<infinisil>
pushqrdx: Only if you pass options that delete older generations
<infinisil>
E.g. a `nix-collect-garbage` without any options won't delete any old gens
<infinisil>
But passing a `-d/--delete-old` will make it delete all old gens
<pushqrdx>
i was talking about deleting none used packages, without the -d flag, so nix will keep packages used by older gens?
proofofkeags_ has joined #nixos
<infinisil>
pushqrdx: Yeah, just running `nix-collect-garbage` will delete only packages that aren't referenced from any gcroot/profile
cosimone has quit [Ping timeout: 260 seconds]
proofofkeags has quit [Ping timeout: 265 seconds]
werner291 has quit [Quit: werner291]
mbrgm has quit [Ping timeout: 260 seconds]
mbrgm has joined #nixos
<a-h>
Maybe I'm missing something. It's possible to make a Docker image from Nix, right? So, I should be using dockerTools.buildImage to add some stuff in, rather than starting with a Docker file and adding nix content to it?
<a-h>
Even if I just want to run a shell at the end of it,.
<a-h>
But that config won't look much like my home-manager configuration?
notakeyboardsmas has quit [Ping timeout: 245 seconds]
xlei has quit [Ping timeout: 246 seconds]
supersandro2000 has quit [Disconnected by services]
<{^_^}>
[nixpkgs] @veprbl pushed commit from @rmcgibbo to master « bluejeans: 2.17.0 -> 2.19.0 »: https://git.io/JLz8s
fuzzypixelz has joined #nixos
ericsagn1 has quit [Ping timeout: 258 seconds]
<a-h>
Ah, looks like I can't build a Docker image with Nix on a Mac that can run on Linux.
ddellacosta has joined #nixos
<a-h>
Package ‘docker-image-home.tar.gz’ in /nix/store/7jfap9asvggb1il9fr5nk20ckdva0ssn-nixpkgs-21.03pre254870.1121b2259b7/nixpkgs/pkgs/build-support/trivial-builders.nix:7 is not supported on ‘x86_64-darwin’, refusing to evaluate.
<hpfr>
this error look familiar to anyone upgrading recently? https://bpa.st/XE3A
ddellacosta has quit [Ping timeout: 240 seconds]
<duairc>
hpfr: I think you're wrapping Firefox twice
<duairc>
hpfr: Are you using home-manager by any chance?
<hpfr>
yes
<duairc>
You need to set home.stateVersion
<hpfr>
oh, is that new?
<duairc>
I don't think it's new. The default value of home.stateVersion is 18.09, and the default value of programs.firefox.package looks at the stateVersion to decide if it should be a pkgs.firefox or pkgs.firefox-wrapped
<duairc>
That error message happens when you try to wrap an already-wrapped Firefox
<duairc>
I think that might be a consequence of the recent changes to Firefox, but I'm not sure
r3drock has joined #nixos
<duairc>
I spent a long time trying to track down that error when I ran into it though. The fix is to explicitly set home.stateVersion to 21.03 or whatever
printfn[m] has joined #nixos
<hpfr>
oh, I've never actually looked through the firefox module. thanks for the heads up
<duairc>
hpfr: Let me know if setting home.stateVersion fixes it
<{^_^}>
[nixpkgs] @petabyteboy pushed commit from @berbiche to master « bottom: 0.5.4 -> 0.5.6 »: https://git.io/JLzVL
philr_ has quit [Ping timeout: 268 seconds]
<hpfr>
duairc: seems to have worked! never would have found that, thanks
<zeta_0>
i'll come back in about 10 minutes, to see if anyone responded, if not, then I'll just leave this channel and wait a while for anyone to respond to the reddit post!
<andi->
zeta_0: me looks, i've had something similar
<andi->
zeta_0: I do not use emacs but it also stopped working for me in VIM some weeks ago
<duairc>
hpfr: No problem, glad I was able to share that with somebody!
<andi->
zeta_0: actually it works again on recent unstable. does it show up in your PATH?
<magnos>
quit
<{^_^}>
[nixpkgs] @petabyteboy opened pull request #107237 → bottom: fix build on aarch64 → https://git.io/JLzwZ
magnos has quit [Quit: WeeChat 2.9]
<zeta_0>
andi-: when I type python and then press tab, the python-language-server shows up as one of the completions in terminal, but when I open a python file, emacs tells me it does not detect it, but in my .emacs I used the elisp code directly from the nixos instructions on how to install microsoft's python lsp, so at this point, I have no idea?
<andi->
Same, sorry, not an emacs user :/
<zeta_0>
andi-: thanks and it's fine, I posted this issue on reddit, so hopefully people will start responding eventually, to help fix this issue!
<andi->
zeta_0: not sure who checks reddit, maybe try the discourse?
Fare has joined #nixos
<zeta_0>
i've had the best luck with reddit so far(other than the formatting issue), so I think i'll stick with reddit for now.
<andi->
ok
LnL has quit [Ping timeout: 265 seconds]
LnL has joined #nixos
LnL has joined #nixos
pushqrdx_ has joined #nixos
sss2 has quit [Quit: Leaving]
<zeta_0>
well, it's time to give up!
<zeta_0>
andi-: thanks for the help!
pushqrdx has quit [Ping timeout: 264 seconds]
<zeta_0>
andi-++
<{^_^}>
andi-'s karma got increased to 50, it's a crit!
<zeta_0>
later!
sigmundv has quit [Ping timeout: 240 seconds]
sss2 has joined #nixos
marusich has joined #nixos
zeta_0 has quit [Quit: rcirc on GNU Emacs 27.1]
sss2 has quit [Quit: Leaving]
kalbasit has joined #nixos
h0m1 has quit [Ping timeout: 264 seconds]
zakame has quit [Ping timeout: 272 seconds]
h0m1 has joined #nixos
zakame has joined #nixos
m0rphism has quit [Ping timeout: 265 seconds]
c4rc4s has quit [Ping timeout: 240 seconds]
jpo has quit [Ping timeout: 246 seconds]
justanotheruser has quit [Ping timeout: 272 seconds]
LunarLambda has quit [Ping timeout: 260 seconds]
jboyens has joined #nixos
fuzzypixelz has quit [Quit: fuzzypixelz]
<maurer>
For a particular program, I need to generate a config file with a specific name - it only lets me set its config *directory*, not its name. As a result, writeText isn't really working for me.
<maurer>
Any ideas?
<andi->
maurer: writeTextFile { name ="test.txt"; location = "/dir"; text = "test"; }
<{^_^}>
[nixpkgs] @zowoq pushed commit from @0x4A6F to master « dasel: 1.9.1 -> 1.10.0 »: https://git.io/JLzyd
a64 has joined #nixos
<jophish>
domenkozar[m]: I'm sure you're aware, but just in case: cachix is returning 502 errors
<jophish>
arrr, why can't nix fail less hard when a cache isn't working
<jophish>
I wish there was an option to treat an unreachable cache as a miss
jpo has joined #nixos
marusich has quit [Ping timeout: 260 seconds]
schweby has quit [Quit: Das ist möglich, also tun wir es. Später.]
schweby has joined #nixos
jmeredith has quit [Quit: Connection closed for inactivity]
<energizer>
surely it should do that by default, no?
aw has quit [Quit: Quitting.]
spacefrogg has quit [Quit: Gone.]
aw has joined #nixos
fishcl has joined #nixos
spacefrogg has joined #nixos
justanotheruser has joined #nixos
jpo has quit [Ping timeout: 260 seconds]
a64 has quit [Ping timeout: 245 seconds]
<tnks>
jophish: yeah, I'm having the same problem... I don't know how to nixos-rebuild without cachix when Cachix isn't working. This makes me want to only use Cachix user-level in the future.
<jophish>
Try using: --option fallback true
rb2k has joined #nixos
r3drock has quit [Remote host closed the connection]
<energizer>
nix-build --fallback
<tnks>
what about for nixos-rebuild?
Fare has quit [Ping timeout: 260 seconds]
<tnks>
but also the fallback didn't work for me. Not sure what else to try.
<jophish>
tnks: guess you could look at the drv it's trying to build and run nix-instantiate --fallback whatever.drv
<tnks>
I'm trying a user-local ~/.config/nix/nix.conf in hopes that it can override /etc/nix/nix.conf.
marusich has joined #nixos
<tnks>
Okay, that seemed to work.
marusich has quit [Client Quit]
fishcl has quit [Ping timeout: 256 seconds]
<siraben>
Is it just me or caches are down?
<siraben>
`nix-build --fallback -A ...` still seems to try to reach the cache
<jophish>
I'm getting some weird unexpected end of file
<jophish>
also, although nix prints 'disabling cache for 60s' it still tries to fetch it straight after!
<siraben>
Neither --option fallback true nor --fallback work for me
justan0theruser has joined #nixos
justanotheruser has quit [Ping timeout: 272 seconds]
ericsagn1 has quit [Ping timeout: 260 seconds]
<tnks>
I think this is all related to cachix.org acting up.
<siraben>
It should be possible to not use cachix right?
<siraben>
Ok I just commented everything out in ~/.config/nix/nix.conf and it works
Supersonic112 has joined #nixos
Supersonic has quit [Ping timeout: 268 seconds]
Supersonic112 is now known as Supersonic
ericsagn1 has joined #nixos
bn_work has quit [Quit: Connection closed for inactivity]
jpo has joined #nixos
kalbasit has quit [Remote host closed the connection]
jpo has quit [Ping timeout: 240 seconds]
nrb has quit [Remote host closed the connection]
LnL has quit [Quit: exit 1]
LnL has joined #nixos
Jackneill has quit [Ping timeout: 240 seconds]
waleee-cl has quit [Quit: Connection closed for inactivity]
<jophish>
I had to remove /etc/nix/nix.conf and make a manual version without the broken caches
Fare has joined #nixos
Rusty1 has quit [Quit: WeeChat 2.3]
ezioedithore has joined #nixos
rb2k_ has joined #nixos
rb2k has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @Ericson2314 opened pull request #107238 → treewide: Start to break up static overlay → https://git.io/JLgOO
rb2k_ has quit [Ping timeout: 260 seconds]
<{^_^}>
[nixpkgs] @github-actions[bot] pushed 3 commits to staging-next: https://git.io/JLgOw
<{^_^}>
[nixpkgs] @github-actions[bot] pushed 4 commits to staging: https://git.io/JLgOo
rb2k has joined #nixos
cole-h has quit [Ping timeout: 264 seconds]
riksteri has joined #nixos
tinker has quit [Ping timeout: 272 seconds]
tinker has joined #nixos
jboyens has quit [Quit: WeeChat 2.9]
<eeva>
Hey, how do I temporarily rebuild my system config without one of the binary caches in my nix.conf? (It's not responding, and making my nixos-rebuild fail)
<eeva>
oh, I may have the same problem as jophish :D
<jophish>
eeva: perhaps there's a better way, but I made a copy of /etc/nix/nix.conf, and then replaced it with a file without the cache in
<eeva>
:+1: I'll do that
<jophish>
then the rebuild was fine
<eeva>
(my initial attempt was to disable all substituters, but it's gonna take a while to build everything)
<jophish>
note that the rebuild will reinstate the symlink nix.conf, so remove the cache from your configuration.nix first!
<jophish>
eeva: as thanks you could open an issue against nix or nixpkgs (if you feel like it)
<eeva>
jophish: I did fix my configuration.nix first, thanks for the heads up
supersandro2000 has quit [Disconnected by services]
<ronthecookie>
DigitalKiwi: it's the share URL functionality, but now I changed some stuff so I could debug it and it works when I run the daemon manually...
supersandro2000 has joined #nixos
<DigitalKiwi>
share from kde -> device?
<ronthecookie>
yeah
<{^_^}>
[nixpkgs] @flokli merged pull request #101482 → nixos/acme: lego run when account is missing → https://git.io/JTaKU
<ronthecookie>
but I think it's just dbus doing weird stuff so I'll just launch kdeconnectd manually
<{^_^}>
[nixpkgs] @lheckemann opened pull request #107245 → manaplus: init at 1.9.3.23 → https://git.io/JLgyC
<DigitalKiwi>
just ot be sure it's the "share file" on the icon on the dock that should right? (it worked for me but i am on fairly new nixos-unstable)
<ronthecookie>
share file and share url are in the same file in the codebase, i tested share url but i'm guessing behaviour would be similar
<ronthecookie>
anyway I think I figured it out, I'll report back in a bit
kirelagin has quit [Quit: No Ping reply in 180 seconds.]
zupo_ has joined #nixos
kirelagin has joined #nixos
zupo has quit [Ping timeout: 246 seconds]
<DigitalKiwi>
if you figure out how to stop it from making these messages that'd be cool lol Dec 20 10:17:06 mvp-nixos xsession[15597]: file:///nix/store/m30mvxwr1drwfxqszrfpdqwwg9n4c683-kirigami2-5.73.0/lib/qt-5.15.0/qml/org/kde/kirigami.2/templates/AbstractApplicationHeader.qml:102:9: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
neiluj has joined #nixos
neiluj has joined #nixos
neiluj has quit [Changing host]
<ronthecookie>
well I dobut that's happening unless someone here wants to do a deep dive into KDE/qt code
<DigitalKiwi>
...i've started to before
<ronthecookie>
me too, the source alone is like 4gb for qt, kdeconnect isn't that bad but I dobut that's there
<DigitalKiwi>
from what i remember from the last time i think there might be a qmake flag we could set to make that warning not show up but i couldn't figure out how :(
<nrb>
After my dumb question yesterday I find another error, hope it is not a stupid question again: The option value `boot.initrd.luks.devices' in `/mnt/nixos_root/etc/nixos/configuration.nix' is not of type `attribute set of submodules'.
<pinpox>
DigitalKiwi: could you help me out with the derivation? It works when building with nix-build but not when installed as a package
<pinpox>
I don't know where to start looking for the error
jophish has quit [Ping timeout: 256 seconds]
<pinpox>
Running nix-build in this directory (https://github.com/pinpox/nixos/tree/main/packages/mmonit) outputs a ./result with a working program. But then when I install it as a pkgs overlay it errors with "AssertException: Cannot compute the mmonit home directory" when trying to launch it, even though it seems to install fine
gurki has joined #nixos
<gurki>
hello; would it be advisable (or possible) to run nixos within a docker container on some centos host?
<gurki>
google results are a bit ambiguous
<gurki>
from what i understand nixos' approaches to things tend to be a tad different, i cant quite tell whether that affects me there
<colemickens>
Docker's not really meant to run an entire OS
philr has quit [Ping timeout: 240 seconds]
<colemickens>
I suspect you'd have about as much luck as trying to run any other systemd linux distro under docker.
<colemickens>
I've had decent luck running nix-builds for lots of things in a docker container though
<gurki>
in theory having a vm for that woudl be acceptable, but given that i already have a bunch of docker things on that server it would be nicer if i could just use that
<pinpox>
DigitalKiwi: Thank you. I noticed that the path returned by "whereis mmonit" is a symlink to another path in the nix store. When I change to that one and then do ./bin/mmonit in there it seems to work, so I guess I'm installing this wrong, the binary itself seems to be fine
Ikosit has joined #nixos
* DigitalKiwi
wonders why it doesn't make you have a sha
<DigitalKiwi>
pinpox: it looks like it expects to do a lot of writes to the readonly nix store
ericsagn1 has joined #nixos
<pinpox>
DigitalKiwi: So how do I fix that?
<pinpox>
Also, why would it work when run from the path itself?
autrim64 has left #nixos ["Killed buffer"]
dansho has quit [Remote host closed the connection]
dansho has joined #nixos
pbb has quit [Ping timeout: 268 seconds]
pbb has joined #nixos
<genevino>
is there a configuration.nix option to control "DefaultTimeoutStopSec" of systemd? i'm not willing to wait 90sec for my computer to reboot when i type reboot.
fishcl has quit [Quit: Quit]
<genevino>
i already use systemd.watchdog.rebootTime and systemd.watchdog.runtimeTime to circumvent systemd deadlocks when rebooting, but i still get hit by "A stop job is running" messages with a timeout of 90sec sometimes and want to change that.
andi89gi has quit [Remote host closed the connection]
<genevino>
and since this is impossible to actually debug because the behaviour of systemd is entirely non-deterministic, i'm looking for ways to tell it to wait for 10sec max, and if that was un-successful, just let me get on with my life and reboot the thing.
<DigitalKiwi>
i don't know of a generic one but glusterfs has a setting for it that you could make your own based off maybe...
<genevino>
systemd has a setting for that, i just don't see it exposed in nixos options.
<genevino>
i mean hey, at least i was able to reduce the possible time a nixos host requires to reboot to 2min, that's something, but that's still way too much.
<pinpox>
DigitalKiwi: I can't even run 'mmonit --help' that shouldn't have to write anything, but running "/nix/store/hhcqx50b59wfrav0p6v64wii6ghplh5m-system-path/bin/mmonit --help" works, shouldn't that be the same thing?
<vldn1>
Hi someone tested penrose and wtftw tiling wm?
<genevino>
DigitalKiwi: that did the job for said deadlock, thank you. unfortunately it wouldn't be systemd if i wasn't running in just ANOTHER FRESH OTHER deadlock now.
<DigitalKiwi>
it's made me wait at least 10 minutes a few times :(
<{^_^}>
[nixpkgs] @damhiya opened pull request #107254 → notsodeep: init at 2019.5.29 → https://git.io/JL2Ue
<genevino>
DigitalKiwi: yup same here, and i consider that behaviour excessively broken.
<DigitalKiwi>
pinpox: well it has a logs folder and a database folder upgrade script and settings but right now they're all getting put in the same path in /nix/store/* which is read only
<genevino>
open for any further ideas if anyone has some.
<genevino>
DigitalKiwi: reboot --force --force actually "does" reboot forcefully, but that's not what i want obviously.
<pinpox>
DigitalKiwi: I understand, but why does it work when I call it directly?
<pinpox>
DigitalKiwi: Even something like `mmonit -v` fails which just should print the version. Again, running `/nix/store/hhcqx50b59wfrav0p6v64wii6ghplh5m-system-path/bin/mmonit -v` works perfectly
davidv7 has quit [Remote host closed the connection]
dadada_ has quit [Ping timeout: 264 seconds]
da_dada has joined #nixos
<pinpox>
(I got that path from `whereis mmonit`)
<pinpox>
I guess my question is how to install it correctly, so it works with just the `mmonit` command
werner291 has joined #nixos
<DigitalKiwi>
is this different than the monit that there is a module for in nixpkgs
<pinpox>
DigitalKiwi: yes, mmonit is not monit
<pinpox>
I'm considering submitting a PR for mmonit when I have figured out how to package it
<pinpox>
(Explanation: mmonit is a central service for monitoring, monit is an agent that runs on every server and reports it's status to the mmonit instance
<pinpox>
)
<{^_^}>
[nixpkgs] @github-actions[bot] pushed 20 commits to staging-next: https://git.io/JL2TR
<{^_^}>
[nixpkgs] @github-actions[bot] pushed 21 commits to staging: https://git.io/JL2T0
asheshambasta has joined #nixos
davidv7 has joined #nixos
<DigitalKiwi>
oh
andi89gi has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @joachifm pushed 3 commits to release-20.09: https://git.io/JL2Yp
chloekek has joined #nixos
Nigel[m] has joined #nixos
nikivi has joined #nixos
nikivi has quit [Client Quit]
<chloekek>
What’s the easiest way to find the tarball url for the latest unstable? I normally DDG “nixos unstable”, then click some links to get to https://channels.nixos.org/, temporarily enable JS and AJAX, scroll to the correct *-unstable link, and then copy the nixexpr.tar.xz URL, but maybe there’s an easier way.
<simpson>
Ha, wow.
<simpson>
I'm not sure who put JS on channels.nixos.org. They ought not to have done that; it was a poor decision. But CLV.
<__monty__>
betawaffle: Not really, just what I've gleaned from chats. Though someone linked this fork yesterday which seems rather more hardline not in favor of flakes, https://cs.tvl.fyi/depot/-/blob/third_party/nix/README.md Maybe they document why.
<boogiewoogie[m]>
hey, it seems that since recently nix is depending on `aws-sdk-cpp`. why is that?
joesventek has joined #nixos
dstzd has joined #nixos
joesventek has quit [Client Quit]
<gchristensen>
I think it has dependen on aws-sdk-cpp for at least a few years
<patagonicus>
boogiewoogie[m]: If you want to, you can disable it with an override, but then you'll probably have to build nix yourself.
<gchristensen>
nix has built-in support for directly using S3 as a binary cache, that is what it uses it for
shibboleth has quit [Ping timeout: 240 seconds]
dstzd has joined #nixos
<boogiewoogie[m]>
oh, then I just never noticed before. thanks for the replies! =)
<{^_^}>
[nixpkgs] @petabyteboy opened pull request #107256 → libtorrentRasterbar: rename to libtorrent-rasterbar, move around some files → https://git.io/JL2nY
<andi89gi>
Is somebody going to update Calibre to the next current version? Atm nixpkgs.calibe is on 4.22. The current one on the homepage of Calibre is 5.72
<maralorn>
Is there somewhere a toIni or toToml funktion in nixpkgs?
<infinisil>
maralorn: lib.generators.toINI
<maralorn>
infinisil++
<{^_^}>
infinisil's karma got increased to 391
<infinisil>
Or pkgs.formats.{ini,toml}.generato
<infinisil>
generate*
<maralorn>
Just grepping through your PR.^^
<infinisil>
:)
waleee-cl has joined #nixos
zakame has quit [Ping timeout: 272 seconds]
zakame has joined #nixos
vidbina has quit [Ping timeout: 246 seconds]
andi89gi has quit [Remote host closed the connection]
lordcirth__ has quit [Remote host closed the connection]
andi89gi has joined #nixos
sigmundv has quit [Read error: Connection reset by peer]
sigmundv has joined #nixos
ddellacosta has joined #nixos
s1341 has quit [Quit: Connection closed for inactivity]
<{^_^}>
[nixpkgs] @mohe2015 opened pull request #107260 → eclipses: 2020-09 -> 2020-12 and add eclipse-rust → https://git.io/JL2ix
lordcirth has joined #nixos
<dhess>
Where is the `overlays` attribute of a Nix Flake documented? The RFC makes no mention of it. https://nixos.wiki/wiki/Flakes says, "Same idea as overlay but a list or attrset of them," but that implies that if I want to use the overlays of an input Flake, I need to introspect the attribute's type and then adapt accordingly?
<bigvalen>
I think I broke my system, by not really understanding how btrfs mirroring works on nixos.
<bigvalen>
I had a happy box, with nvme0n1p2 as /
<bigvalen>
I got a second drive...an nvme thing, with a sata interface, probably /dev/disk/by-id/.... I can't remember. I just attached it as a mirror, was surprised that it didn't need changes to /etc/fstab or whatever. All good.
<bigvalen>
I tried to fix with nixos-install - from a USB key which seemed to run fine. Is there something I'd need in hardware-configuration to remind btrfs where to find the other part of the mirror ?
<Nigel[m]>
<bigvalen "[ 5.335114] BTRFS error (devi"> Do you have both drives 'online'
<bigvalen>
Well ... probably. The NVME one, with a SATA interface, is brand new. It was online a wee while back
<bigvalen>
I'm trying to imagine what 'online' means in this case. Has something changed, where the ID I added it to btrfs has changed ? Maaaybe
<Nigel[m]>
<bigvalen "I'm trying to imagine what 'onli"> I would boot up with a livecd or similar and use brrfs tools to get them to play together again - then it might reboot as before
<bigvalen>
i booted it from a livecd, could mount the volume no problem - and hence, re-run nixos-install
<bigvalen>
That was the original command... I'll go back into the livecd, see if it's definitely seeing both halves of the mirror
<Nigel[m]>
<bigvalen "That was the original command..."> yes I think btrfs tools should let you see what is going on:
<Nigel[m]>
I use that to monitor devices
<Nigel[m]>
btw: I would expect adding a 2nd drive as raid1 would make no difference at all to NixOS - you might check what changed with your re-run nixos-install
Fare has joined #nixos
ddellacosta has quit [Ping timeout: 256 seconds]
<{^_^}>
[nix] @Ericson2314 opened pull request #4385 → Overall store subclassing → https://git.io/JL2D2
<fgaz>
is there a way to `import` a nix file with a fallback value if the file does not exist?
<sphalerite>
bigvalen: probably a missing driver
andi89gi has quit [Remote host closed the connection]
<sphalerite>
bigvalen: try rerunning nixos-generate-config from your installer system.
<sphalerite>
fgaz: if builtins.fileExists ./expr.nix then import ./expr.nix else {}
andi89gi has joined #nixos
<fgaz>
sphalerite: ooh thank you, I totally missed fileExists
andi89gi has quit [Remote host closed the connection]
dadada_ has joined #nixos
pushqrdx has joined #nixos
andi89gi has joined #nixos
Fare has joined #nixos
kirelagin has joined #nixos
pushqrdx has quit [Excess Flood]
<bigvalen>
Huh. Seems if you have mirrored compressed root btrfs, you need to add ' rootflags=device=/dev/sda3,device=/dev/sdb3 and rootfstype=btrfs ' to your linux command line
pushqrdx has joined #nixos
<duairc>
Are there any resources/documentation for defining a NixOps network as a flake? Is that a thing that you can do?
<shapr>
How do I tell emacs about tools that are only available inside a nix-shell? I can start emacs inside that shell, but I'm used to having emacs running for IRC and other things.
<dhess>
shapr: In my experience, the easiest way is via Emacs's direnv support. Hook direnv into Nix, hook Emacs into direnv.
<rb2k>
For https://github.com/NixOS/nixpkgs/pull/107207 I am adding myself as the maintainer (since it's a new package). Are there any preference to do it in one or two commits?
<DigitalKiwi>
rb2k: i like it in two commits and make sure the one that adds you as a maintainer is first so that a git bisect can't land on a broken package that lists a maintainer that doesn't exist
<isaacr>
I only have experience dealing with installing haskell libraries
<isaacr>
hopefully that's enough information, I assume this is a fairly remedial question about installing a dependency on your nix system...
andi89gi has quit [Remote host closed the connection]
troydm has quit [Quit: What is Hope? That all of your wishes and all of your dreams come true? To turn back time because things were not supposed to happen like that (C) Rau Le Creuset]
andi89gi has joined #nixos
<pushqrdx>
so i just triggered a very nasty side effect of overlays, i wanted to change meson package propagting gcc so i overlayed it
<pushqrdx>
now every shell i enter is rebuilding everything
<pushqrdx>
almost a system wide rebuild :D
<pushqrdx>
is there a way to avoid that in the future
<pushqrdx>
also, installing package is triggering random rebuilds
<pushqrdx>
packages*
<sphalerite>
pushqrdx: not using an overlay like that ^^
<pushqrdx>
yeah but was the alternative, other than upgrading, the issue with meson is that until recently it had a propagated dependency on gcc 9 which caused issues for me (trying to use gcc 10)
<pushqrdx>
this was fixed in a recent PR and merged
<pushqrdx>
the fix was to remove the propagated dependency, so i just did the same in an overlay
SomeoneSerge has joined #nixos
<pushqrdx>
but everything is rebuilding now
<sphalerite>
yeah no way around it really, if you're going to use that overlay everywhere. You could use it only in the contexts where meson propagating gcc9 is an issue though.
mallox has joined #nixos
mallox has quit [Client Quit]
<pushqrdx>
when i first joined here i had question about the use of stuff like lorri etc and the answer was you'll discover it naturally if you had the need, and so far my nix-shell life has been smooth, i don't face huge rebuilds often if never
<SomeoneSerge>
Is there a flag to enable recursive nix build/nix inside nix?
<pushqrdx>
would lorri have helped in that?
<andi89gi>
Hi, mates - I have tried to run nix-build ./ -A calibre - the only error is that by trying to build a derv the SHA-256 are not compatible
<tpw_rules>
i think lorri helps if you garbage collect regularly
troydm has joined #nixos
<pushqrdx>
i don't
<sphalerite>
pushqrdx: yeah not really, it would still have to build all that stuff at some point
Fare has quit [Ping timeout: 260 seconds]
rb2k has quit [Ping timeout: 268 seconds]
domogled1 has joined #nixos
domogled has quit [Ping timeout: 264 seconds]
domogled1 is now known as domogled
respawn_ has joined #nixos
andi89gi has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @Stunkymonkey opened pull request #107270 → MediaElch: init at 2.8.2 → https://git.io/JLaT3
andi89gi has joined #nixos
<{^_^}>
[nixpkgs] @Azd325 opened pull request #107271 → duf: Add darwin supoort → https://git.io/JLaTw
<sss2>
hi all, spice support in qemu builds in nixos/nixpkgs is somehow broken and unusable
domogled1 has joined #nixos
domogled has quit [Ping timeout: 256 seconds]
domogled1 is now known as domogled
zie has joined #nixos
chaoflow has quit [Ping timeout: 256 seconds]
fendor has joined #nixos
pushqrdx has quit [Remote host closed the connection]
jonatanb has joined #nixos
werner292 has joined #nixos
werner291 has quit [Ping timeout: 268 seconds]
werner292 is now known as werner291
ManiacOfMadness has quit [Ping timeout: 256 seconds]
zupo has joined #nixos
noudle has quit []
Dotz0cat has quit [Read error: Connection reset by peer]
cosimone has quit [Remote host closed the connection]
cosimone has joined #nixos
cosimone has quit [Read error: Connection reset by peer]
cosimone has joined #nixos
<proofofkeags_>
OK I'm struggling with getting a systemd unit to run on a system where I have packages installed with nix-env (not NixOS), I am consistently getting permission denied. Has anyone else encountered this before?
<{^_^}>
[nixpkgs] @petabyteboy merged pull request #107256 → libtorrentRasterbar: rename to libtorrent-rasterbar, move around some files → https://git.io/JL2nY
<shapr>
DigitalKiwi: so I could run `niv update` in the fermatslastmargin clone?
KarlJoad has joined #nixos
<KarlJoad>
Is there a way to find the name of a .drv file from its .nar archive?
<gchristensen>
I don't think so, the drv is a lookup to a nar, and many drv's can produce the same nar
seku has quit [Quit: Connection closed]
<KarlJoad>
Oof... Ok. My Raspberry Pi 3 is trying to download some .nar's and keep failing due to a timeout, even though the only thing I changed was an additional authorized SSH key for the user. I'm not sure what's up with that.
oxalica1 has joined #nixos
andi89gi has quit [Remote host closed the connection]
oxalica has quit [Ping timeout: 246 seconds]
oxalica1 is now known as oxalica
<gchristensen>
rather, drv -> narinfo -> nar
andi89gi has joined #nixos
<KarlJoad>
That makes more sense, in my mind now, thanks. I still don't know why the Pi is failing, which is weird. It built fine 3 hours ago.
ericsagn1 has quit [Ping timeout: 268 seconds]
werner291 has quit [Quit: werner291]
civodul has quit [Quit: ERC (IRC client for Emacs 27.1)]
andi89gi has quit [Quit: Leaving]
ericsagn1 has joined #nixos
gxt has joined #nixos
kirelagin has quit [Ping timeout: 260 seconds]
moet has joined #nixos
<moet>
hello! what's the correct way to tell nixops to use a local checkout of nixpkgs (or otherwise override the nixpkgs it uses to be different from the host system)?