<slack1256>
Now I wrote a long text about my shyness of IRC, i bet you are glad. ;_; /s
ddellacosta has quit [Ping timeout: 240 seconds]
<simpson>
rednaZ[m]: I don't think that you're very far off of what should work for pure-Python applications. The main thing to keep in mind is that you'll need to create an actual virtual environment and call the pip from inside there, rather than using a nixpkgs pip with ad-hoc flags.
<tpw_rules>
however qyliss is right, it's best for your experience and a good contribution to the community to write a derivation for it
<rednaZ[m]>
tpw_rules: How come there are no derivations autogenerated for all PyPI packages?
<simpson>
rednaZ[m]: Short answer is that (unlike e.g. Hackage) there's no PyPI metadata that is quite rich enough for that. We've talked with upstream PyPA before and it's non-trivial.
jimmiehansson has quit [Ping timeout: 264 seconds]
<simpson>
And that's only considering pure-Python dependencies. Folks tend to desire native dependencies and extension modules as well.
<rednaZ[m]>
Another questions: I installed pancritic successfully using `pip install --user pancritic`. Of course, I still cannot run it but why does `pip list` not list it?
<simpson>
You *need* to create an actual virtual environment. https://bpa.st/WQ5A
mkaito has quit [Quit: WeeChat 3.0]
Hail_Spacecake has joined #nixos
<rednaZ[m]>
simpson: Thanks. This had worked and I am celebrating but I am just wondering how I purge the leftovers from `pip install --user` from my system.
<rednaZ[m]>
...from before the virtual environment.
shibboleth has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @SuperSandro2000 merged pull request #108046 → openldap: fix build if openssl or cyrus_sasl are overridden to null, add flag for cyrus_sasl, require openssl → https://github.com/NixOS/nixpkgs/pull/108046
<colemickens>
aleph-: what do you mean by importing them?
<colemickens>
aleph-: normally they're on a server and then nix fetches them when it tries to build a certain derivation
<colemickens>
are you trying to bulk import/export or?
stree has joined #nixos
<aleph->
colemickens: So I made a bulk export of the 20.09 channel earlier via https://paste.rs/Q4s trying to set up mostly offline installations for my network.
<colemickens>
Can you clarify what you're trying to do? Do you want to import everything on every machine?
<aleph->
Nope, not everything on every machine. Just have a new machine when spun up able to just import what it needs from my local pkg cache first.
supersandro2000 has quit [Ping timeout: 264 seconds]
mbrgm_ has joined #nixos
rotaerk has quit [Ping timeout: 265 seconds]
<colemickens>
aleph-: in that case yeah, you might want to just import directly to your cache machine's store and then use nix-serve or something like that to make that store available to the rest of your internal machines
<colemickens>
(just to be extra verbose) It might be more efficient to straight mirror the nar/narinfos into a dir and then serve that up via nginx, compared to nix-serve, you'll pull nar, unpack, and then nix-serve would repack on the fly.
ddellacosta has joined #nixos
m0rphism1 has quit [Ping timeout: 246 seconds]
lambda-11235 has joined #nixos
wouter1 has quit [Ping timeout: 258 seconds]
slack1256 has quit [Remote host closed the connection]
<ajs124>
hexa-: re python2/3 (because I don't want to spam #nixos-security). I was trying to figure out if my systems depend on python2, but adding disallowedRequisites = [ pkgs.python2 ]; to baseSystem in nixos/modules/system/activation/top-level.nix doesn't seem to work.
<ajs124>
any better ideas how to determine that? I mean, I guess I could run closureInfo on everything and grep through that, but that output format is kind of weird
<hexa->
nix-store -q --tree $(nix-instantiate '<nixos/nixos>' -A system)
<hexa->
subversion …
<hexa->
neovim
wouter1 has joined #nixos
<hexa->
gvfs
<hexa->
opengl-drivers
<aleph->
colemickens: Planning to have my store over SSH.
<ajs124>
gimp, apparently. hm, those aren't just runtime dependencies though, right?
<aleph->
So I assume I'd need it imported into the binary store
<hexa->
I see setuptools, pip and wheel alot
<hexa->
intel-graphics-compiler, steam-run
<simpson>
ajs124, hexa-: There is of course a Python 2.7 flavor of PyPy which will still receive security updates. (Not relevant to the other channel because it doesn't look like this bug affects PyPy.) It would be prudent of us to determine what our long-term strategy will be, since PyPy will still need a Python 2.7 bootstrap in the future too.
<hexa->
02:14 <ajs124> but yeah, maybe we should just switch to the gentoo fork, instead of using their patches, because what else are we gonna do?
jumper149 has joined #nixos
<hexa->
but honestly, I don't want to put much time into this
<hexa->
it needs to be low effort or not at all
nhs has quit [Ping timeout: 264 seconds]
jumper149 has quit [Quit: WeeChat 2.9]
ddellacosta has quit [Remote host closed the connection]
sigmundv__ has quit [Read error: Connection reset by peer]
ddellacosta has joined #nixos
ddellacosta has quit [Remote host closed the connection]
ddellacosta has joined #nixos
ddellacosta has quit [Remote host closed the connection]
marcusr has quit [Remote host closed the connection]
n-osborn_ has joined #nixos
n-osborne has quit [Read error: Connection reset by peer]
marcusr has joined #nixos
ddellacosta has joined #nixos
ddellacosta has quit [Ping timeout: 240 seconds]
gustavderdrache has joined #nixos
Lord_of_Life has quit [Ping timeout: 256 seconds]
<colemickens>
mount.nfs -o vers=4 works, but vers=3 doesn't. I don't see us disabling v3 anywhere though/
Lord_of_Life has joined #nixos
nhs has joined #nixos
randomuser123 has joined #nixos
poscat has quit [Remote host closed the connection]
h0m1 has quit [Ping timeout: 272 seconds]
h0m1 has joined #nixos
poscat has joined #nixos
ahmedelgabri has joined #nixos
ddellacosta has joined #nixos
andreas303 has quit [Remote host closed the connection]
andreas303 has joined #nixos
<randomuser123>
Can I get a bit of help with my setup? I've got an imperatively-created user account, with home-manager managing my bash profile (programs.bash.enable = true;) and environment variables (home.sessionVariables.EDITOR = "nvim";). When I open a terminal (xfce4-terminal) my env variable isn't set. It *is* set properly, however, if I login to TTY1. What
<randomuser123>
am I missing / how do I get it load my profile everywhere?
andreas303 has quit [Remote host closed the connection]
andreas303 has joined #nixos
SuperSandro2000 has quit [Quit: Ping timeout (120 seconds)]
<randomuser123>
found a setting for xfce4-terminal to make it run it as a login shell. That works I suppose. Now to see if I can fit that into configuration.nix....
<CMCDragonkai1>
The laptop was certified for Ubuntu usage, and during the installation, Ubuntu is even aware that it is a Dell computer
andreas303 has joined #nixos
nhs has quit [Ping timeout: 256 seconds]
ddellacosta has quit [Remote host closed the connection]
zn34 has quit [Quit: Connection closed]
<colemickens>
is there like a nomodeset option on the iso that could be used? could you use the text installer and then be able to iterate in-place easier?
<CMCDragonkai1>
the minimal iso works
<CMCDragonkai1>
but then we want X server to work eventualy
<CMCDragonkai1>
done the full installation
nhs has joined #nixos
<colemickens>
CMCDragonkai1: of course! You might find it easier to SSH in and then be able to iterate easier depending on how bad the tty gets jammed.
<colemickens>
sadly, I'm not super great at debugging X, I haven't had to do that in ages. I'm not even sure what the usual candidates are.
<colemickens>
I am always a bit surprised when I hear about the Ubuntu ISO working in scenarios and the NixOS one not.
<colemickens>
Hopefully we can figure out what it is and improve the installer iso.
<colemickens>
looks like it would just be integrated graphics too, doesn't seem like it ought to be anything weird
nhs has quit [Ping timeout: 240 seconds]
<CMCDragonkai1>
the laptop has intel UHD
<colemickens>
I assume this is a recently-downloaded ISO, CMCDragonkai1 ? Are you trying a stable or nixos-unstable ISO?
<CMCDragonkai1>
and a nvidia gpu
<CMCDragonkai1>
yes
<CMCDragonkai1>
just downloaded
<CMCDragonkai1>
from the frontpage
<colemickens>
oh, it does have a nvidia gpu.
<colemickens>
I wonder if we can blacklist that for now or something.
<colemickens>
A quick google indicates ubuntu might ship the nvidia drivers in their iso, and I don't think we do that.
<veleiro>
i had an old nvidia gpu on a work computer when i first started using nixos
<veleiro>
it wasnt fun
<veleiro>
gnome and plasma are both wayland updated now, im pretty sure
sangoma has joined #nixos
<veleiro>
albeit terribly
nhs has joined #nixos
domogled has quit [Quit: domogled]
<colemickens>
CMCDragonkai1: can you try the iso with the nomodeset boot entry and see if iit helps?
<colemickens>
CMCDragonkai1: or actually, if you have the system installed, I guess you could go ahead and try to use the nvidia driver, if that was your end goal
<veleiro>
maybe its the fault of the displayManager
<veleiro>
CMCDragonkai1: you didn't specify which NVIDIA gen gpu?
<veleiro>
IMO its likely if it is supported by nouveau any NixOS install image would work
<CMCDragonkai1>
its the dell vostro 5402
<CMCDragonkai1>
i wanted to just use the intel UHD graphics
<CMCDragonkai1>
at the very least
<CMCDragonkai1>
not bother activating the GPU
<CMCDragonkai1>
what does nomodeset do?
<CMCDragonkai1>
right now i want step 1 working
<CMCDragonkai1>
which is just a X desktop
<CMCDragonkai1>
step 2 might be NVIDIA PGU
nhs has quit [Ping timeout: 272 seconds]
<veleiro>
ohh, its switchable
<CMCDragonkai1>
it is geforce mx330
<CMCDragonkai1>
but atm
<CMCDragonkai1>
its definitely running on ubuntu without nvidia i think
<CMCDragonkai1>
just intel UHD graphics
nhs has joined #nixos
<colemickens>
idk what nomodeset does, but there are a lot of hits for 'nvidia nomodeset nixos black screen' so it felt like a safe bet to try
<veleiro>
blacklisting nouveau like colemickens would not work i think because its a newer Nvidia GPU generation, that isnt supported by nouveau anyways
tomberek has joined #nixos
<colemickens>
I don't think they have an install though veleiro
<veleiro>
don't you have to 'install' nixos, before you have any graphical interface?
<colemickens>
the internet says that card is a desktop 1030...
<colemickens>
which nvidia probably supports, as it supports my desktop 1080.
<CMCDragonkai1>
veleiro: since like the beginning of time
<colemickens>
I don't think taking a second to blacklist nouvaeu is a waste of time, imo.
<CMCDragonkai1>
veleiro: nixos has had a graphical installer for years
domogled has quit [Quit: domogled]
<veleiro>
that card is a mx330 is what i see
<colemickens>
"The Nvidia GeForce MX330 is a dedicated entry-level mobile graphics card for laptops. It was released in February 2020 and is basically a renamed Geforce MX250 / GeForce MX150 / desktop GeForce GT 1030."
rauno has quit [Remote host closed the connection]
rauno has joined #nixos
ddellacosta has quit [Ping timeout: 272 seconds]
nhs has joined #nixos
solarliner has joined #nixos
respawn has quit [Quit: Leaving]
nhs has quit [Ping timeout: 272 seconds]
<solarliner>
Hi, I'm pretty sure by now it's a known issue, but just in case I wanted to report out that the NixOS cache server is unresponsive for some URLs http://ix.io/2PVU
bdju has quit [Ping timeout: 260 seconds]
h0m1 has quit [Quit: WeeChat 3.0.1]
h0m1 has joined #nixos
b00n has joined #nixos
<{^_^}>
[nix] @regnat pushed to ca/copy-drv-outputs « Let copyPath on RealisedPaths wrap copyPath on StorePaths »: https://git.io/JtyBN
solarliner has quit [Quit: solarliner]
solarliner has joined #nixos
m0rpheus has joined #nixos
nhs has joined #nixos
bdju has joined #nixos
ltl3 has joined #nixos
stree has quit [Ping timeout: 260 seconds]
ManiacOfMadness` has quit [Ping timeout: 265 seconds]
nhs has quit [Ping timeout: 265 seconds]
m0rpheus_ has joined #nixos
ltl3 has left #nixos [#nixos]
ddellacosta has joined #nixos
m0rpheus_ has quit [Client Quit]
lambda-11235 has quit [Quit: Bye]
m0rpheus has quit [Ping timeout: 240 seconds]
vidbina_ has quit [Ping timeout: 246 seconds]
ddellacosta has quit [Ping timeout: 240 seconds]
<b00n>
Hi there dear people (and others), I struggle a bit by trying to set up apparmor on NixOS and hope, someone here could please give me a helping hand. I enabled: 'security.apparmor.enable = true;' and the default profiles: 'security.apparmor.profiles = [ "${pkgs.apparmor-profiles}/etc/apparmor.d" ];' and 'sudo aa-status' tells me, that: '54 profiles are loaded' (and in enforce mode). So it's looking good, but
<b00n>
I fear it doesn't work. How can I be sure to execute my apps/processes through apparmor? I restricted e. g. 'mpv' to taking screenshots for testing reasons, but screenshots working while the modified mpv profile is loaded and my system has rebooted. Does somebody collected experiences with apparmor under Nix?
<{^_^}>
[nixpkgs] @peti pushed 1000 commits to haskell-updates: https://git.io/Jtyaq
nhs has joined #nixos
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/JtyaO
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage2nix: update list of broken builds to avoid evaluation errors on Hydra »: https://git.io/Jtya3
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/Jtyas
<avaq>
Hi folks! I have a question about nix channels and how they play with NixOps. Firstly, it seems that the nix channels available to a nixops-managed system are independent of the channel configured in the NixOps deployment config, is that right?
<avaq>
So I've modified my deployment config to use nixos-20.09, but on the target machine, the channel listed is still nixos-20.03.
<adisbladis>
avaq: NixOps has no channel management built in
lord| has joined #nixos
<adisbladis>
It inherits whatever you have on your system in NIX_PATH
<avaq>
Is there a way to manage nix channels via Nix? So I want to deploy to a target machine, and tell it to adopt a new channel, ideally as specified in its nix config.
<b00n>
If still somebody have (functional) experiences with apparmor under nixos, it would be nice to get in touch with you :)
nhs has joined #nixos
attila_lendvai has joined #nixos
<avaq>
So that after deployment, the channel used by 'nix-shell' and whatnot on that machine is going to be whatever I've configured in the nix expression for that machine.
<adisbladis>
avaq: You can set `nix.nixPath` to achieve that
<adisbladis>
Note that it won't affect the channel used for the deployment, just the channel on the remote once already deployed (it's setting an environment variable)
<avaq>
adisbladis, so once I override the nixPath, I guess using 'nix-channel' won't have any effects anymore?
<adisbladis>
Not really
<adisbladis>
It will still do it's thing, you just won't have a NIX_PATH pointing to whatever mutable location nix-channel manages
<adisbladis>
Tbh I'm not too knowledgeable about nix-channel as I never use it myself
<avaq>
Okay, that seems exactly as I expect it. Thanks. :)
cinimod has joined #nixos
orivej has quit [Ping timeout: 240 seconds]
lord| has quit [Ping timeout: 256 seconds]
lord| has joined #nixos
dev_mohe has quit [Quit: dev_mohe]
<avaq>
And which part is responsible for making the nix-channel binary available on nixops-deployed systems? If possible I might as well remove that when I'm managing nixPath directly, no?
<adisbladis>
It's always bundled in the nix package
<avaq>
Oh well
* etu
have set nix.nixPath to use nix-channels :p
ahmed_elgabri has joined #nixos
terrorjack has joined #nixos
nhs has joined #nixos
df41 has quit [Quit: Connection closed]
* etu
really wanted to change something else without stop using channels in the NIX_PATH :)
ahmedelgabri has quit [Ping timeout: 260 seconds]
<adisbladis>
etu: When I'm talking about channels I'm talking about `nix-channel` & the channel tarballs releases.nixos.org
<adisbladis>
Not the git branches
respawn has joined #nixos
nhs has quit [Ping timeout: 260 seconds]
openhyena has joined #nixos
avaq has quit [Quit: Leaving]
avaq has joined #nixos
avaq has quit [Remote host closed the connection]
terrorjack has quit [Remote host closed the connection]
<openhyena>
Is there a simple method of adding this bin folder to my path declaratively?
<openhyena>
It works except that I am informed ` WARNING: The scripts qcfractal-dashboard, qcfractal-manager and qcfractal-server are installed in '/nix/store/ddhw332lvji5d1lh516yjibnyrnyhx8k-python3.8-qcfractal-0.11.0/bin' which is not on PATH.`
terrorjack has joined #nixos
avaq has joined #nixos
nhs has joined #nixos
tomberek has quit [Quit: Connection closed]
nhs has quit [Ping timeout: 264 seconds]
zupo has joined #nixos
avaq has quit [Remote host closed the connection]
avaq has joined #nixos
<etu>
adisbladis: I know, on some of my systems I've set NIX_PATH to something like this: nixpkgs=/nix/var/nix/profiles/per-user/root/channels/nixos nixos-config=/etc/nixos/hosts/vps04/configuration.nix
<etu>
adisbladis: Then I still use nix-channel ;)
terrorjack has quit [Remote host closed the connection]
terrorjack has joined #nixos
zupo has quit [Ping timeout: 264 seconds]
nhs has joined #nixos
terrorjack has quit [Remote host closed the connection]
terrorjack has joined #nixos
nhs has quit [Ping timeout: 264 seconds]
<{^_^}>
[nixpkgs] @adisbladis opened pull request #113680 → top-level/packages-config.nix: Correctly index package subsets delimited by dots → https://github.com/NixOS/nixpkgs/pull/113680
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage2nix: update list of broken builds to avoid evaluation errors on Hydra »: https://git.io/Jty66
Izorkin has joined #nixos
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/Jty6i
<jansc>
sorry for my dumb question: what is the command to install "pkgs/applications/networking/ncgopher/default.nix" (not even sure if this is the package name) on NixOS? Disclaimer: I'm updating some docs for ncgopher and have no idea about nixos
stree has joined #nixos
<sterni>
jansc: nix-env -iA nixos.ncgopher
<sterni>
jansc: note on non-nixos with nix installed it is nix-env -iA nixpkgs.ncgopher
hyper_ch has joined #nixos
<sterni>
(not sure why this distinction exists in the first place but here we are)
<__monty__>
sterni: The channel names are entirely arbitrary. Traditionally they get a name that hints at the actual channel behind it. nixos-20.03 and nixpkgs-unstable, respectively, in this case, probably (because my nixos is still nixos-19.09 for example).
zupo has quit [Ping timeout: 260 seconds]
cinimod` has joined #nixos
euandreh has joined #nixos
n-osborne has quit [Remote host closed the connection]
<elux>
the other day, Go 1.16 was released, and i'd really like to use it on my Nix system. Any tips / suggestions how I can do that ahead of it landing in the home-manager or nixos-unstable channels (which I am hooked up to)
zebrag has joined #nixos
<__monty__>
elux: An overlay will probably be easiest. Let me see if I have an example.
<adisbladis>
You don't need an overlay
<adisbladis>
Go 1.16 is in nixpkgs already
<adisbladis>
> go_1_16
<{^_^}>
undefined variable 'go_1_16' at (string):477:1
<SuperSandro2000>
less bloat to download if you only need ldd
<sterni>
where the latter wouldn't exist because glibc's out output is not where the binaries are
<sterni>
slby[m]: ^
ManiacOfMadness` has joined #nixos
<sterni>
SuperSandro2000: well, it is also necessary to use if you want to get the output with binaries and write generic / future-proof nix expressions
<jschievink>
hey folks, I'm trying to get rootless podman working, but it's failing with "lchown /etc/shadow: invalid argument" and complaining about insufficient UIDs/GIDs, despite /etc/{subuid,subgid} being set up correctly
ddellacosta has quit [Ping timeout: 272 seconds]
cinimod`` has joined #nixos
ddellacosta has joined #nixos
<jschievink>
it is printing a few warnings about not having a systemd user session, presumably because I run via `sudo -u unprivileged-user` from root?
<jschievink>
but that doesn't look related to me
<adisbladis>
jschievink: I think it may be related since podman assumes systemd as the cgroup manager
<jschievink>
hmm, okay
<adisbladis>
And if you su to another user that user doesn't have any systemd user session running managing the cgroups
<adisbladis>
You might try `podman --cgroup-manager=cgroupfs`
<jschievink>
so is the only way to do this by logging in as the unprivileged user?
cinimod` has quit [Ping timeout: 240 seconds]
shibboleth has joined #nixos
<jschievink>
passing that doesn't seem to do anything
<adisbladis>
Hm.. I don't remember how I got this working in our tests...
<adisbladis>
Ahh, right
<jschievink>
even the systemd warnings stay, so it almost looks like `--cgroup-manager=cgroupfs` is completely ignored
<betaboon>
a host-configuration pulls in pipewire and i have no clue why. the tree just tells me it is being pulled in due to `system-path.drv`. any hints ?
Acou_Bass has quit [Ping timeout: 272 seconds]
<srhb>
betaboon: It's difficult if it's somehow in systemPackages and not in a systemd unit. I'd probably overlay it with a throw and try to trace the error.
<betaboon>
i have it nowhere in my systemPackages. I'll try the overlay+throw idea
stree has quit [Ping timeout: 240 seconds]
<srhb>
betaboon: Just checking, did you try nix why-depends yet?
<srhb>
betaboon: Because if it's not directly in systemPackages, you may get very friendly results from that
<srhb>
betaboon: As in nix why-depends /run/current-system /nix/store/...-pipewire
<slby[m]>
Are there somehwere good examples of how to customize the pkgs.formats.yaml function? I can't seem to escape quotes.
<jschievink>
I rebooted and now rootless podman seems to work fine
<jschievink>
¯\_(ツ)_/¯
<betaboon>
srhb: found it using the throw method. turns out `services.pipewire.media-session.enable` defaults to true since the PR has been merged
davidv7 has quit [Remote host closed the connection]
ambroisie has joined #nixos
n-osborne has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @sternenseemann opened pull request #113696 → ocamlPackages.janeStreet{,_0_9_0}: join the ocamlPackages fix point, allowing overriding to work as expected → https://github.com/NixOS/nixpkgs/pull/113696
<{^_^}>
#97310 (by berbiche, 23 weeks ago, closed): Newly introduced YAML formatter does not allow hexadecimal escapes (or other escapes) in strings
n-osborne has joined #nixos
<slby[m]>
infinisil: Thank you, that could be the source of my problem!
<slby[m]>
adisbladis: I need to have the `""` surrounding the `{}` sadly and I can't seem to escape the `""` for the value. If I try to escape it, I get from `\"` a `\\\"`
walkah has quit [Quit: Connection closed for inactivity]
<infinisil>
Certainly looks wrong to have a structured set within a string within a structured set..
<clever>
yaml is a superset of json, so any yaml parser will also accept json
<tpw_rules>
wait what, don't you have to reindent it?
<clever>
correct, you can just throw toJSON output right into a yaml parser
<clever>
and it just works
<infinisil>
clever: Yeah that's what pkgs.yaml does underneath
<tpw_rules>
how does the indenting work
<adisbladis>
infinisil: Btw, shouldn't generators have preferLocalBuild
<clever>
tpw_rules: indenting is just to let you skip using { and }
<tpw_rules>
oh weird
<slby[m]>
infinisil: I think thats what I started with, that gets me the escape characters also in the generated yaml. Maybe theres something really simple I am overlooking.
<clever>
if you use {} then you dont have to indent
<{^_^}>
[nixpkgs] @SuperSandro2000 merged pull request #112276 → treewide: makeWrapper buildInputs to nativeBuildInputs, makeWrapper: use proper shell when cross-compiling → https://github.com/NixOS/nixpkgs/pull/112276
<{^_^}>
nix#4442 (by domenkozar, 5 weeks ago, open): allowSubstitutes/preferLocalBuild are an anti-feature
<{^_^}>
[nixpkgs] @jtojnar pushed 64 commits to staging-next: https://git.io/Jty5X
<infinisil>
slby[m]: Are you sure?
<infinisil>
Because Nix doesn't pass those escaping chars along
<infinisil>
But they might be escaped again in the represented yaml
<infinisil>
Wait
<adisbladis>
infinisil: Well... I'm on a very high latency connection and it takes several seconds to realise a generator derivation because of the cache round trips
<infinisil>
Yeha no it should work
<adisbladis>
Also that issue is not concluded yet
<adisbladis>
I disagree with preferLocalBuild being an anti feature
shivanatm[m] has quit [Ping timeout: 240 seconds]
phr3nzy[m] has quit [Ping timeout: 240 seconds]
rmcgibbo[m] has quit [Ping timeout: 240 seconds]
bendlas has quit [Ping timeout: 240 seconds]
bsima[m] has quit [Ping timeout: 240 seconds]
fadenb[m] has quit [Ping timeout: 240 seconds]
timokau[m] has quit [Ping timeout: 240 seconds]
emily has quit [Ping timeout: 240 seconds]
<infinisil>
slby[m]: So what's the result you get and what is wrong about it?
<g-w1>
anyone here use emacs with nix and have a way to use nix-shell environments from emacs? I am having trouble with nix-mode and nix-buffer. thanks
<adisbladis>
Pretty much, the popular solution is direnv + one of the two emacs direnv integrations
sangoma has quit [Ping timeout: 260 seconds]
butcher has joined #nixos
avaq has quit [Ping timeout: 265 seconds]
zebrag has quit [Quit: Konversation terminated!]
cseder has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
zebrag has joined #nixos
jonringer has joined #nixos
ahmed_elgabri has joined #nixos
oida has joined #nixos
veleiro has joined #nixos
<slby[m]>
infinisil: I finally found out what happened! There was a rogue `}` in my configuration from experimenting. Somehow the Program still was fine with it but the args were empty `[]` for this particular setting.
<fuzzypixelz>
why do people use nix on alpine in particular?
<fuzzypixelz>
like in the docker image mini tutorial on nixos.org
<g-w1>
adispladis, is there a way to just source a .nix file from emacs? I have a lot of .nix files for my environments and would rather not add the complexity of direnv
leo60228 has quit [Ping timeout: 246 seconds]
revtintin has quit [Quit: WeeChat 1.9.1]
malook has joined #nixos
terrorjack has quit [Read error: Connection reset by peer]
hlz has quit [Quit: WeeChat 3.0]
<g-w1>
how can i use M-x nix-shell (from nix-mode)? I gave it a path to the .nix file but then it asks for a Nix Attr
hlz has joined #nixos
fuzzypixelz has quit [Quit: Connection closed]
<sterni>
slby[m]: every configuration module is passed the config argument if you mean that
<slby[m]>
sterni: Ah, yeah that was very vague, sorry. That's not what I meant.
<slby[m]>
I have a config for a program that I want to pass the program, but I also want the program to execute itself with a different argument but the same configuration.
<slby[m]>
So the configuration is in a way recursive. Because I would need the path of the configuration in the configuration.
<{^_^}>
[nix] @regnat pushed to ca/copy-drv-outputs « Ensure that the ca-derivations bit is set when copying realisations »: https://git.io/JtybC
<sterni>
slby[m]: how do you build the derivation?
ddellacosta has quit [Remote host closed the connection]
<sterni>
non fixed-output derivations can be self-referntial: lib.fix (self: pkgs.writeText "my-config" "path to myself: ${self}")
<slby[m]>
sterni: I was planning on using the pkgs.formats.yaml generator.
<g-w1>
what is the easiest way to reconfigure dynamic binaries (to use the nix dyld path) on the cli?
<lambda-11235>
I am having trouble with patchELF. It seems to leaving the loader as /lib64/ld-linux-x86-64.so.2. Any ideas on what could be going wrong?
<drozdziak1>
How can I eval the borgbackup module? I'm getting the value with unresolved assertions, I need a way to call `lib/eval-config.nix` on an example borg config
<attila_lendvai>
how come that fetchFromGitHub doesn't fail for me when i change the hash to something wrong? could it be that it's fetching the sources from cache.nixos.org and skips the hash check? if so, can i force it somehow?
nisstyre has joined #nixos
<{^_^}>
[nixpkgs] @veprbl pushed to master « rivet: add sansmath to fix make-plots »: https://git.io/JtSJX
sbock has joined #nixos
orivej has quit [Ping timeout: 261 seconds]
<exarkun>
attila_lendvai: How wrong did you make the hash?
<{^_^}>
[nixpkgs] @veprbl pushed to release-20.09 « rivet: add sansmath to fix make-plots »: https://git.io/JtSJ7
<attila_lendvai>
exarkun, i have updated the rev to a new git hash, and didn't update the sha256
<attila_lendvai>
ouch, one more surprise: C-c during a nix-build leaves half-downloaded sources in /nix/store and then subsequent nix-build calls fail until a nix-collect-garbage
<clever>
attila_lendvai: that shouldnt happen
<clever>
attila_lendvai: nix wont register the product as valid in sqlite until the download completes and it matches the sha256 you claimed
<clever>
any non-valid files are ignored when checking if it exists, and are silently deleted upon starting a build
<attila_lendvai>
clever, well, i C-c'd the build, subsequent builds have consistently failed with the same error, and nix-collect-garbage resolved the issue. not sure what happened under the hood. this is a go module.
ddellacosta has joined #nixos
<clever>
attila_lendvai: you can also add `repeat = 1` to /etc/nix/nix.conf, to make nix build everything twice, and fail if it cant product bit-identical results
<exarkun>
Might be because you had the older content for what you claimed was the newer version, and then running gc removed that bogus data from the store because it was unreferenced?
kenran has quit [Quit: leaving]
<attila_lendvai>
clever, good to know, thanks!
<attila_lendvai>
exarkun, could be. i was messing around quickly, and my model of this all is still somewhat foggy, so i can't really think this through
<clever>
attila_lendvai: you should also never edit anything in /nix/store/
sbock has quit [Ping timeout: 240 seconds]
<bqv>
clever: heh I do it every day
<{^_^}>
[nixpkgs] @peti pushed 14 commits to haskell-updates: https://git.io/JtSUZ
<{^_^}>
[nixpkgs] @peti pushed 14 commits to haskell-updates: https://git.io/JtSUl
<attila_lendvai>
do people use some emacs integration with nix to evaluate expressions from the editor, and/or open a repl comint buffer? what's the usual development env? i'm messing in xterm and nix-build stuff, but i guess there must be more sophisticated stuff i never saw anywhere mentioned.
wanamingo has joined #nixos
<clever>
attila_lendvai: there is also `nix repl`
<attila_lendvai>
i'm planning to embark on implementing openvpn-like multi-instance services for bee, and it's complex enough that i will need to take a look at the expansion of my expressions
<attila_lendvai>
also, i'd be grateful for hints what other services do something like openvpn. i'll need almost completely identical service instances, besides their dataDir and another option
lambda-11235 has quit [Ping timeout: 240 seconds]
rj has quit [Quit: rj]
<exarkun>
openvpn itself is available as a nixos service
<exarkun>
so I imagine you can either use that or at least learn a lot from its implementation
n-osborne has quit [Remote host closed the connection]
<attila_lendvai>
clever, re that repeat = 1... i'm on nixos, and /etc/nix/nic.conf is readonly/generated, but the nixos config doesn't have a repeat option. shall i just edit the file until the next nixos-rebuild?
hiro98 has quit [Remote host closed the connection]
meursault has joined #nixos
<attila_lendvai>
and how can i look at the "expansion" of my nixos/modules/services/networking/bee.nix in the repl? i don't even know the proper nomenclature... :|
* attila_lendvai
has managed to callPackage it, and got back a lambda
stree has quit [Ping timeout: 256 seconds]
<attila_lendvai>
no, it's not a lambda, it's a set. but it's not printed deep enough.
<clever>
attila_lendvai: modules cant be loaded with callPackage
<clever>
attila_lendvai: you must instead do something like `nix repl ~/apps/nixpkgs/nixos --arg configuration ./configuration.nix`
<clever>
that will load a local checkout of nixpkgs, and load a configuration.nix from a custom path
<clever>
then you can just eval `config.foo` to see what all of the config contains
Guest50899 is now known as benny
<clever>
configuration.nix can then set options to make the module do something
<attila_lendvai>
clever, i don't want to load them per se. i want to see their intermediate expansion. i.e. what a multi-server openvpn config "expands" to
<attila_lendvai>
clever, that sounds what i want, i'll try
neiluj has joined #nixos
<drozdziak1>
how do I turn a list of strings into an attrset with the strings as keys and a chosen different attrset as values?
neiluj has quit [Changing host]
neiluj has joined #nixos
<colemickens>
does someone have an example of nixos config configuring mimetypes and default applications. I am a bit exasperated with Discord's insistence of opening in the wrong browser.
<clever>
colemickens: ive had even worse problems, chrome and discord will either fail to spawn chrome entirely (the sandbox breaks it), or the new chrome process fails to contact the old one, then double-opens my profile, corruption all cookies
<clever>
steam and discord&
<clever>
*
slack1256 has joined #nixos
<colemickens>
my immediate response to that is not appropriate for the room, that's really awful.
<clever>
colemickens: to make things worse, chrome tends to restore all 4000 of my tabs when that happens
<clever>
if you close without restoring, its toast
<colemickens>
I don't think I've been singlely so mad about a single UX thing ever. Lost so much "data". Browser session is my biggest un-backed-up "liability" these days.
<clever>
colemickens: i use session buddy in chrome to back that up
<clever>
i tried one-tab, and it failed at its ONE BLOODY JOB, and all of its state is reset after each startup, lol
<clever>
the main problem, is that if chrome auto-restores tabs, they arent loaded, so it restores quickly
<clever>
but if session buddy restores tabs, it loads EVERY SINGLE ONE
<colemickens>
alright, I'll have to look into that
<clever>
you have to restore it one window at a time, then reboot chrome
<bqv>
Never been more glad I don't use chrome/firefox
<sshow>
I am packaging a shell script that uses 3 other nixos packages at runtime. Should these be specified as deps in the package derivation? If yes, should I set them as buildInputs?
<sbock>
hey everybody! i found a rather small cli tool which is not packaged in nixpkgs. i wrote a working derivation, however, i don't know if it is worth contributing to nixpkgs or if i should just contribute to nur :) i read through the contributing to nixpkgs guide but this does not specify what packages are wanted. the tool in question is https://github.com/ivaaaan/smug
n-osborne has joined #nixos
<flokli>
sbock: shouldn't be too big... Do you want to send a PR, adding yourself as a maintainer for it?
vrinek has quit [Ping timeout: 272 seconds]
n-osborne has quit [Remote host closed the connection]
ddellacosta has joined #nixos
afreakk has quit [Quit: WeeChat 2.9]
sangoma has joined #nixos
<sbock>
flokli: sure...why not :) thanks for the quick respone!
<sterni>
> builtins.trace (lib.generators.toPretty {} { "foo" = 12; bar = { z = 3; }; }) 12
<{^_^}>
12
<sterni>
({^_^} hides the debug output of course)
<sterni>
toPretty pretty prints any nix value into a string
<attila_lendvai>
clever, so, i'm making progress, but i can only see the config that i provided: x.config.services.bee, but the "expanded" part is empty: i.e. there's no x.config.users.groups.bee that should have been "generated"
n-osborne has quit [Remote host closed the connection]
<clever>
attila_lendvai: did you enable the service in the configuration.nix you supplied?
Dotz0cat has quit [Ping timeout: 256 seconds]
Dotz0cat_ has joined #nixos
<attila_lendvai>
clever, to the best of my knowledge, yes. i even tried to remove the mkIf cfg.enable part
<colemickens>
handy for updating inputs and commiting with a nice message: `nix flake update --recreate-lock-file --no-registries --commit-lock-file`
<clever>
attila_lendvai: you may need to restart `nix repl` after each change
* attila_lendvai
tries
inkbottle has quit [Quit: Konversation terminated!]
inkbottle has joined #nixos
domogled has quit [Quit: domogled]
<attila_lendvai>
clever, that was it, thank you once again!
<clever>
attila_lendvai: `nix repl` caches the contents of any file it parses
<sshow>
what command can I use to see all binaries a package provides? e.g. `pkgs.coreutils`
growpotkin has joined #nixos
<attila_lendvai>
is there a way to increase the print depth in nix repl? 'nix repl print depth' comes up empty...
<clever>
attila_lendvai: :p
<clever>
attila_lendvai: as-in, `:p foo` in the repl, `:?` shows all cmds
* attila_lendvai
feels ashamed for having been unable to find that out by himself
<clever>
attila_lendvai: it also tells you to :? on startup!
ket15 has joined #nixos
fuiltilt has joined #nixos
<ket15>
Hello, I want to launch steam inside of steam-run bash. But it cannot connect to network. How can I do that?
<ket15>
Yes, very strange need
<sterni>
clever: wouldn't :r work as well?
<clever>
sterni: sometimes, but ive found it to be buggy
<sterni>
I expect nothing less
<bbigras>
If I use `imports = [ ]` to split my config and I set `networking.firewall.allowedTCPPorts = [ ];` in 2 files. will they be merged or one will overwrite the other?
<sterni>
bbigras: they will be merged, i. e. appended to each other
<sterni>
bbigras: (unless you use stuff like lib.mkForce)
n-osborne has joined #nixos
<bbigras>
sterni: thanks!
dev_mohe has quit [Remote host closed the connection]
dev_mohe has joined #nixos
n-osborne has quit [Remote host closed the connection]
<attila_lendvai>
how can i merge a list of sets? i.e. i need to generate both systemd.services.foo-n = {} and also systemd.tmpfiles.rules = [ ... "foo-n" ...], one such set for each service instance.
<attila_lendvai>
or shall i just write two generator functions?
<attila_lendvai>
err, no, even with two functions i need a way to merge at the end
<sterni>
attila_lendvai: just generate the services set and then from that the tmpfiles list
<sterni>
would that work?
apache8080 has quit [Quit: Connection closed]
<colemickens>
It seems like our sshd settings are kind of restrictive
<attila_lendvai>
sterni, i don't see how that would work. but i was wrong, i should be able to do this with two functions.
<colemickens>
maybe because they're based on 5 year old articles (I suspect because that's what the Arch wiki also points to)
ATuin has quit [Ping timeout: 240 seconds]
<sterni>
tmpfiles.rules is only a list of the service names, isn't it?
<ambroisie>
Hello people: I am getting into home-manager, and I would like to know if there is a better way to handle my vim configuration
<ambroisie>
When I read the manual, it seems like home-manager can only configure plugins and the `.vimrc`, however I would like to be able to use the whole `.vim` directory file hierarchy
<ambroisie>
I already have a fairly extensive configuration in my old arch-based dotfiles, and would rather keep the idiomatic way of configuring my editor
<ambroisie>
What can I do in this case?
K0kada has quit [Quit: Leaving]
K0kada has joined #nixos
K0kada has quit [Client Quit]
K0kada has joined #nixos
<ambroisie>
And also, what's this about `programs.{neo,}vim.package`? Does it mean I need to use that as my user's `vim` package instead of the `pkgs.{neo,}vim` one ?
<ambroisie>
s/{neo,}/g apparently that's just a vim thing?
<ambroisie>
otherwise yeah, I already saw that "design pattern" when writing previous configurations
<sterni>
ambroisie: might be, then the idea is maybe that you can reuse this in the rest of your configuration to directly invoke the right vim version
gustavderdrache has quit [Quit: Leaving.]
<sterni>
or set your EDITOR / VISUAL env vars? if you have to do that manually
<ambroisie>
I guess it's more that the wording in the documentation was really unusual... I'll go look at the sources
sigmundv__ has quit [Read error: Connection reset by peer]
sigmundv__ has joined #nixos
cseder has joined #nixos
<drozdziak1>
`boot.kernelPackages = pkgs.linuxPackages_5_10` isn't working for me when I try to change my kernel, what do I need to do to install 5.10?
<drozdziak1>
it says that linuxPackages_5_10 9is not of type `package`
<drozdziak1>
crap, okay, that's not about the kernel itself
<drozdziak1>
nevermind then
zakame has quit [Ping timeout: 240 seconds]
davidv7 has joined #nixos
zakame has joined #nixos
ece has quit [Read error: Connection reset by peer]
MachX[m] has left #nixos ["User left"]
ece has joined #nixos
<hexa->
drozdziak1: you probably want pkgs.linuxPackages_latest instead
<samueldr>
hexa-: 5.10 being an LTS should be a fine choice
ahmed_elgabri has quit [Ping timeout: 240 seconds]
zakame has quit [Ping timeout: 240 seconds]
<hexa->
ah fair
zakame has joined #nixos
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
zakame has quit [Ping timeout: 264 seconds]
ddellacosta has joined #nixos
NULLx767 has joined #nixos
sgo has quit [Remote host closed the connection]
NULLx76 has quit [Ping timeout: 246 seconds]
NULLx767 is now known as NULLx76
ddellacosta has quit [Ping timeout: 240 seconds]
<drozdziak1>
when does 5.11 make it to `release-20.09`?