[05:51] normal to lts is probably ok with Prompt=lts, but the last one looks wrong === ara is now known as Guest22932 [08:25] davmor2, ubuntu is rebuilding with all the fixes from yesterday (not sure you got previous msg, it seems I was disconnected) [08:25] jibel: nice [08:26] jibel: can I assume that netboot would be good to test in the meantime? [08:34] davmor2, netboot is ready to test. [08:42] davmor2, ^ [08:52] jibel: \o/ [09:20] jibel: ftr the album Synth Pop is awesome for testing to :) [09:21] jibel: also 64bit UEFI SecureBoot amd64 netboot under way and burning 32 bit too [09:38] davmor2, there'll be another respin of desktop images [09:38] jibel: not effecting netboot though right? [09:38] davmor2, no it's a ubiquity fix [09:38] davmor2, the live session doesn't start when selected from dm [09:39] jibel: cool I'll carry on with netboot then, I'll do a desktop and server install here and wait on an update regarding the desktops then [09:40] davmor2, OK. I'm doing some non-english tests and verifying results for server [10:19] pitti, This is what happens when the installer asks to restart now - http://imgur.com/9atTGQf [10:21] jibel, davmor2 ^^^^ [10:21] flexiondotorg: this does work on Ubuntu and Xubuntu now; this error message looks like it's trying to access something from the squashfs (which is gone at that point) [10:22] flexiondotorg: to prevent that, the casper-shutdown script tries to cache all the plymouth themes [10:22] maybe the mate theme has its themes/plugins in a path which we don't cover? [10:23] pitti, Ubuntu MATE does have plymouth themes [10:23] pitti, Where is the code that does the caching? [10:23] pitti, Maybe I can submit a merge proposal? [10:23] for path in $(which halt) $(which reboot) /etc/rc?.d /etc/default $(which stty) /bin/plymouth /lib/plymouth /lib/*/plymouth /lib/systemd /etc/systemd /lib/*/libnss_files* /etc/nsswitch.conf /usr/share/fonts/truetype/dejavu/DejaVuSans.ttf /etc/fonts/fonts.conf /etc/fonts/conf.d/60-latin.conf; do [10:24] flexiondotorg: it's in bin/casper-stop in casper [10:24] flexiondotorg: I thought by now we'd cache pretty much anything from plymouth; but it looks like it's still missing something there [10:26] Ubuntu MATE puts it's plymouth themes in /lib/plymouth/themes/ubuntu-mate-logo [10:27] that should be covered by /lib/plymouth [10:27] pitti, Indeed. [10:28] I have release manager rights, but haven't requested a rebuild before. When I hit the "Update Rebuild Status" button, does that rebuild for everybody or go to another screen to select? [10:28] balloons: ^ [10:29] pitti, cyphermox has this same issue some days back in a test release of casper I was testing for him. [10:29] flexiondotorg: yes, cyphermox, infinity and I were working on that last week, and updated the list [10:30] flexiondotorg: we don't cachce /var/spool/plymouth/ and /var/lib/plymouth, though [10:30] flexiondotorg: could you boot a live CD, edit casper-stop in-place to add these two, and check if that helps? [10:30] it's not required in the other flavors, but maybe on mate [10:30] pitti, Just testing stock Ubuntu and then I'll do as your request for Ubuntu mATE. [10:31] flexiondotorg: I tested stock ubuntu approximately 30 times yesterday, when fixing the "x gets stuck on reboot" bug [10:31] and I tested xubuntu today's daily this morning [10:31] pitti, Fun :) [10:32] * flexiondotorg create a new Ubuntu MATE test VM. [10:33] flexiondotorg: you can just boot the live CD [10:34] no need to install AFAICS? [10:34] pitti, Understood. [10:35] flexiondotorg: we should get the same if you merely start the live system and shutdown [10:36] pitti, stock Ubuntu just spat out the squashfs error just like Ubuntu MATE. [10:37] flexiondotorg: hm, I'm definitively not getting that in QEMU or on real iron [10:37] pitti, VirtualBox thing? [10:38] pitti, For completeness - http://imgur.com/FSlbpIK [10:39] flexiondotorg: are you testing an iso or from an USB stick? trying to see what's different [10:39] pitti, iso in a VirtualBox VM. [10:39] so it might be usb vs. iso or VB vs. qemu/real hw [10:39] pitti, Emulated DVD drive. [10:40] I can test a DVD here. I'll burn one. [10:42] flexiondotorg: we tested an USB stick on real hw, worked there too [10:46] pitti: I'm testing real hardware and real dvd's, just call me mr old skool [10:46] davmor2: no, I'm grateful for that! [10:47] pitti: I also do a desktop run from usb on real hw too :) [10:47] pitti: I also cover the 3 major gfx and 2 major cpu manufacturers too :D [10:47] pitti, If I just boot the live session and restart, the system doesn't reboot. [10:48] pitti, If I install and restart, the squashfs errors occur. [10:48] flexiondotorg: what do you see exactly? x going down? plymouth? the "hit enter" text? [10:48] pitti, From the live session only, no squashfs errors. [10:48] you should see all three, then pressing enter shld reboot; that's what we see here, anyway [10:49] pitti, I see this - http://imgur.com/btnxiso [10:50] pitti, I do not see a press enter to eject media on either live session reboot nor at the end of an install. [10:50] *sigh* [10:50] pitti, I can't install the DVD on the system. But I can boot the live session and reboot. [10:50] Back in a few... [10:59] pitti, Booting to live session from DVD on real hardware and selecting shutdown work correctly. I got the "Press Enter" and a system shutdown. [10:59] pitti, So, this issue maybe specific to VirtualBox? [11:01] flexiondotorg: yeah, looks like it [11:01] flexiondotorg: I'll install VB here and cross-check (but need to sort out two other things first) [11:01] pitti, Sure :) [11:02] pitti, If you have anyhting the needs testing just ping me. [11:02] flexiondotorg: how's the actual install looking? [11:02] anything broken there still? [11:03] pitti, Install of Ubuntu MATE and Ubuntu were both fine. I also test oem-config for Ubuntu MATE. Also worked :) [11:04] Just need to build new images when the new ubiquity lands so the "Try" option works. [11:04] flexiondotorg: yay! [11:06] pitti, I've ask the OEM to test on there hardware range. [11:06] pitti, Will report back later. [11:13] jibel: so netboot still works as expected with the exception of the keyboard layouts and servers tty7 being blank [11:24] flexiondotorg: I confirm the problem in VB [11:28] flexiondotorg: and when I use a serial console to debug it it of course works [11:32] pitti, Typical. [11:33] pitti: serial console == debug == automatic fix for all woes ;) [11:35] ok, so systemd.debug-shell works [11:35] I see plymouth watch-keystroke running, but as we don't have /cdrom any more it's icky to debug [11:59] flexiondotorg: ok, so under VB plymouthd doesn't listen to /dev/tty7 and thus neither prints text nor receives keystrokes nor shows animation === chrisccoulson_ is now known as chrisccoulson [12:00] under qemu and real hw that works [12:00] pitti, OK, has a memory. [12:01] flexiondotorg: we fixed casper to actually show plymouth during shutdown, I guess that exposed that plymouth bug [12:01] pitti, The VirtualBox video device id is now blacklisted from grub. [12:02] pitti, I actually help test/implement that with the kernel team and cjwatson. [12:03] pitti, http://changelogs.ubuntu.com/changelogs/pool/main/g/grub-gfxpayload-lists/grub-gfxpayload-lists_0.7/changelog [12:03] pitti, Related? [12:04] pitti, If so then that same issue should be present with VMWare. [12:04] pitti: I'm good to rebuild for me now? [12:05] bluesabre, flexiondotorg did you figure out rebuilds this morning? [12:05] elfy: we landed a few other fixes; I think infinity is on it, but please coordinate with him [12:05] balloons, I rebuild Ubuntu MATE. It was published for all. [12:06] balloons, Which is what I expected. [12:06] also, while I have your attention, I was wondering if ubuntu mate was interested in doing a 'show and tell' session at UOS showcasing ubuntu mate? [12:07] balloons, Yes. [12:07] pitti: ack - it'll have to wait - only in for a short time [12:07] popey emailed me. [12:07] balloons, I'll get a response when 15.04 is out :) [12:08] :-) I assumed as much, just double checking [12:22] pitti, The OEM confirms that oem-config now works on all their hardware :) [13:04] flexiondotorg: some good news! [13:04] flexiondotorg: so we'll try to entirely disable plymouth on live system shutdown; this is too hairy to get right at this point [13:05] pitti: any idea why a server install defaults to tty7 but tty7 is a blank screen with just a flashing cursor? [13:06] davmor2: when exactly? [13:08] pitti: install server, I'm doing it from mini.iso. Instead of being in TTY with a login prompt I'm on a blank screen with a flashing white underscore and nothing else. If I switch to tty 1 - 6 I get the login prompt I expect [13:08] pitti: I'm trying to find the bug I filed with steps [13:09] pitti: as in do the install, reboot the system and you are stuck in tty7 as to a more precise timing :) [13:45] elopio, i asked for a review on my tz_tests branch, it seems to fix those flaky tests reliably, i kicked quite a few rebuilds without test failures [13:45] elopio, although some of the jobs failed, but they weren't test failures, infrastructure problems [13:45] elopio, the thing is, i don't really understand why my branch fixed it :) [13:45] elopio, which i don't like [13:53] kenvandine: thanks :) Just removing the call to the other test makes things more stable. But yes, I agree that probably it's not fixed, more likely it just happens less often. [13:53] lets see who's vanguard... [13:54] clearly much more reliable... but also i'm curious why this wasn't a problem until recently [13:54] brendand: can you make the review for ken's branch? or pass it to federico if you won't have time during your slot. [13:54] that call to the other test has been there for 15 months [13:55] it's clearly better not calling that test of course [13:55] https://code.launchpad.net/~ken-vandine/ubuntu-system-settings/tz_tests/+merge/256700 [13:56] kenvandine: yes, but so many things can be happening. As we discussed, the test is too big, so it's hard to analyze it. But I have seen funny things, like an async call that happens always in 9 seconds, after a change it takes 11 seconds 1/10 of the times. Just enough to cross the default timeout. [13:57] true [13:57] i just like understanding my fixes :) [13:57] makes sense though [14:57] jibel: ^ is this the final [14:57] davmor2, not yet [14:57] davmor2, more bugs coming [14:58] oh man === pgraner is now known as pgraner-afk [15:37] davmor2, there are 2 bugs currently, 1. session doesn't shutdown/restart on vbox 2. live session doesn't start on i386 [15:38] 1 is fixed and will be in next build, 2 is in progress [15:39] jibel: oh so only little things then nothing major [15:39] davmor2, nothing really ;) [15:44] davmor2, 2 cannot be fixed at this point but there is a workaround, I'll upgrade the release notes [15:44] update* [15:44] jibel: i386 just needs to die that should fix it right? [16:50] jibel: http://people.canonical.com/~pitti/tmp/gtk_ui.py === chihchun is now known as chihchun_afk [21:30] veebers: can we attach to a row in nfss a subunit stream file? [21:32] elopio: hmm, I'm not sure off the top of my head. You might be able to u64encode it perhaps as part of the json data. [21:32] elopio: I would need to take a look at the code to confirm [21:40] veebers: we are talking about putting in nfss information about each run. I'm wondering where to store the data collected during that run. [22:09] elopio: ah right, yeah an important question.