| *** tiagogomes has quit IRC | 03:37 | |
| *** sambishop has quit IRC | 03:37 | |
| *** mauricemoss_ has quit IRC | 03:37 | |
| *** fay_ has quit IRC | 03:37 | |
| *** tpollard_ has quit IRC | 03:37 | |
| *** nowster has quit IRC | 03:37 | |
| *** paulw has quit IRC | 03:37 | |
| *** flatmush has quit IRC | 03:37 | |
| *** tiagogomes has joined #baserock | 03:37 | |
| *** nowster has joined #baserock | 03:37 | |
| *** paulw has joined #baserock | 03:37 | |
| *** fay_ has joined #baserock | 03:37 | |
| *** franred has quit IRC | 03:37 | |
| *** tpollard_ has joined #baserock | 03:37 | |
| *** flatmush has joined #baserock | 03:38 | |
| *** franred has joined #baserock | 03:38 | |
| *** sambishop has joined #baserock | 03:38 | |
| *** mauricemoss_ has joined #baserock | 03:38 | |
| *** sherm_ has joined #baserock | 05:43 | |
| *** petefoth has quit IRC | 05:47 | |
| *** petefoth has joined #baserock | 05:58 | |
| *** a1exhughe5 has joined #baserock | 07:21 | |
| paulsherwood | afaict devel-system-x86_64-vagrant no longer builds in master - vboxguest specifically http://sprunge.us/NbJa | 07:26 |
|---|---|---|
| *** Albert has joined #baserock | 07:28 | |
| radiofree | the file it's complaining about is supposed to be provided by libXt | 07:36 |
| radiofree | which isn't in x-common | 07:36 |
| radiofree | it was in x-generic sorry | 07:38 |
| radiofree | it was purged during the great wayland cleansing http://git.baserock.org/cgi-bin/cgit.cgi/baserock/baserock/definitions.git/commit/?id=f46ab53b310fae1d43a5ea6b959dfcb001931545 | 07:38 |
| SotK | paulsherwood: if you use my XFCE branch everything you need should be there I think | 07:39 |
| SotK | in fact, I may have updated vboxguest when I was trying to use it too actually | 07:39 |
| radiofree | needs libICE, libSM and libXt adding back | 07:40 |
| paulsherwood | SotK: actually, I should sign off on your XFCE branch - master was broken, not your changes | 07:41 |
| paulsherwood | what's the best order to fix these things? | 07:41 |
| radiofree | if SotK's XFCE branch fixes it, and XFCE as well, then merge that? | 07:41 |
| radiofree | other wise just add libICE, libSM and libXT back to strata/x-generic (see http://git.baserock.org/cgi-bin/cgit.cgi/baserock/baserock/definitions.git/commit/?id=f46ab53b310fae1d43a5ea6b959dfcb001931545) | 07:42 |
| paulsherwood | eeee, grandson, i remember the great wayland cleansing of two thousand and fifteen, like it were yesterday :) | 07:42 |
| radiofree | there might be more virtualbox needs though! | 07:42 |
| * paulsherwood checks | 07:43 | |
| *** paulw has quit IRC | 07:44 | |
| *** paulw has joined #baserock | 07:44 | |
| *** mdunford has joined #baserock | 07:46 | |
| *** bashrc_ has joined #baserock | 08:05 | |
| paulsherwood | sadly, SotK's XFCE branch does not fix build of devel-system-x86_64-vagrant | 08:07 |
| *** mariaderidder has joined #baserock | 08:07 | |
| Kinnison | I'm not surprised that the vagrant system now fails to build | 08:08 |
| Kinnison | it was a proof-of-concept and seriously hacked about with the vbox build system in order to try and only build the headless guest modules | 08:08 |
| Kinnison | the dependencies on X libraries are purely because I could not work out how to disconnect that bit from the rest of the build system | 08:08 |
| paulsherwood | ok then :) | 08:09 |
| *** jonathanmaw has joined #baserock | 08:11 | |
| *** rdale has joined #baserock | 08:17 | |
| radiofree | Kinnison: it looks like it's depending on X for the guest additions though | 08:17 |
| radiofree | for the copy-and-paste one | 08:17 |
| Kinnison | radiofree: it's headless (or IIRC supposed to be) so I tried to excise those, but it was too hard in the time I had available | 08:18 |
| Kinnison | radiofree: If you can make it only build the core and shared filesystem stuff (which was my goal) then it should be good | 08:18 |
| radiofree | personally i think we should revert the "remove these bits of X" | 08:18 |
| Kinnison | But wayland is the future(tm) and screw anyone who can't keep up for whatever reason | 08:19 |
| Kinnison | </pred> | 08:19 |
| radiofree | speaking of wayland, i think mesa is going to have to be updated for this to work with the 4.0.0 nouveau | 08:20 |
| Kinnison | Does wayland have an X emulation layer? | 08:20 |
| radiofree | yeah, XWayland | 08:20 |
| Kinnison | Okay | 08:20 |
| * Kinnison was worried for a bit that he'd reach a day when his chosen XWM wouldn't behave | 08:20 | |
| paulsherwood | Kinnison: i fear this is the way life is going to be, whether we like it or not (ie screw anyone who can't keep up for whatever reason) | 08:21 |
| radiofree | would an optional "arch" field for chunks in strata be ok? | 08:21 |
| paulsherwood | radiofree: what for? | 08:21 |
| radiofree | well we need two versions of mesa, one for a jetson, one for x86 | 08:21 |
| radiofree | however if you do that you end up with mesa-x86, mesa-jetson, then everything that needs mesa-* needs to branch | 08:22 |
| radiofree | however if we had one mesa-common, and defined mesa-x86, arch:x86, mesa-arm, arch:arm | 08:22 |
| Kinnison | radiofree: is there no way to unify the source trees and switch on build? | 08:22 |
| radiofree | i'd rather keep up with mesa where we can, rather than cherry-pick patches onto an older one | 08:23 |
| Kinnison | So the issue is that jetson can't move forward? | 08:23 |
| radiofree | indeed, because they removed gallium-egl in later versions of mesa (that the one we're using) | 08:23 |
| radiofree | which means no testing EGL clients in a vm | 08:24 |
| Kinnison | By paulsherwood's remark "screw Jetson" is to be our approach I guess then | 08:24 |
| paulsherwood | Kinnison: i was speaking philosophically, nothing to do with the current situation | 08:24 |
| Kinnison | :) | 08:25 |
| * Kinnison was liberally extending :) | 08:25 | |
| rjek | The question is, how many screws does the average Jetson have? | 08:25 |
| paulsherwood | heh | 08:25 |
| radiofree | wait a minute, the issue was x86 can't move forward | 08:35 |
| radiofree | we need to move forward *for* the jetson :) | 08:35 |
| Kinnison | Oh right | 08:36 |
| Kinnison | And the only reason we can't move x86 forward is because it removes the ability to run EGL in a VM? | 08:36 |
| radiofree | yes, people (e.g javier) use that for testing gtk, gnome, weston etc... | 08:36 |
| radiofree | and we do release a genivi vm image which wouldn't work with the examples we ship with it | 08:37 |
| paulsherwood | SotK: does your XFCE system work, btw? | 08:41 |
| paulsherwood | (ie is it usable) | 08:41 |
| *** gary_perkins has joined #baserock | 08:48 | |
| *** a1exhughe5 has quit IRC | 08:48 | |
| *** a1exhughe5 has joined #baserock | 08:49 | |
| *** ssam2 has joined #baserock | 08:55 | |
| *** ChanServ sets mode: +v ssam2 | 08:55 | |
| SotK | paulsherwood: yeah | 09:01 |
| SotK | run `startxfce4` | 09:01 |
| SotK | you may need to configure the desktop somewhat before its easily usable though | 09:02 |
| * SotK doesn't know how to change the initial layout | 09:02 | |
| SotK | s/layout/layout by default/ | 09:02 |
| SotK | I must have updated vboxguest when I managed to get it to build, I'll look when I get home | 09:06 |
| *** Krin has joined #baserock | 09:09 | |
| *** lachlanmackenzie has joined #baserock | 09:15 | |
| *** wschaller has joined #baserock | 09:29 | |
| bjdooks | 33 | 09:40 |
| jmacs | 45 | 09:41 |
| richard_maw | 128 | 09:41 |
| richard_maw | sorry, what game are we playing? | 09:41 |
| Kinnison | 2048 | 09:41 |
| * Kinnison wins | 09:41 | |
| nowster | numberwang? | 09:41 |
| jjardon | 24 | 09:43 |
| Kinnison | Oh dear | 09:43 |
| jjardon | 42, sorry :P | 09:44 |
| rjek | This is window 17. | 09:44 |
| SotK | don't be silly, this is window 4 | 09:45 |
| * richard_maw agrees with SotK | 09:45 | |
| rjek | richard_maw: That I shouldn't be silly, or that this is window 4? | 09:47 |
| *** mdunford has quit IRC | 09:47 | |
| richard_maw | it's window 4 | 09:47 |
| perryl | this channel isn't right anywhere other than window 4 | 09:48 |
| nowster | cashier number 3, please | 09:48 |
| * richard_maw happens to have two window 4s, but that's because he runs two irssi sessions to split work identity from non-work | 09:48 | |
| bashrc_ | mornington crescent! | 09:48 |
| Zara | this is window 11 | 09:49 |
| nowster | ceci n'est pas une fenêtre | 09:50 |
| rjek | Mao. | 09:51 |
| richard_maw | rjek: http://i.imgur.com/nhmydxK.jpg | 09:52 |
| Zara | :0 | 09:52 |
| * rjek flees | 09:54 | |
| *** ssam2 has quit IRC | 10:05 | |
| straycat | if anyone can spare just a little time to review https://gerrit.baserock.org/#/q/status:open+project:baserock/baserock/definitions+branch:master+topic:ripsum/ntpd-returns that would be of huge help! | 10:07 |
| straycat | doesn't have to be every patch either, just whatever you might have time for :) | 10:08 |
| pedroalvarez | straycat: assuming that `useradd .. -u 1000` is for the userid, is that really needed? | 10:09 |
| straycat | if we ever want to run ntpd in a chroot (and i do want that one day) then yes | 10:10 |
| *** ssam2 has joined #baserock | 10:20 | |
| *** ChanServ sets mode: +v ssam2 | 10:20 | |
| jjardon | straycat: shouldt ntp be a system account? (0-999 range) | 10:28 |
| jjardon | straycat: I think the way to create those is with the -r option | 10:29 |
| straycat | jjardon, ahh yes good point | 10:31 |
| jjardon | any feedback about https://gerrit.baserock.org/#/c/325/ ? if you agree with the change, It would be great to merge it soon before some change in the x86 bsp is needed | 10:35 |
| *** ssam2 has quit IRC | 10:38 | |
| straycat | jjardon, quite a big change, will try to look at it later :| | 10:39 |
| jjardon | straycat: sure, np | 10:40 |
| straycat | jjardon, i'll resubmit that change with hardcoded uid/gids swapped for -r | 10:40 |
| straycat | since ntpd's not running in a chroot atm anyway | 10:41 |
| SotK | fixing merge conflicts with devel system strata lists is pretty annoying | 10:41 |
| SotK | wq | 10:43 |
| straycat | :) | 10:44 |
| jjardon | straycat: thanks | 10:48 |
| *** ssam2 has joined #baserock | 10:50 | |
| *** ChanServ sets mode: +v ssam2 | 10:50 | |
| ssam2 | I have a couple of Ansible playbooks for managing a distbuild network, and would like to merge them somewhere. Any thoughts where? | 10:51 |
| ssam2 | definitions.git, infrastructure.git or a new repo? | 10:51 |
| ssam2 | I think definitions.git/distbuild/admin would be a good enough place for them | 10:51 |
| jjardon | ssam2: +1 | 10:54 |
| ssam2 | maybe they could be installed by distbuild.configure into /usr/lib/distbuild-admin in fact | 10:54 |
| Kinnison | paulsherwood: I've gotten my first system artifact out of ybd which is complete, incl. running the system integration commands | 10:59 |
| * Kinnison probably ought to stuff an 'ldconfig' into there too | 10:59 | |
| * Kinnison checks the build logs in case it already did one | 10:59 | |
| Kinnison | aah it did that for itself zo/ | 11:00 |
| *** mdunford has joined #baserock | 11:03 | |
| paulsherwood | Kinnison: is that a partial w00t? :) | 11:11 |
| ssam2 | i'm very impressed by how clean all the openstack patches are. well done everyone involved with that! | 11:33 |
| pedroalvarez | ssam2: hehe, took me 18 hours to organise them :) | 11:35 |
| * pedroalvarez realises that he is testing openstack with kernel 4.0.0 | 11:43 | |
| *** wschaller_ has joined #baserock | 11:43 | |
| *** wschaller has quit IRC | 11:43 | |
| * paulsherwood hopes pedroalvarez is excited | 11:44 | |
| pedroalvarez | I am :) | 11:44 |
| pedroalvarez | given that almost nothing was failing :P | 11:44 |
| paulsherwood | Kinnison: did you get any chance to look at that pygobject weirdness? | 11:46 |
| paulsherwood | (i'm aware you may have more pressing and interesting things to do) | 11:46 |
| paulsherwood | can somone please add ed.cragg as an email alias for the chap works here and seems to go by that name? | 11:51 |
| rjek | paulsherwood: Wrong channel? | 11:52 |
| paulsherwood | oops... wrong channel indeed. i'm an idiot. quote me :) | 11:52 |
| ssam2 | it's taking my Jetson more than 30 minutes to download the linux-stable.git tarball from git.baserock.org ... | 11:55 |
| paulsherwood | erk | 11:55 |
| ssam2 | could be a connection issue, but seems weird | 11:56 |
| paulsherwood | ssam2: re-start it | 11:56 |
| pedroalvarez | ssam2: I noticed yesterday some slow connectivity with cache.baserock.org | 11:56 |
| paulsherwood | i've seen that with morph occasionally (not jetson-specific) | 11:56 |
| paulsherwood | ie sometimes morph has *stopped* while getting tarball for me... never managed to reproduce it though | 11:57 |
| ssam2 | :( | 11:59 |
| ssam2 | downloading a large file with wget on this Jetson is going equally slow, so it is the connection | 11:59 |
| ssam2 | or the board | 11:59 |
| rjek | ssam2: I've seen bad git performance from here today too, might be congestion at our end. | 11:59 |
| ssam2 | it's much faster on my laptop. So I think it's the board | 12:00 |
| ssam2 | laptop and Jetson are plugged into the same switch | 12:00 |
| rjek | Oh, hmm. | 12:00 |
| ssam2 | i would restart it, but it's not far off done and I don't want to have to start from the beginning :) | 12:01 |
| radiofree | ssam2: that's interesting, try installing the rtl_nic/rtl8168g-2.fw firmware and test again | 12:01 |
| paulsherwood | ssam2: can you confirm the tarball is still downloading, not stoppped? | 12:01 |
| rjek | heh | 12:01 |
| ssam2 | it's still downloading, 100KB/sec or so now which isn't so bad | 12:02 |
| ssam2 | radiofree: i'm not sure i'm going to have time to do that right now, i'll note it down though | 12:02 |
| radiofree | i'm getting equally crap performance downloading http://releases.ubuntu.com/14.04.2/ubuntu-14.04.2-desktop-amd64.iso on my laptop and jetson | 12:02 |
| ssam2 | maybe it's just our network. | 12:03 |
| paulsherwood | why would you want that on a jetson, radiofree ? :) | 12:03 |
| straycat | jjardon, thanks for that review | 12:03 |
| radiofree | paulsherwood: it's the largest "quick i need a big file to download from a fast server" i could think of | 12:03 |
| straycat | i think on-failure might be better than always, but it's kind of academic | 12:03 |
| radiofree | which is odd considering i use fedora | 12:04 |
| jjardon | straycat: yw | 12:04 |
| *** Albert has quit IRC | 12:41 | |
| *** Albert has joined #baserock | 12:41 | |
| Kinnison | paulsherwood: sadly I've had way too much fun with system integrations to look at pygobject yet | 12:46 |
| * Kinnison supposes he could take a pause | 12:46 | |
| Kinnison | paulsherwood: using the command you gave me: | 12:46 |
| Kinnison | 2015-04-15 12:46:39 [pygobject] ERROR: No definition found for pygobject | 12:46 |
| robtaylor | rdale: ooi, have you got systemd building with musl yet? | 12:50 |
| rdale | no, i'm not sure if systemd can be built with musl | 12:51 |
| rdale | so i haven't actually tried | 12:51 |
| paulsherwood | Kinnison: are you using master? | 12:54 |
| * paulsherwood blushes, having fixed a bug recently | 12:54 | |
| robtaylor | rdale: i'd be interested to hear how it does if you do :) | 12:54 |
| rdale | ok, i'm just looking at the systemd mail archives where there is a discussion about musl | 12:55 |
| robtaylor | rdale: yeah looks like Tom applied the patchset | 12:58 |
| Kinnison | paulsherwood: aah, I'll workout some clean commits and rebase then | 13:00 |
| pedroalvarez | jjardon: my answer here is a bit messed up: https://gerrit.baserock.org/#/c/310/1 | 13:03 |
| pedroalvarez | but is my answer anyway, I don't want to reply the same in all the components. | 13:03 |
| ssam2 | radiofree: this is what you need to deploy a Mason: http://git.baserock.org/cgi-bin/cgit.cgi/baserock/baserock/infrastructure.git/tree/baserock_mason_x86_64 | 13:32 |
| ssam2 | deploy the .morph file with `morph deploy`, then once the system is running, copy distbuild.conf to /etc/distbuild and mason.conf to /etc/mason, then reboot | 13:33 |
| ssam2 | you also need an ssh key called worker.key in /etc/distbuild in fact, which isn't used for anything (unless you need to build private repos) | 13:35 |
| jjardon | pedroalvarez: thanks ,I commented in gerrit | 13:42 |
| pedroalvarez | jjardon: wow, patch the scripts that generate the configuration to configure it for ourselves? | 13:43 |
| pedroalvarez | heh, that is going to be hard work | 13:43 |
| radiofree | ssam2: thanks | 13:44 |
| jjardon | pedroalvarez: if thats going to stop you, ignore my comments; I only wanted to flag it because we will have to do that work at some point when you try to upgrade and the format of those configuration files have changed | 13:46 |
| jjardon | pedroalvarez: how other openstack instances personalize the configuration files? | 13:47 |
| pedroalvarez | jjardon: that would be a good starting point. | 13:48 |
| pedroalvarez | and yes, upgrading to kilo is not going to be straightforward now | 13:49 |
| radiofree | hmm i probably should submit a patch to the jetson clusters to use BOOT_DEVICE | 13:50 |
| Kinnison | paulsherwood: It appears to be related to tput returning 1 | 13:51 |
| Kinnison | paulsherwood: which might be a bug in gnome-autogen.sh | 13:51 |
| paulsherwood | interesting... i wonder why morph would be happy, then? | 13:52 |
| * Kinnison is digging | 13:52 | |
| jjardon | what version of gnome-common are you using? It got a lot of patches recently | 13:53 |
| paulsherwood | whatever is in definitions | 13:53 |
| paulsherwood | (master, ish i think) | 13:53 |
| jjardon | 3.12, that's ancient. I will write a patch | 13:54 |
| franred | pedroalvarez, jjardon, we assume that the Openstack system runs at fist boot without any other manual configuration - which implies we need to modify the configuration files, and for tracking the changes are better to put them into the repository | 13:54 |
| franred | that makes upgrading them a little bit more difficult but much more traceable :) | 13:55 |
| paulsherwood | which repository, franred ? | 13:55 |
| jjardon | franred: Im not suggesting that a manual configuration is needed | 13:56 |
| Kinnison | paulsherwood: I *think* we're encountering a difference between ybd's use of subprocess.call() vs. morph's use of cliapp.runcmd() | 14:00 |
| paulsherwood | wow! :) | 14:00 |
| Kinnison | Not particularly wow | 14:01 |
| Kinnison | it's also a difference between the 'tty' command and the 'tput' command :) | 14:01 |
| Kinnison | it's an icky combination | 14:01 |
| Kinnison | But ybd isn't sealing off the build quite cleanly | 14:01 |
| franred | paulsherwood, https://gerrit.baserock.org/#/q/status:open+project:baserock/baserock/definitions+branch:master+topic:baserock/openstack-in-baserock - hopefully master definitions soon ;-) | 14:03 |
| Kinnison | paulsherwood: I'm not going to claim it's *clean* but http://paste.baserock.org/elocuxotud makes pygobject able to be built | 14:03 |
| *** sherm_ has quit IRC | 14:04 | |
| ssam2 | SotK: is there an easy way to test your OStree branch? | 14:10 |
| ssam2 | I basically want to avoid having to run 'git-review -d' for each patch in the series | 14:10 |
| SotK | ssam2: its pushed to g.b.o as baserock/adamcoldrick/ostree I think | 14:11 |
| ssam2 | cool, thanks | 14:12 |
| *** sherm_ has joined #baserock | 14:16 | |
| *** mike has joined #baserock | 14:17 | |
| paulsherwood | Kinnison: i don't even begin to understand this, so if it works and you're ok with it, i'll accept that patch :) | 14:17 |
| *** mike is now known as Guest25758 | 14:17 | |
| Kinnison | paulsherwood: it's to do with how stdin is processed in the runcmd pipelines | 14:18 |
| Kinnison | paulsherwood: this is the smallest patch to get a similar effect | 14:18 |
| paulsherwood | well so long as it doesn't break the rest, i'm ok with it :) | 14:19 |
| richard_maw | it won't | 14:20 |
| paulsherwood | excellent | 14:24 |
| * straycat would maybe like to document the simple-network conf ext on the wiki | 14:24 | |
| straycat | but not today :) | 14:25 |
| straycat | ssam2, if you git review -d the final change in the topic you get all of them? | 14:31 |
| ssam2 | oh, i didn't think of that | 14:32 |
| ssam2 | SotK: the ostree branch seems to download the .stratum files from the cache but not the stratum.meta files | 14:51 |
| ssam2 | but then a deploy fails because the .meta file isn't available | 14:51 |
| ssam2 | seems it was easy to fix, but I wish that Morph just always downloaded whole artifacts | 14:53 |
| ssam2 | ie. .meta, .build-log, everything | 14:53 |
| ssam2 | I have a few fixes for the ostree branch for things i've come across, should I push them somewhere for you to look at? | 14:53 |
| SotK | ssam2: please do :) | 14:54 |
| ssam2 | ok, morph.git branch sam/ostree-fixes-1 | 14:57 |
| ssam2 | i've found these while trying to `morph deploy` a system on x86 that was built with distbuild on a Jetson | 14:57 |
| ssam2 | so downloading everything from the remote cache | 14:57 |
| *** Guest25758 has quit IRC | 14:57 | |
| Kinnison | paulsherwood: I've pushed a bunch of commits to my ybd repo on github -- I'm not satisfied with them all yet, but you can see where I'm going with things | 15:02 |
| radiofree | ssam2: do you get a snazzy web-ui if you you follow that? | 15:02 |
| ssam2 | yes | 15:02 |
| paulsherwood | Kinnison: careful... upstream is trigger-happy... they may get merged :) | 15:02 |
| ssam2 | the snazziest one | 15:02 |
| radiofree | sorry didn't realise i was scrolled back so far, that was in relation to the mason stuff | 15:02 |
| jjardon | SotK: is ostree already in definitions? I cant find it | 15:04 |
| Kinnison | paulsherwood: I won't cry if that happens, but I'm not asking for it any time soon | 15:05 |
| * Kinnison is still running tests on some of the stuff | 15:05 | |
| paulsherwood | k | 15:05 |
| Kinnison | I want to run a full build with a fresh artifact cache and then compare the results | 15:05 |
| SotK | jjardon: not yet, there is a branch for it in definitions somewhere though | 15:05 |
| Kinnison | fr.ex. | 15:05 |
| SotK | I'm currently rebasing said branch onto master and carrying out a couple of fixups | 15:05 |
| rjek | Kinnison: While you're there, do you fancy testing richard_maw's DES password fix? | 15:07 |
| paulsherwood | Kinnison: bah! 2015-04-06 21:08:31 [librabbitmq] ERROR: Git submodules problem | 15:08 |
| paulsherwood | (ignore the timewarp) | 15:08 |
| Kinnison | rjek: ybd isn't able to deploy yet | 15:08 |
| Kinnison | rjek: that's next on my list | 15:09 |
| Kinnison | paulsherwood: is that my fault? | 15:09 |
| paulsherwood | nope, i don't think so | 15:09 |
| paulsherwood | i'm just moaning | 15:09 |
| rjek | Kinnison: ;-) | 15:09 |
| paulsherwood | i shouldn't even be here | 15:09 |
| Kinnison | ♫ ... I don't belong here ... | 15:11 |
| paulsherwood | lol | 15:14 |
| *** zoli__ has joined #baserock | 15:15 | |
| straycat | https://gerrit.baserock.org/#/c/332/ is a very quick fix i need for simple-network conf ext | 15:17 |
| straycat | i realise we should put all that stuff in a dict | 15:17 |
| straycat | but if we can use this for now i'd appreciate it | 15:17 |
| straycat | jjardon, ^ | 15:17 |
| straycat | merged thanks | 15:19 |
| jjardon | wow, other were faster than me :) | 15:19 |
| paulsherwood | heh | 15:19 |
| straycat | :) | 15:20 |
| jjardon | but Id not use that simple-network configuration file if you plan to use systemd: but the install-files instead: simply create the .network file and you are done | 15:20 |
| jjardon | the rework in simple-network was done to be able to upgrade old system that do not use systemd-networkd | 15:21 |
| jjardon | straycat: ^ | 15:21 |
| straycat | jjardon, i hadn't thought of that actually, but i guess i can continue using simple-network for now | 15:22 |
| ssam2 | oops, i forgot about that as well | 15:23 |
| ssam2 | SotK: does the ostree branch handle the case where an old-style system artifact is available in the remote artifact cache? | 15:23 |
| SotK | ssam2: no | 15:23 |
| ssam2 | i've noticed that my deploy has downloaded a full system tarball from the remote cache and is now committing it to ostree. I'm excited whether it will work or not :) | 15:23 |
| jjardon | straycat: sure, but with the install-files approach you have the full potential of networkd without the need of change any script (you can add .match files as well) | 15:24 |
| SotK | heh, I don't see why it wouldn't work | 15:24 |
| straycat | jjardon, that's very true, and now that install files can be templates you wouldn't even need to hardcode anything in the install files themselves | 15:25 |
| SotK | it will mitigate the speedups a bit though | 15:25 |
| ssam2 | indeed | 15:25 |
| ssam2 | might be good to rebuild the system... although since this was built on ARM, it can't | 15:25 |
| ssam2 | as long as it works, I guess this is only a temporary problem | 15:25 |
| ssam2 | great, I seem to have broken the glibc build :( | 15:27 |
| SotK | :( | 15:27 |
| ssam2 | with .http://git.baserock.org/cgi-bin/cgit.cgi/baserock/baserock/definitions.git/commit/?h=master&id=43ae529a39e96602335e0282d9f294614f69d5e4 ... no idea how that broke it on x86 | 15:27 |
| ssam2 | although, the last logged failure is about 9AM, so I guess it's a random failure | 15:28 |
| radiofree | hd space? | 15:28 |
| ssam2 | seems to be some weird race condition where the mason shell script tries to deploy the system before it's finished building | 15:29 |
| radiofree | "2015-04-15 09:15:39 Finished building glibc-doc" | 15:29 |
| SotK | ssam2: that seems to happen semi-frequently :/ | 15:30 |
| ssam2 | bug in Mason I guess | 15:30 |
| ssam2 | 12.1GB is free on cache.baserock.org | 15:30 |
| ssam2 | time-to-running-first-.write-extension for the distbuild system I'm deploying seems to have gone down to 7s with OSTree, was over 3 minutes before | 15:35 |
| ssam2 | it did take quite a lot of time to download the artifacts and commit them into the ostree artifact cache though | 15:36 |
| ssam2 | but that only needs to be done once, not on every deploy | 15:36 |
| SotK | download will be quicker when the remote cache is also using ostree | 15:36 |
| ssam2 | mm, true | 15:36 |
| jjardon | o_O | 15:36 |
| ssam2 | can we do that? :) | 15:37 |
| radiofree | can i use this wonderful 3minute+ -> 7s technology? | 15:37 |
| * SotK doesn't know what will happen if old morph talks to a morph-cache-server from my ostree branch actually | 15:38 | |
| ssam2 | radiofree: yes, but you have to redeploy your system from the ostree branch. I think adam sent a mail to the list with instructions | 15:38 |
| *** Krin has quit IRC | 15:50 | |
| *** petefoth_ has joined #baserock | 15:57 | |
| *** zoli__ has quit IRC | 15:58 | |
| *** petefoth has quit IRC | 15:58 | |
| *** petefoth_ is now known as petefoth | 15:58 | |
| *** a1exhughe5 has quit IRC | 16:01 | |
| *** wschaller_ has quit IRC | 16:32 | |
| jjardon | Any suggestions about where pygobject should be? would it be ok to put it in foundation? or should we create another stratum? what about libsoup? | 16:32 |
| franred | jjardon, python-common? | 16:32 |
| jjardon | SotK: maybe you already have some idea/patch about this ^ | 16:32 |
| ssam2 | one of the python strata, rather than foundation | 16:35 |
| ssam2 | I think that if in doubt we should create a new stratum for each thing | 16:36 |
| ssam2 | and in future, make strata optional | 16:36 |
| ssam2 | i know it's painful right now having lots of strata, but it's something that we need to fix, one way or another | 16:36 |
| jjardon | sure, problem is that all the python strata depends on core, but glib is in foundation | 16:38 |
| jjardon | maybe is better to create a separate stratum? | 16:38 |
| *** mariaderidder has quit IRC | 16:39 | |
| jjardon | and maybe move there libsoup as well? we could call it gobject-libraries? mmm, not sure I like this idea though | 16:40 |
| ssam2 | no, classifying libraries by whether or not they use GObject doesn't seem useful | 16:41 |
| franred | jmacs, pedroalvarez, I've updated the commit message in https://gerrit.baserock.org/#/c/308/ to show where the files come from | 16:42 |
| rdale | i think most of the gnome libs in foundation are there to support systemd, and maybe there should be a systemd stratum | 16:42 |
| franred | s/files/config files/ | 16:42 |
| jjardon | would it be ok move libsoup to the connectivity stratum? | 16:42 |
| ssam2 | jjardon: this is really an edge case for the 'metapackages' concept: we can either have an 'ostree' stratum that has stuff in it that other things need, or we can admit that trying to group everything into metapackages isn't always possible | 16:42 |
| jjardon | rdale: agree | 16:42 |
| jmacs | franred: cool | 16:44 |
| jjardon | ssam2: sure, what I want to avoid is to repeat chunks in different strata | 16:44 |
| pedroalvarez | franred: thanks | 16:48 |
| *** bashrc_ has quit IRC | 17:05 | |
| *** tpollard_ has quit IRC | 17:10 | |
| *** mdunford has quit IRC | 17:10 | |
| *** gary_perkins has quit IRC | 17:15 | |
| *** ssam2 has quit IRC | 17:17 | |
| SotK | why not make the python strata depend on foundation? | 17:27 |
| straycat | jjardon, sorry only just spotted your other comment about the template, have responded | 17:28 |
| straycat | the other obvious alternative is to patch ntpd to set a higher rlimit | 17:28 |
| straycat | i'd rather use what we have for the moment | 17:29 |
| jjardon | I'd like to avoid unnecessary dependencies as well | 17:31 |
| straycat | :s | 17:31 |
| straycat | i don't follow? | 17:32 |
| straycat | jjardon, ? | 17:33 |
| SotK | what is the problem with chunks in multiple strata anyway ooi? | 17:35 |
| straycat | which one do you include in the system? | 17:37 |
| straycat | jjardon, i'm just going to leave this in the configure extension | 17:38 |
| straycat | we can change it later if we really need to, i'd like to try and get ntpd working in a chroot anyway at some point | 17:38 |
| SotK | ah, yeah | 17:39 |
| *** zoli__ has joined #baserock | 17:45 | |
| franred | SotK, it doesn't make much sense to add foundation as a dependency of python-core or python-common, these strata should be pretty simple and depending on core, IMHO | 17:51 |
| franred | s/much sense/much sense to me/ | 17:51 |
| straycat | jjardon, i've updated the changes in https://gerrit.baserock.org/#/q/status:open+project:baserock/baserock/definitions+branch:master+topic:ripsum/ntpd-returns to match your suggestions | 17:55 |
| straycat | ahh forgot to update the commit msg for one of the changes, neat that gerrit lets me do that through the web interface | 17:56 |
| *** zoli__ has quit IRC | 18:07 | |
| *** lachlanmackenzie has quit IRC | 18:21 | |
| jjardon | straycat: sorry, I was replying to SotK comment about make python strata depend on foundation | 18:23 |
| jjardon | Thanks for the rework, I will take a look when I get home | 18:24 |
| straycat | jjardon, thanks! | 19:08 |
| *** Albert has quit IRC | 19:42 | |
| *** zoli__ has joined #baserock | 20:01 | |
| SotK | paulsherwood: looks like I was building using v4.3.26 of VirtualBox | 20:05 |
| *** rdale has quit IRC | 20:46 | |
| straycat | what, gerrit didn't merge my change into baserock/openstack-v7 | 20:50 |
| SotK | what do you mean? | 20:51 |
| straycat | i mean https://gerrit.baserock.org/#/q/status:merged+project:baserock/baserock/definitions+branch:baserock/openstack-v7 isn't in baserock/openstack-v7 on gbo | 20:51 |
| richard_maw | it takes a few minutes to sync | 20:52 |
| straycat | it was merged this afternoon | 20:52 |
| richard_maw | hrm | 20:52 |
| SotK | oh, that's weird | 20:52 |
| SotK | does the gerrit-replication plugin only push certain branches maybe? | 20:52 |
| straycat | i have a horrible feeling that might be the case | 20:53 |
| straycat | okay i'm going to go ahead and push that change manually | 20:53 |
| pedroalvarez | straycat: i took your change and included it in the initial patch series | 21:45 |
| pedroalvarez | Sorry I didn't flag that error, it has been a busy day. But I did make sure that your patch was in :) | 21:47 |
| *** zoli__ has quit IRC | 22:17 | |
| *** robtaylo1 has joined #baserock | 23:34 | |
| *** robtaylor has quit IRC | 23:38 | |
| *** bfletcher has quit IRC | 23:39 | |
| *** perryl has quit IRC | 23:39 | |
| *** perryl has joined #baserock | 23:40 | |
| *** bfletcher has joined #baserock | 23:42 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
