<hodapp>
I appear to have found a Python package that claims to require the Python package opencv-python, which I can't really install, and it ignores the presence of opencv3
<hodapp>
not sure what I can do about this. I don't see a requirements.txt I can edit requirements in
iqubic has joined #nixos
clacke_movim has joined #nixos
asheshambasta_m has joined #nixos
thomasd has joined #nixos
<thomasd>
hi all, I'm coming from ubuntu 16/18 and am very used to gnome-terminal. which terminal (or shell, or terminal emulator) might be best for me w/ XMonad on NixOS?
sir_guy_carleton has quit [Quit: WeeChat 2.4]
<jackdk>
thomasd: I'm using one called "termite" with a different lightweight WM. I don't have strong feelings about it in either direction
asheshambasta_m has quit [Ping timeout: 248 seconds]
<thomasd>
im so used to stuff like ctrl+shift-c and ctrl-shift-v for copy/paste respectively, but in xterm it's all different
<thomasd>
maybe I can configure xterm some way
mexisme has joined #nixos
<gchristensen>
gnome terminal is fine t,,
<iqubic>
jackdk: I also use termite.
Synthetica has quit [Quit: Connection closed for inactivity]
polman has quit [Read error: Connection reset by peer]
<samueldr>
andi-: for ~ I was even thinking of declaring dotfiles bankruptcy and have it something like /Users/samuel/Configuration/Legacy and set XDG_CONFIG (or the right one) to /Users/samuel/Configuration
<samueldr>
the only annoying thing in that situation is how the "home" buttons now don't mean home anymore :/
<markasoftware>
is it the set of arguments to the nixpkgs expression?
<andi->
`import <nixpkgs>` imports the default.nix from your nixpkgs channel. The `{}` applies an empty set of attributes/parameters to it. E.g. you call the result of default.nix (which is a function) with no additional attributes.
<andi->
yes
<markasoftware>
ah
<andi->
> import <nixpkgs>
<{^_^}>
<LAMBDA>
<andi->
> :p import <nixpkgs>
<{^_^}>
<LAMBDA>
asheshambasta_m has joined #nixos
<markasoftware>
do all nix expressions take a set as an argument?
<markasoftware>
it's not possible to write a nix expression with no arguments?
<clever>
though in both cases, i'm cheating a bit, by using things like writeText, fetchFromGitHub, and haskellPackages.ghcWithPackages
<clever>
its more of an example of how to do a specific task, rather then re-implementing all of nixpkgs!
asheshambasta_m has joined #nixos
asheshambasta_m has quit [Ping timeout: 258 seconds]
<markasoftware>
is it possible to explicitly list additional runtime dependencies?
<clever>
markasoftware: refer to them somehow in your output
<markasoftware>
so i could make a txt with just one derivation per line and that would add them as dependencies?
<clever>
yep
<markasoftware>
noice
<clever>
but the program itself wont know where to find those things at runtime
<clever>
enless it knows to read that file
<markasoftware>
yeah, forgot about that
<clever>
mruntime
<markasoftware>
will you look at that, all-packages.nix has exactly 1,000 contributors!
<clever>
,runtime
<clever>
markasoftware: github probably has a cap on how many it will report
<clever>
,runtimedeps
<{^_^}>
In order of preference: Patch source OR ((if it uses PATH -> wrap with new $PATH) AND (if it uses dlopen, (patchelf --set-rpath in postFixup OR wrap with new LD_LIBRARY_PATH)))
<jchw>
maybe the SDL2 package would be a good reference here? SDL2 dlopens a lot of dependencies at runtime lazily
<clever>
depends on if its PATH or library
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
Glider_IRC__ has quit [Ping timeout: 272 seconds]
asheshambasta_m has joined #nixos
mkoenig has quit [Remote host closed the connection]
asheshambasta_m has quit [Ping timeout: 259 seconds]
rauno has quit [Remote host closed the connection]
abathur has quit [Ping timeout: 244 seconds]
Myhlamaeus has quit [Ping timeout: 252 seconds]
<ploupiboulba[m]>
is it possible to generate a node environment with packages similar to python ? the doc is very scarce. Grepping nixpkgs, I see nodeEnv but no actual use of it
fusion809 has joined #nixos
<Miyu-saki>
ploupiboulba[m]: node2nkx?
<ploupiboulba[m]>
Miyu-saki: I want to create a neovim node environemnt with packages required by the different plugins. Does it mean I will have to write the json myself ?
<adamantium>
Hey-- So i'm trying to use services.xserver.displayManager.startx.enable = true; and "startx" to run my wm without a dm / and as a regular user. But i keep getting the error: (EE) parse_vt_settings: Cannot open /dev/tty0 (Permission denied)
<adamantium>
Works as root, though...
<adamantium>
Does anyone know how to make this work?
asheshambasta_m has quit [Ping timeout: 246 seconds]
mkoenig_ has joined #nixos
mkoenig has quit [Ping timeout: 258 seconds]
<etu>
adamantium: Whats the permissions on /dev/tty0?
<etu>
Hmm, and I assume that you're not in the tty group or so. So that's why you can't write to it. But you shouldn't need to be in that group either.
<adamantium>
just users wheel disk audio video networkmanager systemd-journal libvirtd
weedloser has joined #nixos
adamantium has quit [Remote host closed the connection]
<v0id72[m]>
hi just wondering if someone could point me where to read more about using nix-shell on macOS? As I have read the info on the site. But are there any real world scenario articles? From what I understand it kinda sandboxes the apps from my macOS, is this correct? Looking to do dev in nix-shell and keep all of it isolated from my macOS - as I mess it up constantly. :)
<EdLin>
I've been having problems installing nixos (in a boxes vm for testing prior to full installation). It keeps giving me syntax errors about a semi-colon and equals sign being in unexpected places
Makaveli7 has joined #nixos
<johanot>
the systemd module doesn't have a reloadTriggers? I have a case, where I want to trigger reload, but I still want service restarts for other changes.
jasongrossman has joined #nixos
ym555 has joined #nixos
<johanot>
EdLin: can you paste your config somewhere?
<EdLin>
that's going to be kind of hard... it's in a VM with no IRC installed.
<EdLin>
not sure how to share files from boxes
<EdLin>
I just uncommented a few of the examples in the example config
<johanot>
can't you c/p the config out of the vm and into a github gist on your host system or something?
<tobiasBora>
I would need the adb executable. Some people say it should be in androidsdk. However, I tried to add "androidsdk" in my systemPackages, and I've an error "unknow variable androidsdk
gentauro has quit [Read error: Connection reset by peer]
<eddyb>
has anyone looked at VSCode "Insiders" (i.e. their unstable flavor) and the new remote mode, where it SSHes onto a server and runs a version of itself there?
m0rphism has joined #nixos
<eddyb>
someone is jumping the gun a bit after getting access to a NixOS server and I suspect there might be a lot of work to be done before anything works
<tobiasBora>
adisbladis: whoo thanks, but why such a complicated path? Also, I found programs.adb.enable = true; is it supposed to let the user use adb, or to let people from outside connect to the device using adb?
<adisbladis>
tobiasBora: That option adds androidenv.androidPkgs_9_0.platform-tools to environment.systemPackages and sets up udev rules to let users use adb.
bahamas has joined #nixos
<adisbladis>
I dont remember.. Upstream packaging used to be simpler around this.
<tobiasBora>
adisbladis: oh, so it's the recommended way to install adb I guess?
<adisbladis>
Yes for sure go with programs.adb.enable :)
<{^_^}>
[nixpkgs] @ktf opened pull request #62215 → igprof: init at v5.9.16 → https://git.io/fj0Oa
<turion>
Since lately, my `services.cassandra` doesn't start up properly anymore. journalctl says:
<turion>
Mai 29 09:39:10 nixos systemd[1]: cassandra.service: Failed with result 'exit-code'.
<turion>
Mai 29 09:39:10 nixos systemd[1]: cassandra.service: Main process exited, code=exited, status=126/n/a
<turion>
Mai 29 09:39:10 nixos cassandra[1020]: /nix/store/97gl5mlbfj91n1b64bp5lngnp8cig13n-cassandra-3.11.3/bin/cassandra: line 5: /nix/store/97gl5mlbfj91n1b64bp5lngnp8cig13n-cassandra-3.11.3/bin/cassandra: Argument list too long
<turion>
Mai 29 09:39:09 nixos systemd[1]: Started Apache Cassandra service.
<turion>
When I look at systemctl show cassandra, I can notice a few long lines, but nothing exceptionally long
<turion>
ARG_MAX=999999999 systemctl restart cassandra doesn't help
<turion>
The last line of /nix/store/97gl5mlbfj91n1b64bp5lngnp8cig13n-cassandra-3.11.3/bin/cassandra reads: exec "/nix/store/97gl5mlbfj91n1b64bp5lngnp8cig13n-cassandra-3.11.3/bin/cassandra" "${extraFlagsArray[@]}" "$@"
<turion>
So I'm suspecting that extraFlagsArray is somehow too long? Or whatever $@ is?
<{^_^}>
[nixpkgs] @flokli merged pull request #62069 → nixos: Make 'nesting.clone' work in NixOS tests → https://git.io/fjRfE
<gchristensen>
oh you have to disable ublock origin
<bgamari>
heh, sigh
<bgamari>
yes, it works with chromium
<bgamari>
so I guess that explains it
Shoubit has quit [Ping timeout: 268 seconds]
selfsymmetric-mu has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @CohenCyril opened pull request #62217 → recurseIntoAttrs of coqPackages → https://git.io/fj038
Makaveli7 has quit [Quit: WeeChat 2.4]
Synthetica has joined #nixos
dycan has joined #nixos
cfricke has quit [Ping timeout: 248 seconds]
<yorick>
does... nix log only work on succesful builds?
<andi->
no, but it usually only works for local builds
<yorick>
how do I get the nix log for a failed build?
<andi->
let me verify what I just said.. 1 sec
<yorick>
I've seen nix-store --query-failed-builds
linarcx has quit [Quit: WeeChat 2.4]
<andi->
works for me `nix log /nix/store/....-foo.drv`
<Shados>
yorick, andi-: 'nix log /nix/store/stuff....drv' should work for remote builders too, assuming they're in nix.conf or you specify them with --builders in the nix log command
<andi->
Shados: ahh yes, it was -K which never worked and I always wnated that to copy it to my machine
<yorick>
"error: build log of '/nix/store/asdf.drv' is not available"
<yorick>
oh, only the directly failing builds
<Shados>
yorick: Failures due to transitive dependency failures don't generate a log to begin with -- the logs are the output from the builder, which never runs if the dependencies aren't satisfied :)
cfricke has joined #nixos
asheshambasta_m has joined #nixos
<Shados>
andi-: Yeah, that one is a bit annoying. Doing `--keep-failed --builders ''` works, but then you don't have your remote builders doing the heavy lifting...
bahamas has joined #nixos
linarcx has joined #nixos
<andi->
yep
<andi->
while generally super happy with "magical" remote builds (imporive my notebooks battery time by a lot) those are the little things.. Especially painful if I did compile for >30min and failed..
Shoubit has joined #nixos
asheshambasta_m has quit [Ping timeout: 258 seconds]
<yorick>
andi-: this gives "command too long" sorts of errors
asheshambasta_m has quit [Ping timeout: 268 seconds]
<andi->
use xargs instead of the subshell?
<yorick>
ew.
knupfer has quit [Ping timeout: 248 seconds]
<yorick>
include-outputs worked well and even skipped missing outputs
<delroth>
after an upgrade+reboot cycle (on master) I can't get nix to eval my nixops config anymore, "out of memory". Before I spend more time looking into what's going on, is that a known issue already?
<delroth>
(that same nixops config had no issue building on nixpkgs master from 7 days ago, afaict)
Ariakenom has joined #nixos
Matthieu_ has quit [Ping timeout: 248 seconds]
<delroth>
also it takes less than a second to tell me it's OOM, which is kinda suspicious on a 16GB RAM machine, and I don't see any memory consumption peak
<lordcirth>
Ah, hardening. It always breaks something.
<delroth>
well, not usually :)
bahamas has quit [Ping timeout: 244 seconds]
<delroth>
I kind of expected it to be some hardening change since the problem was clearly not super widespread... but that one was better hidden than the others :p
mexisme has joined #nixos
arjen-jonathan has quit [Ping timeout: 272 seconds]
drakonis_ has joined #nixos
zeta_0 has joined #nixos
mexisme has quit [Ping timeout: 246 seconds]
drakonis has quit [Ping timeout: 258 seconds]
ricekrispie2 has joined #nixos
<zeta_0>
hello guys, how do i setup the aspell `nix` package with emacs? am i good with just the `aspell` package or will i need `aspellDicts.en`(for english) package as well?
bahamas has joined #nixos
ddellacosta has quit [Ping timeout: 248 seconds]
nschoe has quit [Ping timeout: 272 seconds]
ricekrispie has quit [Ping timeout: 248 seconds]
davidak has joined #nixos
<zeta_0>
and will this work with emacs out of the box or will i need add some elisp to .emacs?
Matthieu_ has quit [Quit: WeeChat 2.4]
<davidak>
is there a way to use the ip of eth0 (DHCP) in the webserver config (caddy) using configuration.nix?
<riottest000[m]>
Can anybody help me? I get `error: cannot auto-call function that has an argument with a default value ('stdenv')` when running nix-build
<andi->
yorick: whats the state of gnat? :D (I am trying to compile coreboot…)
Neo-- has quit [Ping timeout: 250 seconds]
<adisbladis>
I looked at it a few months ago (also to build coreboot) and gave up
drakonis_ has quit [Ping timeout: 252 seconds]
<infinisil>
,callPackage riottest000[m]
<{^_^}>
riottest000[m]: If a Nix file foo.nix starts with something like `{ stdenv, cmake }:`, you can build it with `nix-build -E '(import <nixpkgs> {}).callPackage ./foo.nix {}'`
<andi->
I am currently building it on a 17.09 channel version..
<andi->
Hope that works
<riottest000[m]>
<=7b^_^=7d "riottest000: If a Nix file foo.n"> Hmm, how can I integrate that in my default.nix?
<riottest000[m]>
So I don't have to use a long command just to build my thingy
pie___ has quit [Ping timeout: 272 seconds]
zeta_0 has quit [Quit: rcirc on GNU Emacs 26.1]
<infinisil>
riottest000[m]: Instead of taking arguments as { stdenv, ... }, make it `import <nixpkgs> {}`
<infinisil>
I mean
<infinisil>
`with import <nixpkgs> {};`
<{^_^}>
[nixpkgs] @lukebfox opened pull request #62226 → colorls: init at 1.2.0 → https://git.io/fj0cw
<infinisil>
Would be easiest
ng0 has joined #nixos
asheshambasta_m has joined #nixos
bahamas has quit [Ping timeout: 245 seconds]
<yorick>
andi-: the roommate is still working on it one day per week
Shoubit has quit [Ping timeout: 272 seconds]
asheshambasta_m has quit [Ping timeout: 246 seconds]
lordcirth has quit [Remote host closed the connection]
lordcirth has joined #nixos
toppler`` has quit [Ping timeout: 246 seconds]
toppler`` has joined #nixos
infinee has quit [Quit: leaving]
nschoe has joined #nixos
FRidh has quit [Quit: Konversation terminated!]
asheshambasta_m has joined #nixos
clacke_movim has left #nixos [#nixos]
horner has quit [Quit: zzZzZzZzZZZzzz…]
asheshambasta_m has quit [Ping timeout: 258 seconds]
clacke_movim has joined #nixos
horner has joined #nixos
jabranham has quit [Ping timeout: 246 seconds]
freeman42x has joined #nixos
justanotheruser has quit [Ping timeout: 272 seconds]
<{^_^}>
[nixpkgs] @delroth opened pull request #62227 → nixos/malloc: add scudo from LLVM compiler-rt → https://git.io/fj0cx
<stphrolland>
hi. When I launch nix-shell on a shell.nix file with buildInputs = [glibc], the binaries of glibc are exposed in my $PATH . But the includes are not exposed. I can see in my nix-store that the files are here (for example the file I need to access is ...glibc-...-dev/include/features.h), but the include directory is not part of the $PATH enriched by nix-shell. it's not the same guid used also. Any hint on what to include in buildI
<marek>
gchristensen: thank you! that's an awesome graph :)
<{^_^}>
[nixpkgs] @joachifm merged pull request #61994 → cpp-gsl: remove in favor of microsoft_gsl → https://git.io/fjBVX
<marek>
we should make "nixos bookmarks" as every day I'm discovering a new awesome link with something interesting
alex`` has joined #nixos
zeta_0 has quit [Quit: busy]
<stphrolland>
what is strange is that for ncurses, AFAIR I had not to make any distinction between the binaries and the headers, they were available just doing buildInputs = [ncurses]. It does not seem to be the same behaviour with [glibc]
<gchristensen>
marek: that is a cool idea
asheshambasta_m has joined #nixos
clacke_movim has left #nixos [#nixos]
<symphorien>
stphrolland: glibc is already in stdenc you shouldn't have to add it to buildInputs
<symphorien>
Export NIX_DEBUG=1 to see what flags are added by the cc wrapper to cc
Mr_Notheruser has quit [Ping timeout: 252 seconds]
zeta_0 has joined #nixos
justanotheruser has joined #nixos
asheshambasta_m has quit [Ping timeout: 248 seconds]
rsa has joined #nixos
<stphrolland>
@symphorien with NIX_DEBUG=1, which command should I launch to check cc-wrappers flag? using nix-shell, I can only the initial and modified final $PATH.
<symphorien>
Call the compiler
<catern>
Does /dev/stdout exist in the Nix build sandbox?
<clever>
adding -l reveals its a symlink to fd1 on yourself
<catern>
Yeah I just did that test, but here's the weird thing
<stphrolland>
@symphorien I'm trying to use cling, which is based on clang. I have no trace when launching cling. But when I launch the compiler clang, I clearly see the the flag -B/nix-store/...-glibc-...-dev/include
<catern>
"echo foo > /dev/stdout" fails
<symphorien>
Is cling wrapped ?
<clever>
catern: that will be the stdout of the bash that is running echo
<clever>
catern: which would be identical to just `echo foo`
<catern>
clever: sure, nevertheless it should still work
acarrico has joined #nixos
<catern>
but instead: /build/.attr-0: line 1: /dev/stdout: Permission denied
<stphrolland>
@symphorien but in the case of cling, I am almost sure it will use the clang compiler. WHat I mean to be in the wrapper ? I have opened the files by which gcc, and which clang. Currently trying to figure out what they do superficially
asheshambasta_m has quit [Ping timeout: 245 seconds]
asheshambasta_m has joined #nixos
fendor has quit [Remote host closed the connection]
<stphrolland>
@symphorien I think you point me in a good direction. What I've briefly seen from this post is interesting: https://discourse.nixos.org/t/working-on-compilers-in-nixos/2566 it deals about calling clang without the proper wrapper. I come back after some reading..
<catern>
clever: here's a non-nix reproducer if you care: sudo unshare --fork -U -p --mount-proc sh -c 'ls -l /proc/self/fd/; echo > /proc/self/fd/1'
ottidmes has quit [Ping timeout: 246 seconds]
selfsymmetric-mu has quit [Remote host closed the connection]
asheshambasta_m has quit [Ping timeout: 245 seconds]
<{^_^}>
nix#2878 (by edolstra, 1 week ago, open): Run builds in a pseudo-terminal
<catern>
niksnut: aha! awesome!
<clever>
ah, nice!
<catern>
niksnut: do you know that for a fact? because I don't think this is necessarily related to running with a pipe rather than a pty; that reproducer fails even in a pty
<{^_^}>
moby/moby#31243 (by kbrwn, 2 years ago, open): `/dev/stdout` is inaccessible by an unprivileged user in a container started as root
asheshambasta_m1 has joined #nixos
asheshambasta_m has quit [Read error: Connection reset by peer]
<catern>
Oh, nice find
<catern>
In that case I think the program I'm packaging will probably want to workaround this upstream - they currently check if /dev/stdout exists before running the /dev/stdout tests, probably they want to also check that they can write to stdout
<clever>
catern: ls and stat claim you can write to it, so that will be difficult to test, without just checking for errors
<catern>
clever: access(2) should reject it, which is how you should check
asheshambasta_m1 has quit [Ping timeout: 248 seconds]
<clever>
catern: ah, that sounds good
turion has quit [Quit: Leaving.]
drakonis_ has joined #nixos
asheshambasta_m has joined #nixos
orivej has quit [Ping timeout: 272 seconds]
asheshambasta_m has quit [Ping timeout: 246 seconds]
<{^_^}>
jwilm/alacritty#2243 (by doubleloop, 8 weeks ago, closed): Alacritty v0.3.0-rc2 does not start (Error creating GL context)
justanotheruser has quit [Read error: Connection reset by peer]
<mnoronha>
When I run `nix-channel --list`, I see nothing. What channel does this mean I'm using?
<mnoronha>
I'm on nixos
asheshambasta_m has joined #nixos
<circ-user-dLdoV>
It was triggered by a bug in glutin. glutin is just part of the cargo.toml, not a explicit dependency in nix. So I thought, if I rebuild the source, it might pick the correct glutin version with the issue fixed.
jackdk has quit [Ping timeout: 272 seconds]
freeman42y has quit [Ping timeout: 252 seconds]
asheshambasta_m has quit [Ping timeout: 258 seconds]
infinii has joined #nixos
horner has joined #nixos
MmeQuignon has quit [Ping timeout: 272 seconds]
<clever>
circ-user-dLdoV: even if you tell nix to rebuild from source, if you didnt add glutin, it will produce the same binary with the same problem
<clever>
circ-user-dLdoV: if you tell nix to override things and add glutin, it will automatically build from source and not use the cache
<circ-user-dLdoV>
there is no such package as glutin, I actually had a hard time to understand how it worked...
infinii has quit [Ping timeout: 248 seconds]
asheshambasta_m has joined #nixos
<clever>
circ-user-dLdoV: is glutin a cargo crate?
<circ-user-dLdoV>
yes
<nh2>
I'm installing NixOS on a new OVH dedicated server from the rescue system.
<nh2>
It's throwing a Python stacktrace for `systemd-boot-builder.py` doing `.../bootctl", "--path=/boot", "install"`
<nh2>
Can I use `nixos-install` without `/boot` partition?
<clever>
nh2: is the efivars fs mounted?
<nh2>
clever: would it appear in `mount | grep efi`? If so, then not
<clever>
circ-user-dLdoV: then it just needs to be in the cargo.toml, and the cargosha256 has to be updated
<clever>
nh2: then the problem is with /sys/firmware/efi/efivars, not /boot
ndarwinc1 has quit [Quit: WeeChat 2.4]
ndarwinc1 has joined #nixos
ndarwinc1 has quit [Client Quit]
ndarwinc1 has joined #nixos
ndarwinc1 has quit [Client Quit]
<nh2>
clever: I don't quite understand. The last thing the script prints is `Failed to check file system type of "/boot": No such file or directory`. I have the dir `/sys/firmware/efi/efivars` but it's empty
xantoz has quit [Ping timeout: 252 seconds]
mnoronha has quit [Ping timeout: 256 seconds]
<clever>
nh2: have you mounted /boot within the chroot?
asheshambasta_m has quit [Ping timeout: 258 seconds]
<nh2>
clever: there is no chroot, I have just mounted the target root partition at `/mnt` (which is an empty file system tree), and run `PATH="$PATH" NIX_PATH="$NIX_PATH" `which nixos-install` --root /mnt` as per manual
<nh2>
(I can mount `mount -t efivarfs efivarfs /sys/firmware/efi/efivars` if that helps somehow)
zeta_0 has joined #nixos
<clever>
nh2: ah, you also need to mount the efi system partition to /mnt/boot/ in that case
fusion809 has quit [Remote host closed the connection]
<nh2>
clever: I don't have an EFI system partition, do I need one? I think i'd prefer to just do a normal single-partition install
mbrgm_ has joined #nixos
<clever>
nh2: if you want a single-partition install, then you cant use EFI
<clever>
nh2: and systemd-boot is EFI only, so you need to switch to grub.enable = true;
<clever>
you will also need to ensue your using MBR partitioning, legacy on GPT needs another (unformatted) partition
mbrgm has quit [Ping timeout: 246 seconds]
mbrgm_ is now known as mbrgm
<nh2>
clever: oh, I see. Indeed `nixos-generate-config` generated me a `systemd-boot.enable = true`. By what means did it choose to do that? Because my given `--root /mnt` is on a GPT disk?