bennofs_ has quit [Read error: Connection reset by peer]
bennofs_ has joined #nixos-dev
jonringer has quit [Ping timeout: 276 seconds]
orivej has quit [Ping timeout: 240 seconds]
__monty__ has joined #nixos-dev
<eyJhb>
Didn't mention this yesterday. But I appreciate the discussion that you started yesterday lukegb :) Always nice with some perspective! lukegb++
<{^_^}>
lukegb's karma got increased to 26
pmy has quit [Read error: Connection reset by peer]
pmy has joined #nixos-dev
Synthetica has joined #nixos-dev
orivej has joined #nixos-dev
<eyJhb>
Maybe our OVI/Container tools should use a external tool, to do all the nice generation (if possible). Making JSON and fun .tar.gz seems like a overhead that would be nice not to have
pmy has quit [Ping timeout: 240 seconds]
pmy has joined #nixos-dev
init_6 has joined #nixos-dev
<dotlambda>
I just remembered that sudo vulnerability and thought: Why not add a NixOS option to make sudo owned by root:wheel with permissions 4550? Would anyone else like such an option and do you know if sudo supports gid != root?
init_6 has quit []
teto has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
teto has quit [Quit: WeeChat 3.1]
jonringer has joined #nixos-dev
orivej has joined #nixos-dev
cole-h has joined #nixos-dev
pmy has quit [Read error: Connection reset by peer]
pmy has joined #nixos-dev
pmy has quit [Read error: Connection reset by peer]
pmy has joined #nixos-dev
Mindavi has quit [Quit: leaving]
rj has joined #nixos-dev
<__monty__>
Maybe someone here has an idea? I'm stuck trying to troubleshoot why darwin-stdenv is referring to llvm 7.1.0, here's what little information I've managed to gather, https://git.io/J3VvA
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
justanotheruser has quit [Ping timeout: 250 seconds]
justanotheruser has joined #nixos-dev
sorear has quit [Ping timeout: 245 seconds]
vdemeester has quit [Ping timeout: 260 seconds]
<samueldr>
anyone with reservations against #119974, speak today... or don't?
<samueldr>
infinisil: probably *their* 14 days is rounded up
<samueldr>
and your 1 week is rounded down
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
<infinisil>
Ah yeah
<samueldr>
and it's a situation where neither are totally right nor wrong :) encoding data in a lossy way is all about trade-offs
orivej has quit [Ping timeout: 268 seconds]
rajivr has quit [Quit: Connection closed for inactivity]
justanotheruser has quit [Ping timeout: 250 seconds]
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
justanotheruser has joined #nixos-dev
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
<maralorn>
Can I, in pure nix, figure out if a derivation has a marked broken dependency or not?
<maralorn>
I am looking into doing something with tryEval, but I am not sure how to differentiate between broken and other errors.
<MichaelRaskin>
Hm. Maybe overriding meta.broken and retrying?
Synthetica has quit [Quit: Connection closed for inactivity]
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
__monty__ has quit [Quit: leaving]
rj has quit [Ping timeout: 240 seconds]
<maralorn>
I am more generally interested in configuring a hydra job in a way that it does not list eval errors caused by marked broken packages.
rj has joined #nixos-dev
Synthetica has joined #nixos-dev
<sterni>
you can pass allowBroken = true :p
<MichaelRaskin>
That sounds like an expensive way to clean up the eval page
<sterni>
yeah it is
<sterni>
a 10000 extra builds way to clean up the eval page in this case
<sterni>
maralorn: ime the eval page is very short if the eval error is worth your time
abathur has quit [Ping timeout: 252 seconds]
abathur has joined #nixos-dev
abathur has quit [Client Quit]
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
<gchristensen>
eval errors by packages marked as broken can be a critical clue to why something isn't showing up in the list anym,ore
<samueldr>
sounds like it's not for Nixpkgs use though
<gchristensen>
I've definitely used it on nixpkgs :)
<gchristensen>
the eval errors page could be come structured, though
<sterni>
gchristensen: yeah the curse of haskellPackages is that we have so many packages marked as broken that it is a bit of a lost cause either way
<samueldr>
maybe though it would be helpful for their use case if "asking for package a, but package a is broken" is not listed, but "asking for package b, but dependency a is broken" gets listed?
<sterni>
gchristensen: peti opted to mark everything as broken which had dependencies which were broken which kept the eval errors clean, so new problems are easily spotted; however it makes fixing a package a terrible chore because you have to search its reverse dependencies as well
<gchristensen>
ack
<sterni>
I think I'll just play around with it a bit in the next time to figure out what works
<gchristensen>
but really, the evaluation errors could become a structured json document pretty easily
supersandro2000 has quit [Killed (tolkien.freenode.net (Nickname regained by services))]
supersandro2000 has joined #nixos-dev
<rmcgibbo[m]>
I've just moved the r-rmcgibbo code into a public open-source repo at https://github.com/rmcgibbo/r-rmcgibbo, so if anyone would like to file bugs or anything, that will now be a good location.
<rmcgibbo[m]>
I haven't been looking too much at it recently -- it seems like it's fairly stable and not annoying anyone too much (or crashing too much), but if anyone feels differently please let me know. I was a bit busy with offline commitments so I may have missed some things.
rj has quit [Remote host closed the connection]
rj has joined #nixos-dev
<rmcgibbo[m]>
The next major thing would be moving "r-rmcgibbo" (bad name, lol) to some sort of nix-community configuration and changing the way it reports results to use the green check marks API, but that's going to require a pretty significant amount of delicate work.
rj has quit [Ping timeout: 240 seconds]
rj has joined #nixos-dev
<samueldr>
#121450 and #121720 are a good pair of PRs that could get reviews... but not sure who else can judge them :)