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 | 18.09 release managers: vcunat and samueldr | https://logs.nix.samueldr.com/nixos-dev
aanderse has joined #nixos-dev
goibhniu has quit [Ping timeout: 260 seconds]
sir_guy_carleton has joined #nixos-dev
<andi-> seems like we are having logging regarding the fetching(?) deadlocks now: https://hydra.nixos.org/build/82528778/nixlog/7/tail it is just sad that they seem hang anyway :/
lassulus_ has joined #nixos-dev
lassulus has quit [Ping timeout: 272 seconds]
lassulus_ is now known as lassulus
<gchristensen> that message isn't new
<gchristensen> andi-: this is a new and exciting deadlock :)
<gchristensen> oh, no, Ike didn't get the new nix. looking.
Ericson2314 has joined #nixos-dev
orivej has quit [Ping timeout: 252 seconds]
sir_guy_carleton has quit [Quit: WeeChat 2.0]
FRidh has joined #nixos-dev
Ericson2314 has quit [Ping timeout: 252 seconds]
sphalerite has quit [Quit: WeeChat 2.0]
sphalerite has joined #nixos-dev
sphalerite has quit [Quit: WeeChat 2.0]
sphalerite has joined #nixos-dev
sphalerite has quit [Quit: WeeChat 2.0]
sphalerite has joined #nixos-dev
init_6 has joined #nixos-dev
<manveru> time to rename it to 行けない...
orivej has joined #nixos-dev
FRidh has quit [Quit: Konversation terminated!]
init_6 has quit []
goibhniu has joined #nixos-dev
<gchristensen> train wifi isn't the best place to debug broken builders... but it is a fine place to start doing deploys -- the atomic update is perfect.
sir_guy_carleton has joined #nixos-dev
<ekleog> manveru: that pun was… :D
<manveru> was just celebrating that i finally switched from ibus+anthy to fcitx+mozc :)
<ekleog> mozc ftw
<ekleog> (though I'm ibus+mozc currently)
<gchristensen> what did that message say?
<ekleog> time to rename it to « it doesn't work » (which is pronounced « ikenai » in japanese)
<gchristensen> haha
phreedom has quit [Quit: No Ping reply in 180 seconds.]
phreedom has joined #nixos-dev
obadz has quit [Ping timeout: 252 seconds]
<gchristensen> it would be interested to see Nix's evaluator be able to run multiple builtin fetches at once
<andi-> more locking issues? ;-)
<gchristensen> I'm doing an eval with about 23 builins.fetchTarball's and it is ... a bit slow
<gchristensen> error: Nix daemon out of memory oops
<samueldr> #47901 has 5612 changed paths according to ofborg, but **only** aarch64; since they're on separate jobsets on hydra is it fine to merge without going through staging?
<{^_^}> https://github.com/NixOS/nixpkgs/pull/47901 (by dhess, 2 days ago, open): haskell: re-enable aarch64, but disable parallel builds on that arch.
<samueldr> mic92: how do you want to proceed for an update to alacritty? I see it's using your fork
<samueldr> confirming that by using the upstream 0.2.1 release my (only?) issue with alacritty has been fixed
<andi-> wasn't there talks about moving away from the fork?
edef has quit [Remote host closed the connection]
edef has joined #nixos-dev
<samueldr> there was talk recently about alacritty here, don't really remember what it was all about so I'm asking
<gchristensen> go for it, samueldr
<samueldr> good, was wondering if there was yet something I wouldn't foresee
<samueldr> dhess: think you can track the status of this in the coming hydra evals?
<samueldr> (I'll try to remember to check too)
<andi-> I am a bit confused about https://nixos.org/channels/nixos-18.03/git-revision, it shows as d16a7ab which is from 20th of september, the nixpkgs-channels repo points to the commits the IRC bot announces.. When I update the channel locally I also end up on the commit from the 20th september.
<andi-> same is visible on 18.09, nixpkgs-channels branch points to a commit from 3 days ago while the channel is from the 3rd october
* samueldr wonders if his fix for updates going back from the future actually worked
<andi-> two days ago the 18.03 channel on this machine was on a newer machine.. :/
<andi-> s/machine\./revision./
<samueldr> :)
<andi-> How does that process actually work? can an old evaluation cause a "new" channel bump with an old revision?
<samueldr> give me 2 secs
<andi-> k
<samueldr> or a bit longer
<andi-> take your time, but not too long :P
Ericson2314 has joined #nixos-dev
<{^_^}> nixos-channel-scripts#22 (by samueldr, 6 weeks ago, merged): Fixes channel going back in time due to incomplete change.
<samueldr> as vcunat said earlier today: seems to have happened with 18.09beta
<samueldr> which builtins.compareVersions should have handled right
<samueldr> so my fix might not have been enough, or something else I wasn't able to psychically diagnose happens
<{^_^}> #47592 (by Mic92, 6 days ago, merged): alacritty: 2018-08-30 -> 0.2.0 (backport)
<samueldr> 0.2.1 was released 4 days ago
<samueldr> :)
<Mic92> it is not using my fork so it can be updated as usual
<samueldr> I still see mic92/alacritty.git
<Mic92> samueldr: ah right, the fork-less version was in staging
<Mic92> because it required a fix in buildRustPackage
<samueldr> I mainly don't want to step on toes
<Mic92> 0.2.1 can be backported as well
<Mic92> it fixes a security issue
<samueldr> if you meant to 18.09, even without security issues, it would have been; "patch" level update with bug fixes are important
<samueldr> so uhm, I'm still unclear on how to proceed due to the other change being in staging; do we wait until staging is merged back into master? otherwise whatever your fork fixes will be lost for a small while
lassulus has quit [Ping timeout: 272 seconds]
<andi-> It seems not the script goes back in time but hydra... https://hydra.nixos.org/job/nixos/release-18.03/tested/latest-finished
<gchristensen> hydra does definitely, when jobs are restarted. the script should protect itself flom that though, and has code to do that
<andi-> ok
<gchristensen> ie, if a job for the latest finished jobset is restarted, it is no longer finished:)
<andi-> I just wasn't sure wether hydra pushes some hook or if there is a regular job interval on nixos.org that pulls. It seems like it pulls regulary on such a url for each release?
<andi-> back to perl then..
<{^_^}> nixos-channel-scripts#22 (by samueldr, 6 weeks ago, merged): Fixes channel going back in time due to incomplete change.
lassulus has joined #nixos-dev
orivej has quit [Ping timeout: 272 seconds]
orivej has joined #nixos-dev
lopsided98 has quit [Quit: Disconnected]
lopsided98 has joined #nixos-dev
lopsided98 has quit [Client Quit]
lopsided98 has joined #nixos-dev
<andi-> So the script does look okayish.. So the question is: Was it ever deployed? ;-)
<andi-> (the newer version)
<gchristensen> mmmaybe not
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-dev
pie_ has joined #nixos-dev
<samueldr> hmm, if it wasn't that would explain the behaviour not being fixed :/
Ericson2314 has quit [Ping timeout: 260 seconds]
<elvishjerricco> Is there someone better to get input from on this PR? https://github.com/NixOS/nix/pull/2452 I know eelco is generally quite busy.
<{^_^}> nix#2452 (by ElvishJerricco, 1 week ago, open): Fix overflow when verifying signatures of content addressable paths
obadz has joined #nixos-dev
obadz has quit [Ping timeout: 244 seconds]
Ericson2314 has joined #nixos-dev
<niksnut> elvishjerricco: I'll have a look tomorrow
<elvishjerricco> niksnut: oh thanks a bunch :)
obadz has joined #nixos-dev
obadz has quit [Ping timeout: 252 seconds]
obadz has joined #nixos-dev
obadz has quit [Quit: WeeChat 2.2]
pie_ has quit [Ping timeout: 260 seconds]
goibhniu has quit [Ping timeout: 252 seconds]