrade from Mga4 does not update vmlinuz or initrd.img symlinks and new initrd still has Mga4 plymouth
20:55:18 <tmb> Luigi12_work, yeah... it's listed as a workaround for other virtualization platforms too
20:57:13 <tmb> 15486 needs to be re-tested... atleast the new plymouth is also now forced with proper conflicts...
20:57:31 <ennael> ok so special focus with coming isos
20:58:19 <ennael> damned collision :)
20:58:41 <ennael> https://bugs.mageia.org/show_bug.cgi?id=15583
20:58:50 <ennael> Installing to '/' in UEFI writes into ESP and changes default bootloader
21:00:22 <ennael> so it seems bug fix proposed upstream
21:00:37 <DavidWHodgins> I thought grub2-efi had to be installed to the mbr.
21:00:39 <tmb> upstream thread is dead for now ... no changes done to drakx
21:00:55 <ennael> yep so I guess to be added for mageia 6
21:01:28 <marja> DavidWHodgins: no, it is for gpt partioned disks, there is no mbr
21:01:36 <DavidWHodgins> Ah. Ok
21:01:58 <tmb> yeah... addinng efi support really opened up a can of worms.... :)
21:02:06 <DavidWHodgins> Yep
21:02:35 <DavidWHodgins> I really need to learn more about gpt and nvram handling.
21:02:35 <Luigi12_work> just to be clear on 3910, vbetool being broken is a new regression in mga5, so it hasn't always been like that
21:02:48 <tmb> as I initially also stated that there is a lot of "assumptions" in drakx code that went invalid
21:04:28 <tmb> Luigi12_work, yeah, vbetool needs fixing, maybe testing different compilation/linker flags if its the toolchain that broke it
21:05:20 <ennael> so what about the last bug?
21:06:26 <tmb> errata if we cant fix it n time... we have more important bugs to squash
21:07:02 <Luigi12_work> I can try to experiment with vbetool compilations on Thursday I think
21:07:21 <Stormi> would be nice if someone else than me could test https://bugs.mageia.org/show_bug.cgi?id=13901#c10
21:07:29 <Stormi> to see if it goes to errata
21:07:35 <ennael> ok
21:07:42 <ennael> that's all for the list
21:08:00 <marja> DavidWHodgins: you suggested to keep this one open.... is that still needed? https://bugs.mageia.org/show_bug.cgi?id=5746
21:08:24 <wilcal> tmb a quick update on the status of UEFI all of
21:09:13 <DavidWHodgins> marja: It needs to be retested before that can be decided.
21:09:37 <DavidWHodgins> With multiple ways of upgrading, it's tedious to test them all.
21:09:58 <marja> DavidWHodgins: I'll add it to the upgrade tracker, that'll increase the chance someone will test it
21:10:37 <DavidWHodgins> That reminds me. With Mageia 4, installation added cauldron repos instead of Mageia 4 repos. What has to be done, to ensure that doesn't happen again?
21:12:08 <marja> DavidWHodgins: on the mirrors Mga5 links to cauldron, I think now 5 should already be added instead of cauldron
21:12:37 <marja> DavidWHodgins: urpmi.cfg should be checked after upgrading
21:13:02 <tmb> well, until we flip mageia-release to "release mode" and push new mirror api for 5 it will add cauldron
21:14:00 <ennael> ok anything else ?
21:14:03 <tmb> we could maybe push mirror api now as I have the "5" symlink in place...
21:14:13 <wilcal> sorry status of UEFI
21:14:22 <wilcal> where are we at
21:15:01 <tmb> wilcal, well it should work for most at this point, even on intel bios fakeraids :)
21:15:43 <wilcal> Do we need to consider a M5.1 delayed 60/90 Days after M5.0?
21:16:05 <DavidWHodgins> wilcal: Only if we find major problems after release
21:16:10 <tmb> now it wont be perfect on mga5, but good enough until mga6
21:16:24 <wilcal> so lots of stuff in errata?
21:16:30 <tmb> there will always be corner cases...
21:16:42 <wilcal> like don't work very good with Software RAID
21:17:13 <wilcal> works perfect with blank drive and not so perfect with everything else
21:17:32 <Stormi> well not bad with already installed windows, does it?
21:17:40 <tmb> wilcal, well we'll see if/when we release RC
21:20:22 <marja> wilcal: not true, it installs and works fine here (2 mageia cauldrons + win8 on local disk, several Mageias on external disk)
21:20:58 <marja> wilcal: a lot of EFI issues got fixed
21:21:38 <ennael> ok can we close the meeting for tonight ?
21:21:50 <tmb> yeah, for efi normal setups should behave nicely... the corner cases will be down to crappy efi firmware, more "unusual" setups and so...
21:22:05 <DavidWHodgins> Should we discuss the aws and release schedule, or leave those for the mailing list, or next meeting?
21:22:35 <tmb> no matter how much we want to try and cover there will also be someone finding new ways of doing stuff :)
21:22:44 <marja> :-)
21:23:04 <ennael> DavidWHodgins: we are submitting last fixes before rebuilding what should be last isos for RC
21:23:35 <DavidWHodgins> Sorry, meant support schedule, as Luigi12_work posted to the mailing list.
21:24:23 <marja> DavidWHodgins: date of next release + 3 months, correct?
21:24:38 <DavidWHodgins> Yes, instead of 18 months.
21:25:03 <marja> sounds like a good idea to me
21:25:35 <DavidWHodgins> Same here
21:26:16 <DavidWHodgins> We are missing a few people here tonight though, so perhaps it would be better to decide using the mailing list.
21:27:00 <marja> DavidWHodgins: I'm not sure all new council members can already post to the ml (Stormi and i don't count)
21:27:34 <DavidWHodgins> new council members need to prompt sysadmins to update ldap
21:27:43 <marja> Latte isn't in the list, for instance
21:28:20 <marja> nor is Otto or sebsebseb
21:28:40 <marja> funny enough, obgr_seneca is in it
21:28:51 <Luigi12_work> the support lifecycle thing can be discussed next week I guess (I'll probably miss that meeting though)
21:28:53 <marja> ah, because he's on the board
21:29:30 <Luigi12_work> I'm not really proposing it as a permanent change either, just for mga5, but it's just an idea.  I want all the teams input to be represented in any decision.
21:29:52 <Stormi> 14 months seems short to me, from a user point of view. 18 is short already.
21:30:01 <Stormi> of course if we can't, then we can't
21:30:51 <DavidWHodgins> We need someone from each team to ensure leaders and deputy leaders are listed in http://people.mageia.org/g/mga-council.html and prompt sysadmins to add missing people to ldap
21:31:01 <Stormi> Actually I would have suggested to envision extending the supported time because our release cycle has become 12 months or more :)
21:31:16 <Luigi12_work> 18 is about as long as any distro does (ones that claim to do more typically only partially support past about that long)
21:31:20 <Stormi> but I know > 18 months is hard, patches become harder to do
21:31:23 <Luigi12_work> and our package set was mostly frozen months ago
21:31:44 <Luigi12_work> I've found in the past once you get around 18 months, getting backported patches and support from upstreams becomes much more difficult
21:32:03 <Luigi12_work> and as I also said, I'm not crazy about this and don't really *want* to reduce our support, it's just being pragmatic
21:32:17 <ennael> ok guys sorry to cut discussion but I'd like to sleep a bit :)
21:32:28 <ennael> you can go on that meeting I can chair someone else
21:32:34 <wilcal> time for more testing for me
21:32:41 * marja likes to sleep, too
21:32:42 <Stormi> time for rest for me too
21:32:42 <Luigi12_work> no problem, you can end the meeting
21:32:48 <Luigi12_work> sorry I couldn't really attend
21:32:52 <Luigi12_work> maybe someday :o)
21:32:55 <DavidWHodgins> Let's end the meeting then.
21:33:00 <marja> Luigi12_work: np :-)
21:33:11 <ennael> thanks all for attending that meeting
21:33:13 <Luigi12_work> good night everyone
21:33:29 <wilcal> bye
21:33:30 <ennael> things are not perfeect but it seems we are very near the end of that hell :)
21:33:38 <marja> we are :-)
21:33:41 <ennael> have a good night
21:33:46 <wilcal> this month for sure for M5
21:33:46 <ennael> #endmeeting