gchristensen changed the topic of #nixos-dev to: NixOS Development (#nixos for questions) | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html | 18.03 release managers: fpletz and vcunat | https://logs.nix.samueldr.com/nixos-dev
viric has quit [Ping timeout: 264 seconds]
viric has joined #nixos-dev
Synthetica has quit [Quit: Connection closed for inactivity]
mbrgm has quit [Ping timeout: 248 seconds]
mbrgm has joined #nixos-dev
Sonarpulse has quit [Ping timeout: 256 seconds]
orivej has quit [Ping timeout: 256 seconds]
cstrahan has quit [*.net *.split]
shlevy has quit [*.net *.split]
zimbatm has quit [*.net *.split]
pauldub has quit [*.net *.split]
shlevy has joined #nixos-dev
zimbatm has joined #nixos-dev
pauldub has joined #nixos-dev
cstrahan has joined #nixos-dev
LnL has quit [Ping timeout: 260 seconds]
mbrgm has quit [Ping timeout: 240 seconds]
LnL has joined #nixos-dev
mbrgm has joined #nixos-dev
<peti> Has anyone tried discourse.nixos.org? I heard there is some mailing list-like interface to it, but I don't see how to use that.
<peti> Oh. This sounds like you need to register with discourse before you can subscribe. That's not exactly an improvement over Google Groups. :-(
<ekleog> well, I guess the point is you need to register by clicking “login with your GitHub account,” account you supposedly already have if you're helping for nixpkgs (but yeah that's sad)
<MichaelRaskin> Well, at least they didn't demand to fill more than a single proper mailing list reply would show… I am OK with adding extra entries to my password manager. I also hope that — unlike Google — my account will stay unblocked without attempts to extort a phone number
<sphalerite> Even the password is optional
<sphalerite> you can sign in using just email as well
<sphalerite> the only thing that's missing is being able to subscribe by email
<sphalerite> AFAIK
MichaelRaskin has quit [Quit: MichaelRaskin]
ckauhaus has joined #nixos-dev
orivej has joined #nixos-dev
__Sander__ has joined #nixos-dev
orivej has quit [Ping timeout: 256 seconds]
vcunat has joined #nixos-dev
<niksnut> copumpkin: hah, I just got a kernel panic trying to boot an Ubuntu 18.04 AMI
<niksnut> it's heartwarming to see that much bigger distros also manage to produce broken AMIs ;-)
<vcunat> :-D
orivej has joined #nixos-dev
<__Sander__> hmm any libgdiplus "experts" here?
<__Sander__> I'm curious why both libungif and giflib are a dependency
taktoa has joined #nixos-dev
<__Sander__> I'm running an app that tries to open gif images and it crashes as a result
<__Sander__> removing libungif seems to fix it :)
acowley has quit [Ping timeout: 240 seconds]
acowley has joined #nixos-dev
disasm has quit [Quit: WeeChat 2.0]
ckauhaus has quit [Quit: Leaving.]
disasm has joined #nixos-dev
orivej has quit [Ping timeout: 268 seconds]
<srhb> clever: Eval 1453885 has the crc32c workaround for vms, fingers crossed. :)
orivej has joined #nixos-dev
<clever> srhb: nice
<srhb> Not so nice. Think it broke a bunch of other stuff. Investigating in a bit and then reverting.
<clever> srhb: the warning i added is not printed in that log
<clever> srhb: and that looks like normal stage-1, not vm
<srhb> I thought it would be in a dependency. Maybe I'm mistaken and it's nondeterministic.
<srhb> Or I bisected wrong. Could be!
<clever> it looks like /dev/vdb doesnt exist
<clever> hydra claims my commit is the only change that occured
<srhb> I'll rerun locally, maybe the bisect failure was just nondeterminism and just happened to be on that commit by accident.
<clever> yeah
<srhb> /nix/store/ls8ickgdpw0yr35bb3lf58gy3i7sj5pc-vm-test-run-installer-btrfsSubvolDefault produces succesfully. Sigh... :-) Sorry for the alarm.
<srhb> Oh but hm, that's not quite the same path
<srhb> Shouldn't it be this: nix-build nixos/release.nix -A tests.installer.btrfsSubvolDefault.x86_64-linux ?
<clever> did you use the exact same nixpkgs rev?
<clever> and the install contains a copy of nixpkgs
<clever> so any dirty files in the repo
<srhb> Ah, I did have some dirt. One more try.
<clever> /nix/store/afazcbc6vl1v6i1xi31hdlzzylynslak-vm-test-run-installer-btrfsSubvolDefault
<clever> works here, on the 51998c675a commit
<srhb> According to hydra it should be /nix/store/a3rnr5dqxxshxbxjzx4wxg8v7fz8hw4i-vm-test-run-installer-btrfsSubvolDefault
<srhb> Go figure. I think it's just not deterministic. That's probably okay.
zybell_ has quit [Ping timeout: 256 seconds]
<gchristensen> 12 hour ofborg outage info: https://www.patreon.com/posts/12hr-ofborg-18600393
zybell_ has joined #nixos-dev
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
Jackneilll has quit [Quit: Leaving]
Synthetica has joined #nixos-dev
Jackneill has joined #nixos-dev
<copumpkin> niksnut: nice! same NVME mount issue? also, seeing the content-addressable error again so do you think we could get a 2.0.2 release with that in it?
<niksnut> no, something about a kernel paging failure
<niksnut> 2.0.2 is out
<gchristensen> happy 2.0.21
<gchristensen> happy 2.0.2!
<niksnut> it was a t1.micro instance btw
<niksnut> maybe 600 MB is just not enough for the linux kernel nowadays...
<copumpkin> weird
<gchristensen> O.o
* dtz does lambda dance
<copumpkin> niksnut: any idea why nixpkgs-unstable seems to be behind the latest successful hydra trunk build? https://github.com/NixOS/nixpkgs-channels/tree/nixpkgs-unstable vs. https://hydra.nixos.org/jobset/nixpkgs/trunk
<gchristensen> copumpkin: it takes "a while" for it to update as it builds indexes and what-not
Sonarpulse has joined #nixos-dev
<niksnut> no, but I see that the 17.09 mirroring is failing
<niksnut> May 04 14:40:26 webserver update-nixos-17.09-start[13369]: job 'nixos.iso_minimal.i686-linux' lacks a store path at /nix/store/w6a76wmw3lqa2ykscndcq43r61wgy0rn-nixos-channel-scripts/bin/.mirror-nixos-branch-wrapped line 103.
<gchristensen> we should probably make all the channel building (or as much as possible) happen as a job inside hydra so this can be more transparent
<niksnut> channel building is impure so it can't happen in a job
<niksnut> copumpkin: the reason is that the tarball job is failing
<gchristensen> right. I guess then as much as possible things like "are all the necessary things here" could be (should be) part of `tested`
<copumpkin> niksnut: tarball job looks like worker died?
<niksnut> could be
orivej has quit [Ping timeout: 264 seconds]
pie_ has joined #nixos-dev
<copumpkin> so on the M5 mysterious boot failure
<copumpkin> /dev/disk/by-label/nixos never appears and the wait loop expires, then the system tries to mount it and dies
<copumpkin> lsblk shows nvme0n1p1 but doesn't display a label for it
<copumpkin> who's supposed to read the label and does anyone know why it might not be appearing?
<copumpkin> this feels like it might be some sort of kernel race in NVME land, but the Amazon Linux AMI has no issues whatsoever
<niksnut> udev is supposed to create the by-label link
<niksnut> but I think it calls lsblk
<niksnut> or blkid
<copumpkin> yeah, it's all very strange
<copumpkin> the whole /dev/disk thing doesn't appear
<copumpkin> and even the /dev/nvme0n1p1 thing doesn't appear at first
<copumpkin> eventually it does, but lsblk shows no label for it, and /dev/disk never shows up
<copumpkin> my colleague is gonna add 4.17 support and see if it's a kernel bug
<copumpkin> given all the recent nvme patches that landed
<copumpkin> but it's still weird that the Amazon Linux AMI works fine
<copumpkin> oh, and RHEL 7
__Sander__ has quit [Quit: Konversation terminated!]
Jackneill has quit [Quit: Leaving]
Jackneill has joined #nixos-dev
<bgamari> Sonarpulse, do you know how to portably depend upon locale data?
<bgamari> glibcLocales seems like the wrong solution
<bgamari> but it seems like that's what everyone does
<bgamari> surely this breaks on musl
taktoa has quit [Remote host closed the connection]
<Sonarpulse> bgamari: I do not
<bgamari> ool
<Sonarpulse> but matthew bauer or dtz I thought
<Sonarpulse> mentioned something similar in a bug recently
<Sonarpulse> about e.g. using more gnu i18n things with musl or similar
<Sonarpulse> certainly Matthew Bauer's meta-packages stuff would be a good approach if we had multiple providers
orivej has joined #nixos-dev
vcunat has quit [Ping timeout: 255 seconds]
Lisanna has quit [Quit: Lisanna]
Cale has quit [Ping timeout: 264 seconds]
zybell_ has quit [Ping timeout: 240 seconds]
MichaelRaskin has joined #nixos-dev
zybell has joined #nixos-dev
<copumpkin> is it possible to get an SSH-able initrd?
<copumpkin> with NixOS?
<copumpkin> like if I get a failure to mount the rootfs (on a remote instance without console access), I'd like to be SSH in and poke at it
<copumpkin> shlevy: you around? :)
globin has quit [Ping timeout: 260 seconds]
goibhniu has quit [Ping timeout: 265 seconds]
Synthetica has quit [Quit: Connection closed for inactivity]
globin has joined #nixos-dev
layus has quit [Quit: ZNC 1.6.5 - http://znc.in]
layus has joined #nixos-dev
globin has quit [Remote host closed the connection]
<obadz> is X possible with NixOS
<obadz> ⇒ yes
orivej has quit [Ping timeout: 260 seconds]
disasm has quit [Ping timeout: 260 seconds]
Sonarpulse has quit [Ping timeout: 268 seconds]