lukegb has quit [Read error: Connection reset by peer]
lukegb has joined #nixos-dev
orivej has joined #nixos-dev
justanotheruser has joined #nixos-dev
mmlb6 has joined #nixos-dev
mmlb has quit [Ping timeout: 246 seconds]
mmlb6 is now known as mmlb
<elvishjerricco>
Huh. Slight oddity I've just noticed. Because nixos runs the activation script at boot before the stage 2 mount units, it will create home directories in the /home directory of the / file system, even if /home is supposed to be on another file system.
<elvishjerricco>
Yea, just tested it. `No directory, logging in with HOME=/`
<elvishjerricco>
Seems like user management should be done after local-fs.target :/
<elvishjerricco>
(yet another reason that most activation scripts should actually be systemd units)
cole-h has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
FRidh has joined #nixos-dev
<FRidh>
Is there a way to get the amount of queued builds per architecture within a certain amount of time from hydra? I have the impression staging-20.09 is more active than it was at previous times.
<FRidh>
Darwin builders are not keeping up (they're all working again though)
pmy has quit [Quit: WeeChat 3.1]
pmy has joined #nixos-dev
typetetris has joined #nixos-dev
srk has joined #nixos-dev
evanjs has quit [Read error: Connection reset by peer]
evanjs has joined #nixos-dev
Jackneill has quit [Ping timeout: 260 seconds]
Jackneill has joined #nixos-dev
gleber has quit [Ping timeout: 248 seconds]
cbarrett has quit [Ping timeout: 248 seconds]
betawaffle has quit [Ping timeout: 248 seconds]
johanot has quit [Ping timeout: 248 seconds]
siraben has quit [Ping timeout: 248 seconds]
ky0ko1 has quit [Ping timeout: 248 seconds]
chvp has quit [Ping timeout: 248 seconds]
cbarrett has joined #nixos-dev
betawaffle has joined #nixos-dev
johanot has joined #nixos-dev
jonringer has quit [Ping timeout: 250 seconds]
__monty__ has joined #nixos-dev
ky0ko1 has joined #nixos-dev
gleber has joined #nixos-dev
siraben has joined #nixos-dev
teto has joined #nixos-dev
cole-h has quit [Ping timeout: 260 seconds]
chvp has joined #nixos-dev
devhell has joined #nixos-dev
Synthetica has joined #nixos-dev
tomberek has quit [Quit: Connection closed]
ScottHDev has joined #nixos-dev
ScottHDev has quit [Ping timeout: 252 seconds]
ScottHDev has joined #nixos-dev
<Taneb>
Test for home-manager in unstable seems to have been broken by updating mosquitto
<lukegb>
WORLDofPEACE: <3 thanks for everything you've done
justanotheruser has quit [Ping timeout: 260 seconds]
teto has quit [Quit: WeeChat 3.1]
justanotheruser has joined #nixos-dev
cole-h has joined #nixos-dev
NinjaTrappeur has quit [Ping timeout: 260 seconds]
rajivr has quit [Quit: Connection closed for inactivity]
avn has joined #nixos-dev
FRidh has quit [Ping timeout: 240 seconds]
typetetris has quit [Quit: Connection closed for inactivity]
FRidh has joined #nixos-dev
justanotheruser has quit [Ping timeout: 240 seconds]
NinjaTrappeur has joined #nixos-dev
confususs has joined #nixos-dev
confususs has quit [Quit: Connection closed]
confususs has joined #nixos-dev
confususs has quit [Killed (Sigyn (Spam is off topic on freenode.))]
confususs has joined #nixos-dev
justanotheruser has joined #nixos-dev
jonringer has quit [Remote host closed the connection]
confususs42 has joined #nixos-dev
confususs42 has quit [Client Quit]
confususs has quit [Ping timeout: 240 seconds]
jonringer has joined #nixos-dev
justanotheruser has quit [Read error: Connection reset by peer]
justanotheruser has joined #nixos-dev
justanotheruser has quit [Read error: Connection reset by peer]
FRidh has quit [Quit: Konversation terminated!]
orivej has joined #nixos-dev
<supersandro2000>
I am having some lost in translation problems with the post from WORLD of PEACE. Basically its a leaving because things changed and trying to do something new?
<gchristensen>
they realized their heart wasn't in it anymore
<supersandro2000>
oh, I think I now understand it a bit better :) thanks graham
<gchristensen>
:)
srk has quit [Ping timeout: 260 seconds]
<abathur>
may not make sense sans context, but wondering if it's feasible to be able to declare a ~dependency on a package's build environment at a specific phase such that: if part of the same evaluation, Nix copies the environment at that point and builds the depender after with access; if not in the same eval, the dependency is re-built up to the requested phase?
<supersandro2000>
abathur: could you make a derivation which stops at that point and disable everything after it?
<abathur>
I think it could be brute-forced by building an override, skipping what you don't want to run, doing whatever you're there to do, and basically making the new output whatever you wanted to output
<abathur>
my feasibility question is more like, is it possible to make that less wasteful if it was going to be done in nixpkgs on hydra, for example
orivej has quit [Ping timeout: 240 seconds]
<abathur>
although I guess it might also be generalizable as a performance improvement for real overrides?
orivej has joined #nixos-dev
<supersandro2000>
wouldn't that mean we would need to store intermediate steps and temporary build artifacts?
<abathur>
~briefly? at least as I imagine it'd be duplicating package A's temporary build dir and shell env to a new build dir at some point during its build, finishing the build, discarding package A's build dir, building package B, discarding B's build dir and the duplicate of A's
<abathur>
I guess it could do read-only access without the copy if you interrupt the build, but it's probably not great for debuggability if there's any chance a failure building B could keep A's build from finishing
<samueldr>
what's Nixpkgs' stance on adding "customized" linux kernels?
<supersandro2000>
Is xanmod pretty small? It only shows 35000 google results which may be an indication.
<supersandro2000>
I don't know what the official answer is but I wouldn't feel good if we would accept every little variant of the kernel
<samueldr>
I have no idea, hadn't heard about xanmod before that
<samueldr>
but not having heard of something is not really a good metric
<samueldr>
in the past I have been stating that we shouldn't do things like add SBC-specific kernel trees to Nixpkgs
<samueldr>
which, hard to say, might have had more end-users than other linux kernel customization projects?
cjb has joined #nixos-dev
__monty__ has quit [Quit: leaving]
orivej has quit [Ping timeout: 268 seconds]
evils has joined #nixos-dev
tokudan has quit [Remote host closed the connection]
Synthetica has quit [Quit: Connection closed for inactivity]
tokudan has joined #nixos-dev
evils has quit [Quit: leaving]
rj has quit [Quit: rj]
evils has joined #nixos-dev
<infinisil>
abathur: Oh, actually a problem with that is that in order for derivation B to reuse something from derivation A, that something needs to be stored in the Nix store
<infinisil>
At least for normal Nix rules
<infinisil>
Ohh
<infinisil>
But yeah, if derivation A stores this thing in a separate output, and B itself then doesn't have a runtime dependency on that output, it won't be propagated down
<infinisil>
So only those who actually need to build B itself need to have those artifacts in their store
supersandro2000 has quit [Quit: Ping timeout (120 seconds)]
<abathur>
I'm not sure, but I imagined it as a bit of an exception to normal rules
<abathur>
to be less cryptic for others reading now or later, I'm curious about something that could support taking a package, doing some deterministic ~static analysis, and producing a cacheable output that other tools could use without having to re-run the analysis every time
<abathur>
this isn't a big deal if it's analysis on the binaries, but if it needs something like the source, intermediate artifacts in the build dir, binaries pre-strip, etc., there may be a substantive cost to re-running large fractions of a long build
<abathur>
it could certainly work by pushing build context into the store, and that'd make the analysis faster to hack on, and that might even be a great debug option, but it at least feels a little wasteful to hold onto unpacked copies and potentially intermediate artifacts as part of regular order?
supersandro2000 has joined #nixos-dev
<abathur>
though if they didn't end up in the cache, it would be a smaller concern I guess? I realized as typing the above that I'm assuming all derivations resulting from an eval end up in the cache
supersandro2000 has quit [Disconnected by services]