| wxl | infinity: anything else i can do to help? | 00:06 |
|---|---|---|
| infinity | wxl: Just be patient. Once this ubiquity makes its way through, I'll respin the world again. | 00:10 |
| wxl | infinity: ok, just checking. wish i could help out :) | 00:10 |
| * wxl sings a song about how heroic infinity is | 01:46 | |
| * bluesabre sings along | 01:47 | |
| infinity | wxl: Those alternates are likely still broken, I shouldn't have rebuilt them. :P | 01:47 |
| infinity | wxl: But desktop CDs should be popping out soon for everyone. | 01:47 |
| wxl | infinity: well, i have hope that something you do MIGHT make something better :) | 01:48 |
| * infinity has an intense need for pizza, wine, and naps. | 01:51 | |
| Mirv | repeating from yesterday a request:”archive admin consultance (preNEW accept) needed for Compiz that adds compiz-mate package, https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-021 - summary of packaging changes https://ci-train.ubuntu.com/job/ubuntu-landing-021-2-publish/32/artifact/packaging_changes_compiz_1%3A0.9.12.1+15.04.20150213-0ubuntu1.diff” | 05:43 |
| Mirv | ie what you'd do normally if it already was in NEW queue | 05:43 |
| wxl | infinity: when you arise from the dead, do you have any ideas on what can be done with ppc desktop? it seems that there's all sorts of dependency hell with xserver-xorg-* packages https://launchpadlibrarian.net/197941262/buildlog_ubuntu_trusty_powerpc_lubuntu_FAILEDTOBUILD.txt.gz | 06:27 |
| === oSoMoN_ is now known as oSoMoN | ||
| jamespage | urgh - please could someone reject golang-pty-dev - got the wrong source package name | 08:52 |
| Riddell | jamespage: rejected! | 09:17 |
| jamespage | Riddell, thanks | 09:29 |
| Laney | Could someone look at grilo-plugins in vivid/NEW to unblock the gnome-photos autopkgtest please? | 10:36 |
| flexiondotorg | Is there anyone in here who can review the Upload Queue? - https://launchpad.net/ubuntu/vivid/+queue | 10:49 |
| davmor2 | infinity: is netboot safe to test at this point? | 11:44 |
| davmor2 | infinity: and does netboot mini.iso hwe include the latest kernels etc | 11:45 |
| flexiondotorg | 😃 | 12:11 |
| rbasak | Could an AA take a look at bug 1417328 please? I think it's blocking proposed migration for MySQL 5.6. | 12:57 |
| ubot93 | bug 1417328 in percona-xtradb-cluster-5.5 (Ubuntu) "Please remove 5.5 versioned MySQL and variants from vivid" [Undecided,New] https://launchpad.net/bugs/1417328 | 12:57 |
| rbasak | I do need to get NEW percona-xtradb-cluster-5.6 uploaded though. Reviewing that now. | 12:58 |
| rbasak | I don't think it should block removal though? | 12:58 |
| rbasak | (of 5.5) | 12:58 |
| highvoltage | o/ | 14:12 |
| highvoltage | Edubuntu 14.04.2 images marked as ready | 14:13 |
| davmor2 | infinity: are you about? | 14:35 |
| === Guest49948 is now known as rcj | ||
| === rcj is now known as Guest20952 | ||
| ogra_ | ogra@anubis:~$ rmadison livecd-rootfs | 15:12 |
| ogra_ | curl: (22) The requested URL returned error: 503 Service Temporarily Unavailable | 15:12 |
| ogra_ | cjwatson, ^^^ is that known ? | 15:13 |
| cjwatson | It might just be unhappy because I'm doing a giant git repack | 15:13 |
| ogra_ | ah, k | 15:13 |
| * ogra_ will wait then | 15:13 | |
| cjwatson | Though the machine is otherwise responsive | 15:13 |
| Laney | Might be something else | 15:14 |
| cjwatson | Grr | 15:15 |
| Laney | I don't seem to be able to get to it from lillypilly | 15:15 |
| cjwatson | cjwatson@lillypilly:~$ GET http://archive-team.internal/ | 15:15 |
| cjwatson | Can't connect to archive-team.internal:80 (Connection refused) | 15:15 |
| cjwatson | Yeah | 15:15 |
| cjwatson | I bet they broke the firewall. Let me check | 15:15 |
| cjwatson | ogra_: escalated to webops | 15:18 |
| rsalveti | http://people.canonical.com/~ubuntu-archive/proposed-migration/ is also giving 503 | 15:18 |
| rsalveti | might be the same issue | 15:18 |
| cjwatson | rsalveti: yes, everything under http://people.canonical.com/~ubuntu-archive/ is broken by this | 15:20 |
| Riddell | bug 1423225 | 15:28 |
| ubot93 | bug 1423225 in base-files (Ubuntu) "lsb-release 14.04.2 still points to 14.04.1" [Undecided,New] https://launchpad.net/bugs/1423225 | 15:28 |
| Riddell | infinity: ↑ isn't that important to change ? | 15:28 |
| infinity | Riddell: Oh, sonofa... I didn't promote base-files to updates before the last respin. :( | 15:36 |
| * infinity wonders how he missed that. | 15:37 | |
| arges | infinity: hey. any merges you guys need help with? | 15:59 |
| davmor2 | infinity: hey dude, the mini.iso are still dated the 30/01 do they need an update at all or is mini.iso and netboot good to test? | 16:01 |
| infinity | davmor2: netboot should be fine. | 16:03 |
| jderose | infinity: so might be a regression introduced in fixing lp:1422864, but now linux-signed-generic-lts-utopic is marked as automatically installed, will be removed with "apt-get autoremove" | 16:03 |
| davmor2 | infinity: and is that the same for the HWE version? | 16:03 |
| infinity | jderose: Erk, really? | 16:03 |
| jderose | infinity: yeah, i just added a comment and screenshot - https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1422864 | 16:04 |
| ubot93 | Launchpad bug 1422864 in ubiquity (Ubuntu Trusty) "14.04.2: UEFI: linux 3.13.0-45 downloaded during install" [Undecided,Fix released] | 16:04 |
| jderose | infinity: does that seem related to you, or is this a separate issue? | 16:04 |
| infinity | jderose: I'm not sure how the change we made last night would have toggled the auto/manual state of that package. | 16:06 |
| infinity | jderose: But it's a bug nonetheless. Will investigate before I do the respin-for-base-files-screwup. | 16:06 |
| infinity | Which image are you testing? ubuntu? | 16:09 |
| jderose | infinity: yes, ubuntu desktop amd64 | 16:11 |
| infinity | jderose: Alright, grabbing it. | 16:13 |
| infinity | davmor2: Should be, yes. | 16:13 |
| davmor2 | infinity: cool | 16:13 |
| infinity | davmor2: d-i was rebuilt for the latest kernels, there shouldn't have been a reason to rebuild it again after. | 16:13 |
| davmor2 | infinity: no worries | 16:14 |
| davmor2 | jibel: ^ I guess that means it's safe to test did you add it to the tracker? | 16:14 |
| cjwatson | ogra_,rsalveti: webops have fixed it now | 16:15 |
| cjwatson | though it'll be very slow until I've finished this git archive maintenance | 16:16 |
| jderose | infinity: also, another problem i'm still trying to narrow down.... so in an oem install, i have a script that makes sure ubiquity-frontend-gtk, oem-config-gtk are installed, because they've sometimes been missing in daily ubuntu+1 ISOs. with a production ISO this script should have no effect.... | 16:17 |
| jderose | infinity: but but when i try to install the two, apt wants to remove grub-efi-amd64 grub-efi-amd64-signed | 16:18 |
| ogra_ | cjwatson, yep, confirmed, works fine (as expected ... after some minutes :) ) | 16:19 |
| infinity | jderose: There are no conflicts there, I can't see why that would be happening... | 16:19 |
| jderose | infinity: still trying to wrap my head around it, reproduce in a more isolated way. but for what it's worth, this script works fine when i do an oem install from 14.04.1 | 16:21 |
| rsalveti | cjwatson: thanks! | 16:24 |
| jderose | infinity: starting to narrow it down. at least when doing a uefi-mode oem install, oem-config-gtk isn't installed - https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1423254 | 16:45 |
| ubot93 | Launchpad bug 1423254 in ubiquity (Ubuntu) "14.04.2: oem-config-gtk not installed after doing an OEM install" [Undecided,New] | 16:45 |
| jderose | checking bios-mode now | 16:45 |
| jibel | infinity, ^ confirmed | 16:47 |
| jibel | cannot do an OEM installation | 16:47 |
| infinity | Lovely. :/ | 16:47 |
| jderose | jibel: so are you testing in bios mode then? | 16:47 |
| infinity | Does either of you happen to remember a time in the past when it did work correctly? | 16:47 |
| jibel | jderose, no, it was an UEFI installation | 16:47 |
| jibel | infinity, last attempt to release 14.04.2 | 16:48 |
| jderose | infinity: well, works with 14.04.1 :) | 16:48 |
| jibel | infinity, 3 weeks ago | 16:48 |
| jibel | or 2 | 16:48 |
| infinity | jibel: ... | 16:48 |
| jderose | ah yeah, worked with the daily from feb 3 | 16:48 |
| infinity | jibel: It worked 2 weeks ago? Not much changed. Grr. | 16:48 |
| jibel | infinity, usually it means that ubiquity and oem-config are out of sync | 16:48 |
| infinity | Well, "not much changed" should make it easier to lay blame, but I really don't like that the blame might be a 4-line ubiquity change that shouldn't relate. | 16:49 |
| jibel | infinity, ubniquity has be updated yesterday | 16:49 |
| jibel | well, today | 16:49 |
| jibel | I'm lost in time | 16:49 |
| infinity | jderose: Well, reproduced the linux-signed meta issue. Scratching my head about the why now. :P | 16:51 |
| infinity | jibel: Oh indeed, you're right, the oem-config in the pool is out of sync. Weird. That really shouldn't happen. | 16:52 |
| * infinity wonders if he has a stale lock on his mirror again. | 16:52 | |
| jderose | infinity: well, it's good i'm not crazy at least :) do you think this is related to the change yesterday? | 16:52 |
| infinity | jibel: So, that might magically fix itself when I abuse nusakan and do a rebuild for the base-files issue. | 16:52 |
| infinity | jderose: I think it could be. I really don't want it to be. :P | 16:53 |
| jibel | infinity, yeah, rebuild until if works ;) | 16:53 |
| jibel | it* | 16:53 |
| infinity | jibel: I was thinking something slightly more elegant like "figure out why my mirror is stale, then rebuild", but sure. | 16:53 |
| infinity | Also, wow, llvmpipe video is awful. Can I get something less crap in an OVMF qemu? | 16:54 |
| jderose | infinity: oem-config-gtk is likewise missing after doing a BIOS-mode OEM install | 17:01 |
| infinity | jderose: Yeah, same problem I'm sure. | 17:01 |
| infinity | And I think I've narrowed down what's happening with linux-signed. Ish. | 17:01 |
| jderose | although when you try to manually install them after the fact in BIOS mode, apt doesn't want to remove grub-efi, so at least it's a slightly better situation | 17:02 |
| infinity | linux-signed-image-generic-lts-utopic *should* be marked auto (and is), but linux-signed-generic-lts-utopic should be installed and manual, and it's... Not there. | 17:02 |
| infinity | Now to hunt down why it's not there. :( | 17:02 |
| infinity | Oh. Balls. It is the change yesterday. | 17:03 |
| * infinity does a fresh install with some live patching. | 17:08 | |
| infinity | How many years ago did "-net user" become the default in qemu? I just realised I forgot to specify it on the CLI, but I have a network. And now I suspect I've been pointlessly typing it for half a decade. :P | 17:17 |
| mlankhorst | no idea, I've had to use -net nic,vlan=0 -net tap,vlan=0 at least since precise :-) | 17:20 |
| infinity | mlankhorst: Yeah, I usually use tun/tap bridge setups too, which might be why I never noticed that user "Just Works" without asking for it. | 17:22 |
| mlankhorst | mostly for netboot | 17:23 |
| infinity | cyphermox: You around? | 17:29 |
| cyphermox | yes | 17:29 |
| cyphermox | just finished my soup | 17:29 |
| infinity | cyphermox: http://paste.ubuntu.com/10293752/ <-- Spot review for me. | 17:29 |
| infinity | cyphermox: Tested by live-patching a VM, and it DTRT. | 17:29 |
| cyphermox | there was an issue with $kernels after all? | 17:30 |
| infinity | cyphermox: (Our first attempt only kept linux-signed-image-$blah, not linux-signed-$blah, which cascaded into a bunch of autoremovals post-install) | 17:30 |
| cyphermox | ok | 17:30 |
| infinity | cyphermox: $kernels only contains *image*... Oops. | 17:30 |
| infinity | cyphermox: And rather than "fix" that and possibly create another bug, better to just use another list, IMO. | 17:30 |
| cyphermox | yes | 17:30 |
| infinity | cyphermox: At least, for a quick fix today. | 17:30 |
| cyphermox | sounds good | 17:31 |
| cyphermox | looks good | 17:31 |
| infinity | jderose: Nice catch, BTW. Not sure most people would have noticed until the first ABI bump when they couldn't boot anymore. :P | 17:31 |
| cyphermox | heh :0 | 17:31 |
| jderose | infinity: only caught it because of some of the automated tooling i use at system76 :) | 17:32 |
| infinity | jderose: Huzzah for your scripts hating on me? | 17:33 |
| jderose | hehe | 17:33 |
| infinity | Self-accepting that upload based on cyphermox's review. | 17:35 |
| cyphermox | anything I should look at? | 17:38 |
| infinity | cyphermox: Check iso.qa for any interesting new bugs filed that seem like something we could fix in an hour or three? | 17:38 |
| cyphermox | sure | 17:38 |
| infinity | cyphermox: But I think we're probably about as good as we're going to get once I fix my mirror skew and rebuild with the new ubiquity and base-files. | 17:39 |
| jderose | infinity: cyphermox: is it just me, or is the timezone detection broken? i'm in us/mountain, and both desktop and server installs are defaulting to us/eastern | 17:43 |
| infinity | jderose: Mine correctly picked Edmonton. | 17:46 |
| infinity | jderose: (which is mountain, but of the Canadian type) | 17:46 |
| infinity | jderose: Probably has more to do with your ISP. It's not an exact science. | 17:46 |
| jderose | infinity: it doesn't seem like it's even trying though... in the desktop install, i'm used to a delay while it does the lookup (when there in internet access). | 17:47 |
| infinity | jderose: I don't remember if I saw a noticable delay (really, it should be quick anyway), but it definitely got it right here, so... | 17:48 |
| infinity | jderose: You're not preseeding, I assume? | 17:49 |
| jderose | no | 17:49 |
| jderose | hmm, i wonder if the lookup is being skipped on oem installs, but not in normal installs... double checking some things | 17:49 |
| infinity | jderose: I thought OEM install was broken for you anyway? Or did you hack around it to keep testing? | 17:50 |
| jderose | hacked around it. the install works okay, but oem-config-gtk isn't installed. you can manually install it after the fact | 17:51 |
| jderose | well, except for the uefi-mode issue. | 17:51 |
| jderose | infinity: never mind, must be something new and magical with my ISP... i'm getting us/eastern when i do a desktop 14.10 install also | 17:55 |
| infinity | jderose: Yeah. Large ISPs shuffle subnets around every once in a while, and the magic geoip databases don't always keep up. | 17:55 |
| infinity | jderose: I've lived in pretty much every timezone in Canada over the last few years without ever moving. :P | 17:56 |
| jderose | hehe :) | 17:56 |
| infinity | In theory, IPv6 will fix all of this. In theory. | 17:56 |
| infinity | Except that most people get their v6 via tunnels right now, and completely break the geographical routing assumptions. | 17:57 |
| cyphermox | jderose: you could check https://freegeoip.net/ to find out if that's it, but it most likely is the reason | 17:59 |
| infinity | cyphermox: ... where I discover that I've been proxying firefox through my colo in San Jose for god knows how long and forgot to turn that off. | 18:00 |
| cyphermox | fwiw I live in a reasonably sized city where the ISP usually does the right thing, but with dynamic IPs, they can mess up geoip databases pretty quickly | 18:01 |
| infinity | cyphermox: Also, creepy. Looking up my real IP actually drills down to the right postal code. I don't know how I feel about that. | 18:02 |
| cyphermox | infinity: meet Big Brother. | 18:02 |
| Riddell | infinity: will you respin for the lsb-release update? | 18:02 |
| jderose | hmm, pretty amazing it gets it right down to that sort of granularity. amazing/creepy, take your pick i guess :) | 18:03 |
| infinity | Riddell: Once this ubiquity winds its way through. So, Soon(tm). | 18:03 |
| cyphermox | infinity: ou could sit at home, and do like absolutely nothing, and your name goes through like 17 computers a day. 1984? Yeah right, man. That's a typo. Orwell is here now. He's livin' large. We have no names, man. No names. We are nameless! | 18:03 |
| infinity | Riddell: Did you have one more thing you wanted to squeeze in, or just curious when you'll see new ISOs? | 18:03 |
| infinity | cyphermox: Great, now I need to watch Hackers in the background while I work. | 18:03 |
| cyphermox | sorry, so do I | 18:04 |
| Riddell | infinity: the latter, just needing to know if I should poke my testers now or wait | 18:04 |
| Riddell | I'll wait :) | 18:04 |
| infinity | Riddell: arm64/armhf binaries for ubiquity need to publish, then a copy-to-updates, another publisher cycle, then buildy buildy. | 18:04 |
| infinity | Riddell: So, triggering in probably about an hour, images as they spit out after that. | 18:04 |
| cyphermox | jderose: take into consideration it could pinpoint your location to the meter, via lat/lon coordinates. Databases don't usually include that precise information, but it could ;) | 18:05 |
| infinity | cyphermox: I love the "17 computers a day" bit, though. That number seemed low even then, and laughably tiny now. | 18:05 |
| cyphermox | indeed. | 18:06 |
| cyphermox | infinity: fwiw I don't see bugs-i-can-fix-in-one-hour so much in iso.qa, unless it was going to be yet another ubiquity upload for bug 1418105 ;) | 18:07 |
| ubot93 | bug 1418105 in ubiquity (Ubuntu) "Trusty: testing .2 the oem session on i386 leaves the oem wifi password in place" [High,In progress] https://launchpad.net/bugs/1418105 | 18:07 |
| cyphermox | .. which isn't even on iso.qa? | 18:07 |
| cyphermox | and not yet fixed in vivid, either, so yeah | 18:07 |
| infinity | cyphermox: That's not a regression, right? | 18:08 |
| infinity | cyphermox: If not, then target for .3 is much more reasonable than trying to squeeze it in. | 18:08 |
| cyphermox | infinity: no, that was already in 14.04.1 too | 18:08 |
| cyphermox | yeah, I know | 18:08 |
| * infinity targets it. | 18:08 | |
| infinity | cyphermox: Fun bug, though. Reminds me of how Every Single Time we get a machine sent to us from IBM, it's configured to run on their internal network. :P | 18:09 |
| cyphermox | my evil plan has been discovered. | 18:10 |
| infinity | Because, apparently, DHCP is hard. | 18:10 |
| cyphermox | well, it's really a lot more of an issue for wifi, assuming people would do OEM installs on a wifi network and all | 18:11 |
| cyphermox | most likely one by one, because preseeding that would be ugly. | 18:11 |
| infinity | cyphermox: Sure, yeah, cause the password is actual sensitive data. But conceptually similar, that's all. :) | 18:11 |
| cyphermox | yes | 18:11 |
| infinity | cyphermox: Though, I think it would be hilarious if driving by System76 with my laptop gave me free Internets. | 18:12 |
| infinity | jderose: Can we consider this a feature instead of a bug? :) | 18:12 |
| cyphermox | infinity: ask jderose, he could hook you up | 18:12 |
| cyphermox | ;D | 18:12 |
| jderose | infinity: cyphermox: it doesn't. now we do all our image mastering with qemu, never have a wifi password in the golden image | 18:13 |
| infinity | jderose: Oh, I didn't assume this bug actually affected you. :P | 18:13 |
| jderose | although i do know that prior to me working at system76... free wifi was included. definitely a feature :) | 18:13 |
| infinity | jderose: You're just the only OEM in the channel right now to bug. :P | 18:13 |
| cyphermox | you'd see house prices right around that building rise up, and mostly taken by Ubuntu developers, that's break demographic statistics badly :) | 18:13 |
| infinity | cyphermox: I dunno, what self-respecting developer would be satisfied sharing WiFi with a bunch of other people? | 18:14 |
| infinity | cyphermox: I consider my 100Mbit slow. :( | 18:14 |
| * ogra_ glares at infinity | 18:15 | |
| infinity | I need to go back in time to when I thought 1.5 SDSL was effin' INCREDIBLE, and just live there forever. | 18:15 |
| cyphermox | 100MBit? | 18:15 |
| * ogra_ still lives there | 18:15 | |
| ogra_ | well ... 2MBit SDSL | 18:15 |
| infinity | cyphermox: Yeah. Well, only 10 up, which is part of the complaint. Cause who downloads stuff, really? | 18:15 |
| infinity | ogra_: At least it's S and not A. I imagine 2M down would ba 256k up or something in A land. | 18:16 |
| infinity | Or worse. | 18:16 |
| ogra_ | yeah | 18:16 |
| ogra_ | thats why i have it :) | 18:16 |
| infinity | ogra_: Also, assuming that's SDSL-over-repurposed-ISDN-infra, isn't the EU standard 2.3Mbit, not 2? | 18:17 |
| infinity | Or are you rounding down? :P | 18:17 |
| ogra_ | nope, its 2MBit :) | 18:17 |
| infinity | Sucks to be wherever you are... | 18:18 |
| ogra_ | they probably swallow the overhead somehow | 18:18 |
| ogra_ | i could get 100M cable ... *if* there was cable in m street ... which my neighbors voted against | 18:18 |
| ogra_ | s/m/my/ | 18:19 |
| infinity | ogra_: Get it two streets over and set up a wireless repeater? | 18:19 |
| ogra_ | haha | 18:19 |
| jderose | infinity: was the problem with base-files just that /etc/lsb-release hadn't been bumped to 14.04.2? | 18:19 |
| infinity | ogra_: That's more or less how lamont used to do things before they finally ran cable to his place. | 18:19 |
| ogra_ | well, I#m pondering getting an additional sat link | 18:19 |
| infinity | jderose: The problem was that I bumped it ages ago and forgot to promote it from -proposed because I'm a derp. | 18:19 |
| ogra_ | for downloads that should be fine ... 60M or some such for ~50€ | 18:20 |
| jderose | infinity: no, not a derp. likely just too busy with other things :) | 18:20 |
| infinity | jderose: I'm sticking with the derp explanation. | 18:20 |
| jderose | fair enough, that's your right :P | 18:20 |
| * ogra_ goes back to scream and shout about nodejs and npm | 18:21 | |
| ogra_ | i'm inclined to say i havent seen something as bad in a long time ... | 18:22 |
| ogra_ | but i think thats wrong ... i havent seen something that bad *evar* ! | 18:23 |
| infinity | ogra_: Whatever you're doing with node, RUN AWAY. | 18:25 |
| ogra_ | hahaha | 18:25 |
| ogra_ | to late | 18:25 |
| infinity | It's never too late. You're in an abusive relationship with nodejs, there's help available. | 18:25 |
| ogra_ | well, created that "hangouts server" for snappy ... webrtc based using socket.io ... so it need node ... up to now i just used a complete node sdk build that i unpackked in the snap before building it ... | 18:28 |
| ogra_ | now i'm trying to get x86 working too and try to pull the node install from the archive ... | 18:29 |
| ogra_ | (which makes it 10:1 smaller ... but is a pain to do in qemu arm chroots) | 18:29 |
| infinity | So use real hardware? | 18:30 |
| infinity | Surely, you have some lying around from a past life. :P | 18:30 |
| ogra_ | heh, i even have a ton of new one (yay snappy) ... | 18:32 |
| ogra_ | but i want a build script that runs on one machine | 18:32 |
| infinity | I think relying on qemu-user-static for this is likely to end in tears. | 18:32 |
| infinity | There are any number of things it just will never do quite right. | 18:32 |
| ogra_ | my target is to have a proper script for people wanting to package node projects | 18:33 |
| infinity | And it'll fail even more miserably the day someone asks you to support powerpc. | 18:33 |
| ogra_ | so they can just call it on their PC and get a snap with the necessary binaries in place | 18:33 |
| ogra_ | the point is that i use binaries from debs ... i *theoretically* wouldnt need to even execute anything foreign | 18:34 |
| infinity | ogra_: As long as the debs don't have postinsts that are necessary, yeah. | 18:34 |
| ogra_ | *if* there wasnt that crappy npm (node package manager) that essentially just downloads crap from github to put it in some module dir | 18:34 |
| ogra_ | right, but even then i could re-implement my own postinst if needed | 18:35 |
| infinity | "Downloading crap from github" is the new black. | 18:35 |
| ogra_ | yeah | 18:35 |
| infinity | It amazes me that people don't see all the flaws in this. | 18:35 |
| ogra_ | thats whyy we have sappy ... just ship the flaws and dont care how broken it is | 18:36 |
| infinity | But I guess our industry is all about making the same mistakes, learning the same lessons, and solving the same problems, every half generation or so. | 18:36 |
| jderose | infinity: i feel the same way. call me old fashioned, but i really like signed pre-built packages. and shared libraries, dependencies :) | 18:36 |
| infinity | jderose: You're old-fashioned. | 18:37 |
| infinity | jderose: And also 100% correct. | 18:37 |
| infinity | jderose: And I have no bias whatsoever. | 18:37 |
| infinity | *cough* | 18:37 |
| jderose | hehe | 18:37 |
| * ogra_ loves the "a package is a project" approach you can go with snappy ... | 18:37 | |
| ogra_ | it fulfills a different purpose indeed | 18:38 |
| infinity | ogra_: I don't at all discount the need for the "app store" model to let people do crazy things, and to let upstreams and end users get on with things without me getting in their way. | 18:39 |
| infinity | ogra_: I was specifically being grumpy about how the "github culture" distributes software. | 18:39 |
| ogra_ | yeah | 18:40 |
| infinity | Blows my mind that it's become both common and *accepted* to say "wget this ruby script and pipe it to sudo, what could possibly go wrong?" | 18:40 |
| infinity | I kinda want to blame Apple. Just cause. | 18:40 |
| ogra_ | haha | 18:40 |
| infinity | The OSX generation of UNIX developers seem to leave much to be desired when it comes to learning all the same things we learned 15 years ago. | 18:41 |
| ogra_ | dunno, can you blame apple for misbehaving opensource devs ? | 18:41 |
| jderose | infinity: i'm noticed a very strong correlation between using osx and being a fan of this approach :) | 18:41 |
| infinity | ogra_: No, I can blame them for switching to UNIX (which was the right thing to do, but argh) | 18:41 |
| * ogra_ would expect them rather in the linux camp | 18:41 | |
| ogra_ | for switching UNIX you mean :) | 18:41 |
| infinity | That too. | 18:42 |
| wxl | infinity: speaking of apple, any luck on lubuntu desktop ppc? :) | 18:42 |
| ogra_ | lol | 18:42 |
| wxl | ftw XD | 18:42 |
| ogra_ | stop being on topic ! | 18:42 |
| infinity | wxl: Oh, err. No. | 18:42 |
| ogra_ | :) | 18:42 |
| wxl | i know, so lame | 18:42 |
| infinity | wxl: We might just have to skip that image this round, and PPC users can install with 14.04.1 | 18:42 |
| infinity | wxl: I think we're running out of time to unwind what went wrong. | 18:43 |
| wxl | infinity: well the log kind of suggested the same problems as with the others but *shrug* | 18:43 |
| infinity | wxl: To be fair, PPC users also have about 0 reason to want/need the HWE stack. | 18:43 |
| infinity | wxl: No, the log suggests entirely new and exciting problems. :P | 18:43 |
| wxl | hm | 18:43 |
| wxl | not what i saw | 18:43 |
| infinity | wxl: Well, none of the other failed to build, they just got weird things installed. | 18:43 |
| infinity | wxl: ppc is actually just refusing to install what we ask it. I'd need to spend some quality time in a PPC chroot and sort out what mlankhorst broke. :P | 18:44 |
| * infinity fires up a PowerStation in the corner to look. | 18:44 | |
| infinity | wxl: No promises on fixing it, but I'll look. | 18:45 |
| infinity | wxl: But, like I said, I don't think it's world-ending either. The HWE stack is for people who have Really New Hardware, which is the exact opposite of your PPC users. | 18:45 |
| wxl | infinity: yeah, but if we can fix it, it would be nice | 18:45 |
| wxl | infinity: looking at the END of https://launchpadlibrarian.net/197941262/buildlog_ubuntu_trusty_powerpc_lubuntu_FAILEDTOBUILD.txt.gz it's apt that's unhappy | 18:45 |
| infinity | wxl: Yes, I've read the log. | 18:46 |
| infinity | wxl: Just need to drill down a lot deeper to find out why. | 18:46 |
| infinity | And it's probably that the X metapackages are broken on PPC. | 18:46 |
| infinity | But fixing them means respinning everyone, so as we get closer to the wire, I'm more inclined to say "oh well, next time". | 18:47 |
| wxl | meh | 18:47 |
| wxl | well you're the boss :) | 18:47 |
| infinity | wxl: I wouldn't leave one of your x86 images high and dry like that, but yeah. PPC users probably shouldn't want the HWE stack anyway. It solves nothing for them, except pretty version number bumps. | 18:48 |
| infinity | Anyhow, I'll look now. If it's a quick fix, I'll slam it in. | 18:48 |
| wxl | that's what i'm thinking it might be | 18:49 |
| wxl | and if's not, then yeah, oh well :) | 18:49 |
| infinity | wxl: Thankfully, we can just keep publishing the 14.04.1 PPC ISO anyway, and other than the weird version mismatch, no "normal end user" will really notice the difference. :P | 18:50 |
| infinity | Install, update, it says "14.04.2" when they log in, just old kernel and X, no big deal. | 18:50 |
| wxl | yep, sounds good | 18:50 |
| * infinity needs to get some faster SAS disks for his PPC machine... | 18:53 | |
| infinity | Or more RAM and a massive tmpfs. | 18:53 |
| wxl | infinity: what ram you need? | 18:55 |
| wxl | if you tell me what hardware you need for that ppc machine, i'll figure out a way to get it to you | 18:55 |
| infinity | wxl: Hah. Not sure, TBH. I'd have to go digging a bit. | 18:55 |
| wxl | infinity: well, let me know | 18:56 |
| wxl | brb | 18:56 |
| infinity | wxl: But being old server class hardware, it's probably something annoyingly expensive, like the disks. :P | 18:56 |
| infinity | I'm really just waiting for more OpenPower kit to become available. | 18:56 |
| wxl | infinity: oh you have a server, huh. maybe i could ship you my g5 :) | 18:56 |
| wxl | you and randall ross, eh, infinity ? | 18:56 |
| infinity | wxl: It's essentially the same as your G5, but with some IBM oddities. | 18:56 |
| wxl | brb | 18:57 |
| infinity | wxl: Okay, throwing a quick test at my PPA. If this builds for all arches, I'll let it in, but I make no promises that it will boot, install, or not suck. | 19:03 |
| wxl | infinity: sounds good, thanks | 19:03 |
| * infinity taps his foot waiting for LP to publish that so he can kick off quick tests. | 19:09 | |
| infinity | rsalveti: Are you doing manual builds on nusakan right now? | 19:15 |
| infinity | Okay, rebuilding everything but lubuntu for the ubiquity and base-files bugs. Will respin lubuntu after I've tested this livecd-rootfs hack for PPC. | 19:28 |
| infinity | Riddell: Images should start vomiting out shortly. | 19:28 |
| * infinity goes to hunt lunch. | 19:29 | |
| wxl | thanks infinity | 19:29 |
| infinity | wxl: Okay, got it building. Once this livecd-rootfs makes its way through LP, I'll rebuild lubuntu. | 19:56 |
| infinity | wxl: Like I said, I'm making no promises about the PPC ISOs being usable, this is our one chance. If they suck, oh well. | 19:56 |
| infinity | wxl: But the manifest diffs look sane, so they might be alright. | 19:56 |
| infinity | Riddell: Lucky you, your ISOs were first out ^ | 19:57 |
| elfy | alphabetical? going to propose AXubuntu :p | 20:02 |
| infinity | elfy: No, entirely random (ish). | 20:02 |
| elfy | :) | 20:03 |
| infinity | elfy: We throw all the livefs builds at the buildds and the first ones to return get turned into ISOs. | 20:03 |
| infinity | elfy: So, in theory, the smaller livefses would usually win. | 20:03 |
| elfy | right | 20:03 |
| infinity | elfy: (And that would always be server, except that it takes a while to get the powerpc livefs back across the ocean...) | 20:03 |
| infinity | elfy: Really, though, there's a certain element of "whee, computers! parallelism is hard!" to it that makes it random enough to be fun betting on. | 20:05 |
| elfy | ha ha | 20:05 |
| jderose | infinity: 20150218.1 desktop amd64: uefi mode install seems correct; also no problems with OEM install (now to test customer setup bit) | 20:32 |
| jderose | infinity: oem-config customer setup bit seems solid too | 20:34 |
| infinity | jderose: \o/ | 20:50 |
| infinity | jderose: I love it when a plan comes together. | 20:50 |
| jderose | infinity: thanks for all the help and quick fixes! | 20:51 |
| infinity | jderose: Are you logging your testing on the ISO tracker? I see 0/9 for all ubuntu desktop tests, always nice to see more than just jibel filling those in. ;) | 20:53 |
| jderose | infinity: no, i haven't been. for some reason ubuntu sso wont let me signinto the tracker, i think because i used the same user name before tracker was integrated with sso | 20:54 |
| infinity | jderose: Oh, special. I thought all instances of that had been fixed. | 20:54 |
| infinity | stgraber: ^-- You know what's up with that? (if you're not too busy at Collab) | 20:54 |
| infinity | wxl: Your builds are on their way now. | 21:03 |
| wxl | thx infinity | 21:03 |
| wxl | um desktop ppc's not there infinity https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/trusty/lubuntu | 21:27 |
| wxl | i mean i see the tracker thinks it's rebuilding | 21:28 |
| infinity | Bah, maybe it failed differently. Looking. | 21:34 |
| infinity | Or the build never happened... | 21:34 |
| wxl | that's what i'm thinking infinity | 21:35 |
| wxl | i've never seen it not show up there if it's in process | 21:35 |
| infinity | I think the tracker rebuild thingee lost its mind. Let me force the issue. | 21:35 |
| infinity | wxl: Doing a manual build to restore sanity. | 21:37 |
| wxl | thank you again, infinity | 21:37 |
| davmor2 | infinity: netboot on 32,64,32hwe,64hwe done | 21:37 |
| infinity | wxl: It might end up renumbering amd64 and i386 to .2, but they'll be the same build as before. | 21:37 |
| infinity | davmor2: Ta. | 21:37 |
| infinity | davmor2: Any idea who's taking responsibility for testing server ISOs? | 21:38 |
| davmor2 | infinity: server team normally | 21:38 |
| infinity | davmor2: I can do ppc and ppc64el, since no one else seems to, but we usually have a ton of Canonical QA folks doing x86. | 21:38 |
| davmor2 | infinity: yes I'm one of them ;) I was getting the netboot out the way while iso's were respinning | 21:39 |
| infinity | highvoltage: Sorry to invalidate all your testing, but there's a new edubuntu for you to spot check and smoketest. | 21:40 |
| infinity | davmor2: Check. | 21:40 |
| davmor2 | infinity: tomorrow I think the plan is that jibel will hit i386 and amd64 and I'll do a quick run on mac, I think server team have their own qa who auto/manually test server at least that is what has happened in the past jibel might know more though | 21:42 |
| wxl | i see it building, so thanks infinity | 21:42 |
| infinity | davmor2: I'd better test ppc/ppc64el tonight, so I know if the world's generally broken. | 21:43 |
| infinity | davmor2: But I don't really anticipate issues, the HWE stuff is both (a) trivial for server compared to desktop and (b) been included in dailies for months without much complaint. | 21:44 |
| infinity | superm1: Is anyone testing mythbuntu point release ISOs? | 21:46 |
| davmor2 | infinity: famous last words ;) QA Motto if something can go wrong it will and will take out the universe with it :D | 21:46 |
| infinity | davmor2: Don't jinx it. | 21:46 |
| infinity | zequence: Ditto for you, is anyone testing studio point release ISOs? You're running out of time. | 21:47 |
| infinity | davmor2: What's the public/community QA channel again? Either I parted it, or I can't remember what it's called. :P | 21:47 |
| elfy | #ubuntu-quality | 21:48 |
| elfy | that the one you mean? | 21:48 |
| davmor2 | infinity: #ubuntu-quality | 21:48 |
| davmor2 | meh elfy beat me to it :) | 21:48 |
| elfy | :) | 21:48 |
| rsalveti | infinity: I was building for touch | 21:55 |
| elfy | infinity: what sort of time UTC are you expecting to release tomorrow for trusty? | 21:55 |
| wxl | you're utc-5 infinity ? | 21:55 |
| elfy | not working - so I'll be able to catch my own stragglers and could do other's empties for an hour or so today | 21:55 |
| infinity | rsalveti: Yeah, I figured. Stellar response time, though! ;) | 21:55 |
| rsalveti | infinity: sorry, started the image and went to get some food :-) | 21:56 |
| infinity | wxl: I'm -0700 right now. | 21:56 |
| wxl | infinity: oh, you're close to my time, good. :) | 21:56 |
| infinity | elfy: I haven't picked a time, and won't until I get more results in. We could always be respinning again if something Very Bad is found. :/ | 21:56 |
| wxl | so my guess is something like 1200-0700 | 21:57 |
| elfy | infinity: yep understand that - but it's not going to be something daft like 10:00UTC is what I'm after here :) | 21:57 |
| wxl | elfy: hahah yeah it's not going to follow the ubuntu on air schedule XD | 21:57 |
| infinity | elfy: Oh, no. It'll be much later than that. | 21:59 |
| elfy | infinity: ok - cool, I can clean up my own stuff tomorrow then - have a bit of time now - can smoketest studio's installs at least | 22:00 |
| elfy | and I mailed the studio list too | 22:00 |
| infinity | elfy: You're a champion. | 22:00 |
| elfy | infinity: I could actually mark the studio stuff for release, in the -release team to get access on trackers to help them set up package stuff | 22:02 |
| elfy | not that I actually would | 22:02 |
| infinity | elfy: If $people test it, I'll happily mark it ready myself. I'm really just looking for either user-harmful showstoppers (you tried to install it and your computer started speaking in tongues before throwing a fireball at your cat) or usability showstoppers (the xfce desktop was replaced by fvwm95, and no one's sure how). | 22:03 |
| infinity | elfy: If it all looks "good enough", I'm not terribly picky. | 22:04 |
| elfy | :D | 22:04 |
| elfy | not looking forward to beta vivid - we got a vbox issue on 32 bit that's going to have people failing it :) | 22:05 |
| * wxl waits for the successfully built desktop ppc image to show up on the tracker already· | 22:08 | |
| infinity | elfy: Tell the virtualbox testers to use kvm instead? :P | 22:09 |
| infinity | wxl: It's still thinking hard about it over here. | 22:10 |
| elfy | infinity: I looked at that - too much to think about for this old man | 22:10 |
| infinity | elfy: I'm sure there are pretty GUI frontends to KVM too... | 22:11 |
| infinity | Not that I use them, but I'm sure they exist. | 22:11 |
| elfy | probably - and probably not as 'hard' as it appeared | 22:11 |
| elfy | but if they didn't test with it - we'd not have found a bunch of stuff last cycle :D | 22:11 |
| infinity | elfy: Usually, what we find with virtualbox is that virtualbox is crap. | 22:12 |
| elfy | yea | 22:12 |
| infinity | elfy: Which is an interesting test, but less relevant when 99% of our expected users are either installing on real hardware or in cloud instances, which are invariably every hypervisor in the world EXCEPT virtualbox. | 22:13 |
| cjwatson | infinity: hey, fvwm95 was pretty good. | 22:13 |
| infinity | (kvm, xen, hyper-v, vmware...) | 22:13 |
| cjwatson | At least, you know, in 1998. | 22:13 |
| infinity | cjwatson: Wash the pink paint off your shades. | 22:13 |
| elfy | unfortunately people use it - and till I got rid of testdrive references in the current wiki - QA recommended vbox sort off :) | 22:13 |
| infinity | cjwatson: Also, windowmaker 4 lyf, yo. *gang sign* | 22:13 |
| elfy | s/off/of | 22:13 |
| cjwatson | I tried using windowmaker in 1999 and never understood it | 22:14 |
| stgraber | jderose: hey | 22:14 |
| infinity | cjwatson: Understanding it wasn't important, it was undocking apps and watching the icon explode. You could do that ALL DAY LONG and never tire of it. | 22:14 |
| elfy | infinity: understand your point - but only Ubuntu will be worrying about cloud :) | 22:14 |
| jderose | stgraber: hey :) | 22:15 |
| infinity | elfy: Sure, but most desktop people don't worry about any virt (err, except I guess Mac people and whatever thing they insist on running Linux under...) | 22:15 |
| elfy | :) | 22:15 |
| stgraber | jderose: ok, so let me take a look at the tracker logs real quick | 22:17 |
| wxl | OMG PPC JUST BUILT | 22:17 |
| elfy | cjwatson: have you got a really quick minute in PM? | 22:17 |
| wxl | elfy: infinity: isn't virt-manager the GUI for kvm? | 22:17 |
| infinity | wxl: Dude, that was a whole 6 minutes ago. | 22:17 |
| wxl | infinity: i was involved in other conversation :) | 22:18 |
| elfy | he's been refreshing the tracker infinity :) | 22:18 |
| cjwatson | elfy: sure | 22:18 |
| wxl | and i like virtualbox honestly | 22:18 |
| cjwatson | though I'm mostly a spectator here now | 22:18 |
| infinity | wxl: And sure, if you say so. I run qemu from the commandline because I hate learning anything that involves a mouse and isn't a video game. | 22:18 |
| wxl | of course i use the one from oracle's repos | 22:18 |
| wxl | infinity: word, homie. | 22:18 |
| stgraber | jderose: can you login now so I get a new log record to investigate? | 22:20 |
| wxl | is there anywhere that the upload to cdimages are tracked (publicly)? | 22:21 |
| jderose | stgraber: done: The name jderose is already taken. | 22:21 |
| infinity | wxl: I'm not sure what you mean by "upload to cdimages". | 22:22 |
| wxl | infinity: once the build is done, it uploads to cdimage, right? | 22:22 |
| infinity | wxl: Well, it's done on a machine we call "cdimage", and published to its disk. | 22:22 |
| infinity | wxl: Then mirrored via rsync to all the frontends you know as cdimage.ubuntu.com | 22:22 |
| wxl | so this is not made public at all, infinity ? | 22:23 |
| infinity | wxl: But there's nothing being uploaded to cdimage, per se. | 22:23 |
| infinity | wxl: I'm not sure what you're asking for. :P | 22:23 |
| infinity | wxl: The rsync logs to the frontends? No, no reason for every mirror to make their logs public. | 22:23 |
| infinity | wxl: The cd image build logs, those are public. | 22:23 |
| infinity | wxl: http://people.canonical.com/~ubuntu-archive/cd-build-logs/ | 22:23 |
| cjwatson | although only synced hourly or so | 22:24 |
| wxl | oh that's good to know cjwatson | 22:24 |
| cjwatson | oh, every 15 minutes, not so bad | 22:25 |
| stgraber | jderose: try now | 22:29 |
| jderose | stgraber: it worked, thanks! | 22:30 |
| stgraber | good | 22:31 |
| elfy | oops installer crashed in studio | 22:31 |
| elfy | knew I should have just gone to bed | 22:31 |
| infinity | ... | 22:31 |
| stgraber | not sure what was going on, I just checked if there was any content linked to the old account and seeing how there was none, I removed it :) | 22:31 |
| elfy | mmm | 22:32 |
| infinity | elfy: Like, python exploded, or whole desktop went poof? | 22:32 |
| elfy | won't let me report it either ... | 22:32 |
| elfy | crashed mid install | 22:32 |
| infinity | But just ubiquity, is what I was asking? | 22:32 |
| elfy | trying again | 22:32 |
| elfy | I guess - I'd set it all up and it was in the middle of the actual install | 22:33 |
| elfy | wouldn't let me report it - 3rd party package warning from apport | 22:33 |
| infinity | Oh, apport, you scamp. | 22:34 |
| elfy | :p | 22:34 |
| elfy | redoing it | 22:34 |
| elfy | 32bit install passed, the livesession tests passed as much as I could | 22:34 |
| ianorlin | Hi apport from running the installer at the menu doesn't seem to working on a package during the install I think the install completed but is saying the packages on the release are not offical package on the installation media | 22:38 |
| wxl | do you have any idea what you might have done to cause that issue, infinity ? :) | 22:39 |
| wxl | ianorlin: that's desktop amd64 or i386? | 22:39 |
| ianorlin | desktop i386 | 22:39 |
| ianorlin | I orignally just wanted to report a papercut but apport being broken in install seems to be worse bug | 22:40 |
| infinity | I have no idea why apport feels the urge to even perform that check, nor why it's half broken. :/ | 22:40 |
| elfy | but funny I just had the same | 22:40 |
| infinity | bdmurray: The above is all yours. Why does apport suck at life? | 22:43 |
| elfy | infinity: mmm - so it crashed again - apport gave up again - best I can do is show what it was trying to install when it crashed | 22:43 |
| elfy | http://i.imgur.com/5C75Q1i.png | 22:43 |
| infinity | elfy: The file from /var/crash could be handy. | 22:44 |
| elfy | I'll run it again ... | 22:44 |
| elfy | ;) | 22:44 |
| * infinity downloads the image too.. | 22:44 | |
| infinity | elfy: amd64? | 22:44 |
| bdmurray | ianorlin: which desktop image? | 22:44 |
| elfy | yep | 22:44 |
| elfy | checked the md5sum - that was good | 22:44 |
| bdmurray | have you run apt-get update yet? | 22:45 |
| elfy | bdmurray: I get the same apport issue from the studio 64 bit trusty image - apport fail that is | 22:45 |
| ianorlin | trusty-desktop-i386 | 22:45 |
| ianorlin | ah that works around it | 22:46 |
| bdmurray | ianorlin: apt-get update? | 22:47 |
| ianorlin | yes from the installer | 22:47 |
| bdmurray | infinity: its because apport doesn't know where the packages come from | 22:47 |
| infinity | bdmurray: Which means? | 22:47 |
| infinity | bdmurray: Seems like a bug if you can't file bugs from the livecd because apport's being too strict about how it decides a package is "official". | 22:48 |
| elfy | bdmurray: if I'm getting the same - better to abort and apt-get update and start again? | 22:49 |
| bdmurray | elfy: getting the same what? | 22:49 |
| elfy | apport issue in studio's 64 bit trusty | 22:50 |
| elfy | but half way through install | 22:50 |
| ianorlin | improve error message to make apport tell you to run sudo apt-get update if in the installer and you have ubiquity running? | 22:52 |
| elfy | infinity: ok so crashed again, can see the crash file - which is 13Mb | 22:55 |
| infinity | elfy: That's a bit larger than I'd hoped, but if you can toss that $somewhere that I can get at it, that might be nice. | 22:56 |
| infinity | elfy: I'm about to try to reproduce myself anyway, though. | 22:57 |
| elfy | trying to ;) | 22:57 |
| infinity | cyphermox: Want to grab an ubuntustudio trusty ISO, on the off chance that I need help hunting down WTF is going on? | 22:57 |
| infinity | It might just be that studio doesn't get to release with this point release if we can't sort it in a hurry, I'm not going to delay everyone else because they had no testing. :/ | 22:58 |
| cyphermox | infinity: ah, yes, I might as well start downloading | 22:58 |
| cyphermox | I was going to run off in a few minutes for karate class though, but I guess I can reschedule that to tomorrow | 22:58 |
| infinity | cyphermox: amd64. Seems elfy can only reproduce this on amd64, for some odd reason. | 22:58 |
| * infinity is hoping he can't reproduce under kvm and can just blame it on virtualbox somehow. | 22:59 | |
| cyphermox | does virtualbox/qemu matter? | 22:59 |
| cyphermox | ah, answers my question | 22:59 |
| infinity | cyphermox: Well, http://i.imgur.com/5C75Q1i.png was in vbox. | 22:59 |
| infinity | cyphermox: But I doubt it actually matters. :P | 23:00 |
| cyphermox | yeah, I saw | 23:00 |
| elfy | infinity: paste.ubuntu.com/10298040 | 23:00 |
| cyphermox | hmm, for a crash like that, and the four lines in the screenshot look fine too, I doubt it's vbox | 23:00 |
| infinity | cyphermox: Yeah, I know. | 23:00 |
| cyphermox | ... | 23:01 |
| infinity | bdmurray: So, is this behaviour a new regression? Given that ubiquity *has code* to call into apport when it crashes, either this has never worked (unlikely), ubiquity hasn't crashed for anyone in ages (unlikely), or something changed to break this. | 23:01 |
| cyphermox | ah, I see. paste.u.c is slow because the paste is ginormous. | 23:01 |
| elfy | yea ... | 23:02 |
| cyphermox | elfy: you... ran out of space? | 23:02 |
| infinity | elfy: Your disk image is too small. | 23:02 |
| elfy | orly | 23:02 |
| infinity | Feb 18 22:51:03 ubuntu-studio ubiquity: Processing triggers for man-db (2.6.7.1-1ubuntu1) ... | 23:02 |
| infinity | Feb 18 22:51:04 ubuntu-studio ubiquity: /usr/bin/mandb: | 23:02 |
| infinity | Feb 18 22:51:04 ubuntu-studio ubiquity: can't write to /var/cache/man/10957 | 23:02 |
| infinity | Feb 18 22:51:04 ubuntu-studio ubiquity: : No space left on device | 23:02 |
| infinity | elfy: studio is kinda huge. :P | 23:02 |
| elfy | mmm | 23:02 |
| bdmurray | infinity: I've tested it in utopic / vivid live cds and it (ubuntu-bug apport) works for me. I did find a bug about it failing when the installation is off-line. | 23:02 |
| elfy | yea yea - it says it wants 8.6Gb - made drive 10 | 23:02 |
| infinity | Their estimate is clearly wrong. :P | 23:03 |
| infinity | But that's a minor bug compared to an installer crashing. | 23:03 |
| infinity | So, phew. | 23:03 |
| ianorlin | I am online I kust started it before it ran apt-get update on installing | 23:03 |
| elfy | infinity: :D | 23:04 |
| ianorlin | I got apport to work using apt-get update | 23:04 |
| elfy | rerunning it for the last time ... | 23:04 |
| bdmurray | infinity: apt-cache policy apport when booting a Trusty Live CD doesn't show the version being from archive.ubuntu.com rather /var/lib/dpkg/status | 23:04 |
| infinity | cyphermox: That said, can you file a ubiquity bug for it to do a df of target before popping up the "crash" dialog and maybe suggesting something more appropriate like "you ran out of disk, doofus" instead of "please file a bug"? | 23:04 |
| infinity | bdmurray: Yes, that's expected on a livecd. | 23:04 |
| cyphermox | yeah, sounds like a good idea. | 23:05 |
| cyphermox | I'll use non-infinity wording for the error message though ;) | 23:05 |
| bdmurray | infinity: it didn't happen with utopic or vivid | 23:05 |
| infinity | bdmurray: I think we may have changed livefs creation recently to keep the indexes instead of wiping them. Don't recall now. | 23:06 |
| infinity | bdmurray: But trusty has always been the way it is, we didn't change how IT builds. | 23:06 |
| infinity | bdmurray: So either apport's always been doing this in trusty, or a change was backported that wasn't tested in that environment? | 23:06 |
| bdmurray | infinity: I'll double check the 14.04.1 release then | 23:07 |
| infinity | bdmurray: It's entirely possible this has always happened in trusty, and it's just that no one's crashed anything in the live session and complained about it. :P | 23:07 |
| elfy | infinity: 64bit must be a lot bigger than 32bit then - 32bit didn't fail on same size drive | 23:07 |
| infinity | bdmurray: But that seems unlikely, y'know? We know a lot of whiney people. | 23:07 |
| infinity | bdmurray: And ship a lot of shoddy software. | 23:07 |
| infinity | elfy: 64-bit binaries are a fair chunk larger, the i386 install might have been borderline. | 23:08 |
| infinity | elfy: If you're using the automagic partition layout, it's possible the 64-bit one also reserved more swap space, giving you a smaller root. | 23:08 |
| elfy | possible - certainly didn't look afterwards :) | 23:08 |
| elfy | aah yes | 23:09 |
| ianorlin | bdmurry you don't need me to report a problem with apport from the live session do you or should I continue to make the lubuntu i386 install is complete and boots? | 23:09 |
| bdmurray | ianorlin: no, I don't need any more information | 23:09 |
| bdmurray | infinity: it also happens on 14.04.1 image | 23:12 |
| elfy | infinity: yea :) 32 bit ... 92% of / used | 23:12 |
| infinity | bdmurray: Okay, cool. Then not a regression for .2, officially don't care today. | 23:13 |
| infinity | bdmurray: But we should sort out what the right fix is for .3 | 23:14 |
| elfy | I didn't really think about drive size, space it wanted and swap tbh - too late in the day | 23:15 |
| bdmurray | infinity: oh and I guess crashes won't be reported from point release CDs either, because of /etc/apport/crashdb.conf problem_types setting | 23:16 |
| elfy | infinity: ok - passed that now, so studio have 32/64 bit installs done and some livesession | 23:16 |
| infinity | elfy: Well, to be fair to you, the installer is monumentally unhelpful about informing you that you've run out of space. So, at least we got a cosmetic bug out of it. :P | 23:16 |
| elfy | not doing anymore today :) | 23:16 |
| infinity | elfy: Thanks for that! | 23:17 |
| elfy | infinity: :) | 23:17 |
| elfy | and your welcome :) | 23:17 |
| elfy | remember how helpful I've been when we still can't get 32bit to run on vbox next week :D | 23:18 |
| cyphermox | fwiw it's bug 1423377 | 23:19 |
| ubot93 | bug 1423377 in ubiquity (Ubuntu) "Should show a more meaningful error message on full disk" [Undecided,New] https://launchpad.net/bugs/1423377 | 23:19 |
| elfy | can I mark that affects me more than once? | 23:19 |
| elfy | :p | 23:19 |
| infinity | elfy: There might be a new kernel landing before Beta 1, if the kernel team's ducks are all in a row, maybe that'll magically fix your vbox issues. Maybe. | 23:20 |
| infinity | Not that I know what your issues are, and I don't want to know until this point release is out. :P | 23:21 |
| elfy | cyphermox: ta - I made a note on the tracker and referred bug | 23:21 |
| wxl | so is this whole bug an issue that's going to require respins? | 23:21 |
| * infinity suspects he might need beer to fuel the rest of his evening. | 23:21 | |
| infinity | wxl: There was no bug. The bug was elfy being tired. | 23:21 |
| elfy | infinity: do you think I've been drinking tea? | 23:22 |
| wxl | infinity: oh. well ianorlin too then? | 23:22 |
| * wxl is drinking tea | 23:22 | |
| infinity | wxl: Oh, the apport thing? That's not a regression, we're going to leave it be and fix it later. | 23:22 |
| wxl | infinity: k cool | 23:22 |
| elfy | infinity: and yes - magic is good too | 23:22 |
| infinity | wxl: 14.04.1 images have the same apport issue, so I officially don't care until tomorrow. :P | 23:22 |
| wxl | infinity: the workaround will be included in the main release notes, ya? | 23:22 |
| wxl | hahaha | 23:23 |
| elfy | ok - well - nice to spend some time with you gents - but I'm going to wander off now :) | 23:23 |
| elfy | night all | 23:23 |
| infinity | elfy: 'Night. | 23:24 |
| wxl | nite elfy | 23:24 |
| bdmurray | infinity: bug 1423382 | 23:24 |
| ubot93 | bug 1423382 in apport (Ubuntu) "unable to use apport to report a bug from point release Live CDs" [Undecided,New] https://launchpad.net/bugs/1423382 | 23:24 |
| wxl | ianorlin: could you make sure to include that in your report ↑ | 23:24 |
| infinity | bdmurray: The solution might be to have indexes on the livefs, if that's what we're doing in U/V... But we can talk about it when I'm not a release crunch mess. | 23:24 |
| elfy | I did | 23:24 |
| * cyphermox back in ~1h30 | 23:44 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
