<rmcgibbo[m]>
obviously there's a lot i don't know here, but (a) it's a lot less expensive in terms of server time than i realized (b) for big PRs, we really only need to build a subset of the changed attrs for every PR. (some subset of) the direct-dependencies are sufficient to uncover most problems.
b42 has quit [Quit: WeeChat 2.9]
rajivr has joined #nixos-dev
<sterni>
which is probably why it's wise not to use hydra
<{^_^}>
starlink in ~40 minutes: https://www.youtube.com/watch?v=DpsrExXlQrA: Ping for space stuff (edit this command to add yourself, see ",help"): infinisil Taneb ldlework etu philipp[m] eyJhb gchristensen __red__ red red[evilred]
<hexa->
very red
<infinisil>
And they're live!
<samueldr>
(wrong channel?)
<infinisil>
samueldr: Nope?
<infinisil>
T-12
<sorear>
seems like more of a -chat than a -dev thing?
<infinisil>
Ohhh!
<infinisil>
Damn it, sorry!
jonringer has quit [Remote host closed the connection]
mkaito has quit [Quit: WeeChat 3.0.1]
<samueldr>
:)
<worldofpeace>
gchristensen: hope u already went through with the block. literally a troll
jonringer has joined #nixos-dev
jonringer has quit [Remote host closed the connection]
jonringer has joined #nixos-dev
cole-h has quit [Quit: Goodbye]
cole-h has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
LnL- is now known as LnL
cole-h has quit [Ping timeout: 260 seconds]
tom39291 has quit [Quit: WeeChat 2.9]
tom39291 has joined #nixos-dev
Jackneill has joined #nixos-dev
jonringer has quit [Ping timeout: 240 seconds]
BaughnLogBot has quit [Ping timeout: 245 seconds]
BaughnLogBot has joined #nixos-dev
<rnhmjoj>
is there something like udev.packages for installing polkit rules?
<rnhmjoj>
looks like environment.systemPackages should work
<supersandro2000>
building of '/nix/store/0whpr7n08fa15c30v2sv2pkizaxd3d0s-avr-stage-static-gcc-debug-10.2.0', '/nix/store/7fhf7j7f4w21kslj6pjqzb66cnn82xii-avr-stage-static-gcc-debug-10.2.0-man', '/nix/store/d2ma669gjn90m3f9mymh5mcapx241n9q-avr-stage-static-gcc-debug-10.2.0-info', '/nix/store/sv53x65711hbp4mwa0s4h5bghl61q1cs-avr-stage-static-gcc-debug-10.2.0-lib'
<supersandro2000>
killed after writing more than 67108864 bytes of log output
<gchristensen>
holy crap
<hexa->
on staging-next, is it preferred to do merges to fix conflicts or are rebases also fine
<hexa->
there is a redundant package update in there, that is already on master
<hexa->
so I'd usually just drop the redundant commit with a rebase
<hexa->
but obviously this involves force-pushing
<supersandro2000>
I would prefer a rebase personally because it would make the history easier to understand
<hexa->
yep, but a merge commit is more transparent
<supersandro2000>
I think the others usually do it with a merge
<symphorien[m]>
There are so many merges in staging next that I would be surprised if you could do a rebase at all
<hexa->
I did
<hexa->
the merge commits vanish during the rebase
<siraben>
supersandro2000: uh oh big log
evils has quit [Ping timeout: 240 seconds]
hplar is now known as aei
rj has quit [Ping timeout: 268 seconds]
aei is now known as hplar
mkaito has quit [Ping timeout: 260 seconds]
<qyliss>
hexa-: I'd rather staging-next wasn't rebased, because it's a branch other people might be working on
<qyliss>
it means we can't find which PR introduced a commit
<qyliss>
because its hash will have changed -- I don't think commit hashes should change once a PR is merged
<qyliss>
(in fact I'm a little surprised staging-next isn't a protected branch)
<lukegb>
yeah, git doesn't have good support for commit hashes changing; I wouldn't really rebase 'public' commits
<qyliss>
hexa-: how would you feel about having staging-next become a protected branch? Would make it so that there was a clear answer to this in future.
mkaito has joined #nixos-dev
mkaito has quit [Changing host]
mkaito has joined #nixos-dev
<hexa->
qyliss: I've only been involved in the staging-next cycle once or twice
<hexa->
having to go through a pull request while resolving a conflict sounds like job-creating measures
<lukegb>
that's not really a requirement of protected branches
<lukegb>
the main thing is denying force-pushes
<qyliss>
yeah I'm not suggesting require PRs
<qyliss>
just that we disable force pushes
<hexa->
oh ok
<hexa->
that would totally work for me
<hexa->
should clear that with fridh, vcunat and roberth though
<roberth>
I mostly keep an eye on staging-20.09, not master's staging branches. The only risk I see now is that it'd be harder to undo an accidental push that messes things up, so it'd be good to know which maintainers have the permissions to correct such a push
<cole-h>
I think you'd still be able to correct such a push, but only if you don't force-push the "accidental push" away and just fix it up
BaughnLogBot has quit [Ping timeout: 264 seconds]
BaughnLogBot has joined #nixos-dev
rj has joined #nixos-dev
rj has quit [Ping timeout: 268 seconds]
rj has joined #nixos-dev
rj has quit [Ping timeout: 268 seconds]
rj has joined #nixos-dev
Jackneill has joined #nixos-dev
<abathur>
domenkozar[m]: curious what you think about the cachix cache-name idea in nix#4577
<domenkozar[m]>
abathur: love it! will merge tomorrow
rajivr has quit [Quit: Connection closed for inactivity]
<abathur>
<3 domenkozar[m] ty :)
<{^_^}>
domenkozar[m]'s karma got increased to 38
<gchristensen>
does Nix's CLI expose enough primitives that I could implement my own garbage collector? like could I say "Nix, I promise this path is freeable -- don't check on your own, delete exactly this one path" ?
<{^_^}>
nix#1225 (by expipiplus1, 4 years ago, open): Should the daemon be used when NIX_BUILD_HOOK is set?
<{^_^}>
nix#1221 (by edolstra, 4 years ago, open): Remove the build hook mechanism
<catern>
btw, the replacement for NIX_BUILD_HOOK is "just write some C++ linking against libnix and implementing a new Store", right? has anyone done that?
rj has quit [Ping timeout: 268 seconds]
rj has joined #nixos-dev
__monty__ has quit [Quit: leaving]
<infinisil>
catern: I'd imagine it's `{pre,post}-build-hook` in nix.conf