<{^_^}>
#90275 (by IreneKnapp, 17 minutes ago, open): Help Thunderbird find libudev, which is in the systemd lib dir.
<samueldr>
one thing that can help is figure out who touched those derivations recently and cc-ing them in the PR
<samueldr>
looks like taku0 (on github) is the maintainer and actively maintaining
<samueldr>
I belive the bot would have requested a review anyway at some point
<Irenes[m]>
ah, thanks
<Irenes[m]>
I appreciate the help
<cole-h>
Irenes[m]: Slight nit: the commit title doesn't follow our CONTRIBUTING standards -- would you mind rewording it to `thunderbird: libudev is part of systemd` or something that similarly conveys intent?
<Irenes[m]>
oh! sorry, of course
<cole-h>
<3 Irenes[m] thank you (I love to nitpick, sorry :^) )
<{^_^}>
Irenes[m]'s karma got increased to 4
<Irenes[m]>
it's just part of doing it right, I'm happy to learn these conventions
<cole-h>
(But it's also how ofborg knows to build/test a certain package)
<Irenes[m]>
done
<cole-h>
Ah, sorry, I meant the actual commit title -- ofborg doesn't care about the PR title ^^
<Irenes[m]>
oh lol
<Irenes[m]>
got it
<Irenes[m]>
fixed
<cole-h>
Perfect; sorry, and thank you!
<Irenes[m]>
seriously, no need to apologize. I wish every project were scrupulous about details.
<samueldr>
some people take it badly to have their nits combed and picked :)
<Irenes[m]>
yeah, I hear that
<Irenes[m]>
programming is, like... there's a lot more emotions in it than you might naively think.
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-dev
drakonis_ has quit [Ping timeout: 260 seconds]
orivej has quit [Ping timeout: 264 seconds]
drakonis_ has joined #nixos-dev
{^_^} has quit [Excess Flood]
{^_^} has joined #nixos-dev
drakonis_ has quit [Ping timeout: 258 seconds]
drakonis_ has joined #nixos-dev
euank has joined #nixos-dev
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 256 seconds]
orivej_ has joined #nixos-dev
drakonis has quit [Quit: WeeChat 2.8]
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-dev
FRidh has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
cole-h has quit [Quit: Goodbye]
orivej has joined #nixos-dev
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-dev
orivej has quit [Quit: No Ping reply in 180 seconds.]
<FRidh>
nix-shell runs hooks that are meant for the build of a package. That will cause incompatibilities if you want to just use certain programs in a temporary environment, and the reason the experimental nix command has `run` and `dev-shell` commands
<Mic92>
I have not looked into dev-shell, but does it not conceptionally builds the same environement that nix-shell builds?
<FRidh>
yes, exactly
<FRidh>
if you *just* want programs, use `nix run`, if you want to have the environment to reproduce a build, use `nix dev-shell` or `nix-shell`
<Mic92>
I see it as a big advantage of nix that we reuse expressions for both building and development. I think we should stick to it.
<FRidh>
definitely. The issue is that you get more than you've asked for. That is, hooks and such typically (including portions of stdenv) you get even though you don't always want it
<FRidh>
example: SOURCE_DATE_EPOCH
<clever>
the QT hooks for example, mutate some things in the configrePhase
<FRidh>
or depending what you're doing the hook setting up PYTHONPATH and such
<clever>
and they are a mess to deal with if you want qtcreator to work
<clever>
i think it was generating a qmake shell script, based on $buildInputs
<clever>
and you must use that variant of qmake
<Mic92>
but it does not need to be this way. SOURCE_DATE_EPOCH can be easily fixed.
<clever>
ghc fixes the problems by doing a buildEnv over all packages, and re-generating the ghc-pkg index
<clever>
but then the nix expr has to be more involved in the dep management
<FRidh>
not fixed, worked around. You will still export the variable, causing a fixed date, while outside of a nix-build "where you just want to use tools like on another distro" you still don't expect that
<Mic92>
and we probably can fix the qt stuff as well.
<clever>
last time i touched it, you had to both launch qtcreator in nix-shell, and then edit the settings and specify the path to qmake
<FRidh>
clever: yes, that's what I think we need to do more. I would like to see a buildEnv that takes a mapping with functions that decide how a composition needs to be performed
<clever>
but the hooks generate a new qmake every time
<clever>
so you have to mess with qtcreator config, every time you re-launch the shell or wipe /tmp
<clever>
FRidh: i also find the haskell stuff a bit odd, you must shell into .env, not the main derivation
<clever>
FRidh: id like to see that cleaned up in the future, and not spreading to other frameworks
<FRidh>
I think the language-specific builders need to be more aware on how to deal with dependencies, and eventuall pass in only (native)buildinputs, and avoid propagatedBuildInputs.
<clever>
yeah
<clever>
i also see a lot of people mis-understanding propagatedBuildInputs, and think its for final runtime deps
<FRidh>
Still want to do that with python but it breaks everything :(
<FRidh>
yes I think the python builder is to blame for that
<clever>
python doesnt help, because it baked propagatedBuildInputs into the PYTHONPATH in wrappers, so it does actually become final runtime deps
<clever>
FRidh: this one really does become a runtime dep, and nix-env will silently install extra things you didnt ask for
<clever>
and if collisions occur, it silently picks one or the other
<clever>
and then QT programs fail
<clever>
ive even seen the entire desktop-manager fail immediately upon login, because an old library was in nix-env
<FRidh>
Ugh propagated-user-env-packages. It's a whole different approach to handling run-time dependencies.
<clever>
avoid it at all costs :P
<FRidh>
yep
<FRidh>
but still, its the way if you want to have things installable like with other distro's
orivej has quit [Ping timeout: 264 seconds]
orivej has joined #nixos-dev
__monty__ has joined #nixos-dev
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
FRidh has quit [Quit: Konversation terminated!]
orivej has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-dev
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-dev
<gchristensen>
eyJhb: I don't need to, because /etc doesn't exist on startup it just gets created
<gchristensen>
edef: ^
<edef>
ack
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 258 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 265 seconds]
orivej_ has joined #nixos-dev
orivej_ has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-dev
NinjaTrappeur has quit [Quit: WeeChat 2.8]
NinjaTrappeur has joined #nixos-dev
<eyJhb>
gchristensen: wait, what is that in relation to?
orivej has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-dev
<aszlig>
flokli: what do you mean with bad UX for cherry-picking merge commits?
<aszlig>
you basically just specify the parent-number and since we usually don't have too many octopus merges, the number is usually 1
<aszlig>
from reading the commit history however i personally find pointers for individual commits when you backport a whole branch much less readable than having a pointer from the backport-merge back to the original merge
<aszlig>
... since you need to go through them one-by-one instead of getting the bigger picture[TM] when just looking at the parents of the original merge
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-dev
nschoe has joined #nixos-dev
nschoe has quit [Ping timeout: 272 seconds]
orivej has quit [Ping timeout: 260 seconds]
orivej_ has joined #nixos-dev
nschoe has joined #nixos-dev
orivej_ has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-dev
bhipple has joined #nixos-dev
<flokli>
aszlig: you still need to manually specify the parent, as git doesn't know what was the PR, and what was origin/master before the merge.
<flokli>
you get to the PR very quickly, by clicking at the commit id in the "Cherry-picked-from" line, and then at the PR number.
<flokli>
and that's how it has been handled in nixpkgs mostly - so I'd keep it consistent with that.
<mdlayher>
infinisil: thanks again for the reviews! updated my PR to break up a long doc comment and ready for another look when you are: https://github.com/NixOS/nixpkgs/pull/89781
<{^_^}>
#89781 (by mdlayher, 6 days ago, open): nixos/corerad: add settings option to supersede configFile
<mdlayher>
infinisil: thanks so much! looking forward to converting my configs to the new model, it's super nice
<mdlayher>
i subscribed to the PR you had mentioned which would add a blessed Nix to TOML generator, so i'll update when that's ready
<infinisil>
Awesome :)
<mdlayher>
oh! i meant to ask: is there a precedent around setting up a warning message or similar in cases where a user specifies both settings and configFile?
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
orivej_ has joined #nixos-dev
<abathur>
clever FRidh re: mis-understanding propagatedBuildInputs: I find the "specifying dependencies" section of the nixpkgs manual impenetrable and slippery. I hesitate to assume my experience generalizes--but if it does, I have a hard time imagining many people developing a robust understanding without spending much time working below that API layer?
<clever>
abathur: ive also found that the example of how to set builder, misleads a lot of people into not using anything from the stdenv, complicating things
drakonis has joined #nixos-dev
drakonis_ has quit [Ping timeout: 256 seconds]
orivej_ has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
rajivr has quit [Quit: Connection closed for inactivity]
bhipple has quit [Ping timeout: 246 seconds]
drakonis_ has joined #nixos-dev
drakonis has quit [Ping timeout: 272 seconds]
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
bhipple has joined #nixos-dev
drakonis has joined #nixos-dev
justanotheruser has joined #nixos-dev
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-dev
FRidh has joined #nixos-dev
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-dev
cole-h has joined #nixos-dev
orivej_ has joined #nixos-dev
orivej has quit [Ping timeout: 264 seconds]
orivej_ has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-dev
<Ericson2314>
:D
justanotheruser has quit [Ping timeout: 272 seconds]
<{^_^}>
#21403 (by aszlig, 3 years ago, open): [WIP] NixOS integration of nixpart
nschoe has quit [Ping timeout: 272 seconds]
<aszlig>
flokli: well, i do beg to differ there, but i also have a different workflow that doesn't involve the github ui very much, so i'll change the pull request to include individual pointers