<judson_>
I'm having a weird problem that might be caused by not having an rdb module available. I confirm that rdb is not in the modules directory or modules.alias
<judson_>
Is it sensible to be trying to find RDB in NixOS?
<{^_^}>
[nixpkgs] @cdepillabout pushed 2 commits to haskell-updates: https://git.io/JJTC5
<ethnt>
Hi, question from a Nix newbie here: I'm running Radarr on my server with the `services.radarr` option, but the version that's on the 20.03 is out of date and is buggy. Is there any way to use the unstable version for that specific service?
<ethnt>
I suppose the other option is to make my own container, but I'd rather not do that, given how simple it currently is
<{^_^}>
[mobile-nixos] @samueldr opened pull request #177 → Fix Mobile NixOS regressions after the pkg-config changes in Nixpkgs → https://git.io/JJTWg
codygman has quit [Read error: Connection reset by peer]
knupfer has quit [Quit: knupfer]
knupfer has joined #nixos
knupfer has quit [Client Quit]
knupfer has joined #nixos
pr06lefs has quit [Ping timeout: 244 seconds]
codygman has joined #nixos
<{^_^}>
[nixpkgs] @peterhoeg pushed to f/calibre « calibre: use webengine i/o webkit »: https://git.io/JJT8e
<{^_^}>
[nixpkgs] @peterhoeg opened pull request #92118 → calibre: use webengine i/o webkit → https://git.io/JJT8v
h0m1 has quit [Ping timeout: 272 seconds]
simukis_ has quit [Quit: No Ping reply in 180 seconds.]
h0m1 has joined #nixos
simukis_ has joined #nixos
codygman has quit [Read error: Connection reset by peer]
drakonis has joined #nixos
codygman has joined #nixos
codygman has quit [Read error: Connection reset by peer]
codygman has joined #nixos
<etrigan63[m]>
<nh2[m] "etrigan63: I like your suggestio"> I would happily contribute if I could get the system to work. I am able to do the tutorial install but once I boot into the system I have no network.
<nh2[m]>
etrigan63:VM or real hardware? GUI installation (with network-manager?) or terminal? If terminal, need to check: Are the network interfaces named as you expect (`ifconfig`/`ip addr`), routes right (`ip route`), DNS right (`cat /etc/resolv.conf`)?
proofofkeags has quit [Remote host closed the connection]
pr06lefs has joined #nixos
proofofkeags has joined #nixos
Rusty1 has quit [Ping timeout: 264 seconds]
<{^_^}>
[nixpkgs] @samuelgrf opened pull request #92119 → g810-led: init at 0.4.2 → https://git.io/JJT8B
<etrigan63[m]>
Real hardware on my test laptop (no Ethernet). The generated config file has the hardware listed correctly. Everything should be handled via dhcp. If I log in to KDE (a la the tutorial) network manager is empty even though the service is enabled. The wireless is compatible because the graphical setup has the wireless network working correctly. It should not be this difficult. Jon's i3 config gets me lightdm and no
<etrigan63[m]>
desktop.
<{^_^}>
[nixpkgs] @veprbl opened pull request #92120 → hepmc2: 2.06.10 -> 2.06.11, mark as LGPL 2.1 → https://git.io/JJT8R
<nh2[m]>
etrigan63: With "the graphical setup has the wireless network working correctly", do you mean that in the live installer, KDE+network-manager has Internet fully working, but it does not work in the installed system (there the KDE-network-manager list is empty, but the wifi adapter is shown in e.g. ifconfig)?
<etrigan63[m]>
<nh2[m] "etrigan63: With "the graphical s"> Exactly.
endformationage has quit [Ping timeout: 256 seconds]
<aleph->
Stupid question since I can't remember what's the process for including an application in systemPackages that I've patched locally and over-riden?
stephank has quit [Quit: stephank]
stephank has joined #nixos
proofofkeags has quit [Remote host closed the connection]
proofofkeags has joined #nixos
nuncanada2 has quit [Read error: Connection reset by peer]
proofofkeags has quit [Ping timeout: 260 seconds]
<nbathum>
aleph-: what is the form of your override? You said patching, so maybe you have a custom expression, or are you doing like pkgs.foo.override and only modifying args?
<nh2[m]>
aleph-: where have you patched it? You can just do `environment.systemPackages = [ stuff (yourpackage.override { ... }) moreStuff ]`
<aleph->
nh2[m]: Built an entirely new version that's not upstreamed yet. So I have a local default.nix for it.
<aleph->
Just reading up on overlays now. Guess I should go read on overrides
<nbathum>
yeah what nh2 said. the parenthesis are key. so like `environment.systemPackages = [ (import ./my/new/package)) ]`
<nbathum>
since you have a default.nix, you can just provide the path to that local folder
<nh2[m]>
aleph-: If you want your new version to be used everywhere (e.g. not only in `systemPackages`, but also when some other package refers to it), then you should best make an overlay and add it to `pkgs` that way. Then in `systemPackages` you can just stick in that `pkgs.yourPackage`
<aleph->
Got it, yeah. Planning to use it with the service I've written as well.
<aleph->
So that makes sense
nuncanada2 has joined #nixos
sangoma has quit [Ping timeout: 256 seconds]
<etrigan63[m]>
<nh2[m] "etrigan63: https://github.com/Ni"> I'm confused. There is literally zero references to any networking in that configuration file.
<nh2[m]>
etrigan63: in the first file, `networking.networkmanager.enable = true;` and `networking.wireless.enable = mkForce false;` (disabling direct use of wpa_supplicant, probably so that network-manager can rule the device)
<etrigan63[m]>
What first file? I only saw the one file.
Supersonic has quit [Ping timeout: 272 seconds]
Supersonic112 is now known as Supersonic
<nh2[m]>
etrigan63: hmm, in the quote of my message to which you are replying, it does show the 2 files I mentioned (`installation-cd-graphical-base.nix` and `installation-cd-graphical-plasma5.nix`). Maybe Matrix is tricking us here?
pjt_tmp has joined #nixos
<etrigan63[m]>
It literally looked like one link to me.
<aleph->
Okay, so I'm obviously missing something with overlays. Here is where I'm adding my overlays file to my nix path, my overlays.nix file and the service/pkg I'm attempting to overload. https://paste.rs/2XV
<aleph->
So what exactly am I missing here?
knupfer has quit [Ping timeout: 256 seconds]
proofofkeags has quit [Ping timeout: 256 seconds]
nuncanada2 has quit [Read error: Connection reset by peer]
pr06lefs has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @cdepillabout pushed 2 commits to haskell-updates: https://git.io/JJTRc
Izorkin_ has joined #nixos
Izorkin has quit [Read error: Connection reset by peer]
Izorkin_ is now known as Izorkin
pr06lefs has quit [Ping timeout: 256 seconds]
cosimone_ has quit [Ping timeout: 246 seconds]
<jackdk>
I am trying to patchelf sorbet again, which has worked in the past. I tried (which might be wrong): `patchelf --set-interpreter $NIX_CC/nix-support/dynamic-linker bundle/ruby/2.6.0/gems/sorbet-static-0.5.5794-x86_64-linux/libexec/sorbet`. Now anything that tries to invoke that `sorbet` executable gives me "permission denied"
<jackdk>
I also get "permission denied" if I try running `$NIX_CC/nix-support/dynamic-linker $(which true)`, so I might just be patchelfing the wrong interpreter?
<jackdk>
grfn: I think there's some system about telling nix to trust the cache, or some kind of keying thing that you need to set up also. But I don't remember details.
<grfn>
jackdk: I'm setting `trusted-substituters`
lunatera has joined #nixos
<grfn>
and I have `require-sigs = false`, just to test
<nbathum>
grfn: not sure about sigs/keying, mine has a public key configured. but is your user added to `nix.trustedUsers:q
<grfn>
nbathum: yeah that was my first thought - and it is
<nbathum>
whoops wrong window lol
nfip^ has quit []
justanotheruser has joined #nixos
<nbathum>
well it definitely works with keys all setup and specified on both sides
<nbathum>
grfn: if `require-sigs` is the option equivalent to nix.requireSignedBinaryCaches, all that does is allow downloading binaries that are not signed. you'd still need to setup the authentication to connect to the cache
<grfn>
nbathum: yeah but I *can* connect to the cache
<grfn>
`sudo ssh ssh://cache` works but using it as a substituter gives an unauthorized error
<nbathum>
so to clarify what you're experiencing is that a normal ssh connection works, but nix does not?
<nbathum>
like `nix ping-store --store ssh://your-builder` produces an error
nwm_ has joined #nixos
simonthesourcere has joined #nixos
simonthesourcere has quit [Remote host closed the connection]
simonthesourcere has joined #nixos
<simonthesourcere>
Yesterday I was here with some problems getting grub to recognize my second disk with windows 10 on it. If I boot into the grub cli it seems it can't see the partitions on the windows drive at all
<simonthesourcere>
if I do `ls` I'd get `(hd0) (hd1) (hd1, gpt1) (hd1,gpt2)` or something very similar
<simonthesourcere>
where `(hd0)` is my windows disk
palo1 has joined #nixos
palo has quit [Ping timeout: 246 seconds]
palo1 is now known as palo
<simonthesourcere>
but if i do `insmod part_msdos` apparently I can see the partitions
<{^_^}>
[nixpkgs] @kalbasit merged pull request #91055 → nanopb: include C sources → https://git.io/JfFC3
<{^_^}>
[nixpkgs] @kalbasit pushed to master « nanopb: include C sources (#91055) »: https://git.io/JJTuB
<srhb>
simonthesourcere: Don't you want insmod part_gpt on an efisys?
<srhb>
It sounds almost like it's _not_ efi if msdos works
<simonthesourcere>
I don't know, I'm terrible with this stuff
<srhb>
Which I think is weird on Windows 10 (but I don't know much about that ecosystem)
user_0x58 has joined #nixos
<srhb>
simonthesourcere: Did you ever recover the grub config from the alternate machine? (if that was you)
maier has joined #nixos
maier is now known as kenran
<nicolas[m]>
Your nixos disk is partitioned with GPT while the Windows one uses MBR
<nicolas[m]>
AFAIK EFI boot only works with GPT partitions but I could be wrong
<srhb>
I think that's the case too. The chainloading will be different for non-UEFI-booting windows, I'm just confused that it's non-UEFI at all for such a new version.
<srhb>
s/new/recent
<nicolas[m]>
Yeah I believe Windows still defaults to MBR
<simonthesourcere>
hmmm, I'm not sure how to fix it
<nicolas[m]>
I think you could do an offline reconstruction MBR -> GPT
<simonthesourcere>
can I do that with smth like gparted?
<srhb>
Wouldn't chainloading mbr also be possible
<srhb>
Changing the partition table in-place sounds scaryu
<nicolas[m]>
Or add an extra grub entry with the insmod thingy
<srhb>
Yeah, that.
<simonthesourcere>
it doesn't help with the insmod thingy
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos
ericsagnes has quit [Quit: WeeChat 2.7.1]
simonthesourcere has quit [Remote host closed the connection]
orivej has quit [Ping timeout: 244 seconds]
ericsagnes has joined #nixos
alp has joined #nixos
dermetfan has joined #nixos
mtn has joined #nixos
sangoma has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @vbgl opened pull request #92139 → coqPackages.coq-extensible-records: enable for Coq 8.11 and 8.12 → https://git.io/JJTgK
<mtn>
I'm running uwsgi as a service (maps network requests to python functions), and my I can't write to the filesystem (a location outside the nix store)
<mtn>
I've verified the user it's running under can normally write to it without a problem, so I think it has something to do with nix
<{^_^}>
[nixpkgs] @dotlambda pushed commit from @r-ryantm to master « abcmidi: 2020.06.07 -> 2020.06.30 (#92114) »: https://git.io/JJT2N
<mtn>
fwiw, it can write to `/run/uwsgi` which has uid: nginx, gid: nginx, access: 0755
<sphalerite>
mtn: try using strace as energizer suggests. `strace -e file $UWSGI_COMMAND`
cfricke has joined #nixos
<mtn>
sphalerite: So I initially disregarded this when running it, but maybe it's relevant: I don't see any system call to try and open the file, just a python permission denied error
<mtn>
Maybe python is aborting before trying to make the system call, checking permissions on its own?
<sphalerite>
mtn: even then you should see a stat system call.
<sphalerite>
mtn: oh, I did forget `-f` on the strace command, it may well be happening in a child process.
<mtn>
sphalerite: it was a child process, thanks for pointing that out, never would've realized
<mtn>
Maybe the child process is running under a different user, I'll check the pid
<mtn>
Nah, it's also running under user nginx
<mtn>
The openat looks like this: `[pid 28703] openat(AT_FDCWD, "...path...", O_WRONLY|O_CREAT|O_TRUNC|O_CLOEXEC, 0666) = -1 EACCES (Permission denied)`
cfricke has quit [Quit: WeeChat 2.8]
<energizer>
does that user have rx perms on the parent dirs?
<energizer>
all ancestors
<mtn>
Nope
<energizer>
try changing that
<mtn>
When you say all ancestors, how far back do you mean? I have something like `.../labeler/labeler/uploads/` where uploads is owned by nginx, and `..` is the root of my project directory
<{^_^}>
[nixpkgs] @kalbasit pushed commit from @r-ryantm to master « h3: 3.6.3 -> 3.6.4 (#91591) »: https://git.io/JJTaE
dermetfan has quit [Ping timeout: 256 seconds]
<pjt_tmp>
I'm trying to set up git-annex's tor remote with services.tor.hiddenServices, but I don't know what to specify in the submodule for it to work. How might I determine that?
<ScottHDev>
Hello, I'm switching from sddm kde to lightdm qtile (with a new configuration in /etc/nixos/configuration.nix) but when rebooting I still get sddm?
NeoCron has joined #nixos
azure1 has joined #nixos
<sphalerite>
ScottHDev: is your /boot mounted correctly?
alp has quit [Read error: Connection reset by peer]
ixxie has joined #nixos
<nixuser>
Hi guys, I am trying to configure a hydra jobset depending on the Gitlab merge requests feature for a specific project. I am using the Open Gitlab Merge Requests with https://<gitlab>/api/v4/projects/9/merge_requests?state=opened&private_token=<token> . Unfortunately this results in an evaluation error "401 unauthorized". I've tried executing the same command in a browser which returns valid json. Does anyone know how
<nixuser>
to resolve this issue or how to configure a hydra jobset for gitlab merge requests? Thanks!
<{^_^}>
[nixpkgs] @cdepillabout opened pull request #92147 → kaggle: init at 1.5.6 → https://git.io/JJTr3
orivej has joined #nixos
<kenran>
,locate dhall-to-yaml-ng
<{^_^}>
Couldn't find in any packages
<raghavsood>
Are we supposed to do something special for cron? I'm trying to use it to clean up old docker images, but the commands seem to have no side effects. Even a simple command like `date >> /tmp/log.cron` doesn't add anything to the log file, although I can see both this and the docker commands being run according to `journalctl -fu cron`
faly has joined #nixos
faly has left #nixos [#nixos]
sigmundv has quit [Remote host closed the connection]
<clever>
raghavsood: have you tried using a systemd timer instead?
<raghavsood>
Not yet, but that's an option if cron doesn't work - just curious as to why it won't work at the moment
<raghavsood>
It definitely executes, just doesn't... do anything
pjt_tmp has quit [Ping timeout: 256 seconds]
dermetfan has joined #nixos
<duairc>
As far as I understand, if use hiPrio on an option, it will discard other instances of that option, but what I really want to do is just to force a particular ordering before they get merged
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/JJToJ
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
<raghavsood>
clever: the same commands in a systemd timer do work
orivej has joined #nixos
<ScottHDev>
Hello, I have a huge problem, every time I reboot, it seems like I'm rebooting on an old config. And nixos-rebuild switch doesn't change anything I always get to the same config, could you help me troubleshooting this?
<srhb>
ScottHDev: Typically either full /boot or you've switched bootloader without installing the new one?
<srhb>
ScottHDev: Pay very close attention to the last lines of nixos-rebuild boot for the first one.
<ScottHDev>
Actually the config might be the new builded one because I can see that cups is now starting at boot, but none of the packages installed in environment.systemPackages are available after boot
<srhb>
ScottHDev: Example?
<ScottHDev>
example of a package than isn't available at boot?
<srhb>
Yes :)
<ScottHDev>
home-manager
sheepfleece has quit [Quit: Lost terminal]
<srhb>
ScottHDev: And ls -l /run/current-system/sw/bin/home-manager confirms this?
<srhb>
ScottHDev: Compare your current generation to the profiles in /nix/var/nix/profiles/system -- which one are you on vs. which one is latest.
<ScottHDev>
Ok, I'm rebooting, just to be sure
<ScottHDev>
Yes it's confirmed : No Such file or directory
<nixuser>
clever: thanks again for your previous hint regarding hydra. Unfortunately I don't get, how to reference the nix expression (release.nix) inside the 'Open Gitlab Merge Request' input value. Do you have any idea?
<clever>
nixuser: and you then set that as the declarative project file for the whole project, that will auto-create a jobset called .jobsets
<ScottHDev>
Now I'll push the new derivation for qtile 0.15.1 (nixpkgs still build on old 0.13.0...)
<clever>
nixuser: the default.nix (line 6 of my json) will then receive a list of PR's as json, on one of its inputs (the name, and which PR's is defined on line 15)
<clever>
nixuser: that default.nix must then contain a single job, called jobsets, which returns a json file, a set of key=spec, where each spec is similar to the spec.json from before
<ScottHDev>
Hello back, I'm configuring a temporary binary cache. The goal is to rebuild my home-manager session by using as most as I can the store paths of another computer. I got it working but I get warning: substituter <substituer> does not have a valid signature for path
<ScottHDev>
Even with `home-manager --option require-sigs false switch`
<ajs124>
I've recently been getting an "Invalid Filename" error when saving files from my e-mail client (claws-mail), when they contain an umlaut. Any ideas why that might be?
<betaboon>
is there any builtin or library function to get the directory of a file. like coreutils-dirname ?
<pbogdan>
betaboon: there's builtins.dirOf
munksgaard has left #nixos ["WeeChat 1.0.1"]
cjpbirkbeck has quit [Quit: cjpbirkbeck]
kolibri_user has joined #nixos
orivej_ has joined #nixos
kolibri_user has quit [Client Quit]
orivej has quit [Ping timeout: 246 seconds]
<{^_^}>
[nixpkgs] @markuskowa merged pull request #92120 → hepmc2: 2.06.10 -> 2.06.11, mark as LGPL 2.1 → https://git.io/JJT8R
grumble has quit [Quit: Well, would you look at the time. I've almost missed my ambiguous, non-existent appointment that I have scheduled just when I start to lose interest in my current conversation.]
pr06lefs has joined #nixos
grumble has joined #nixos
rtjure has joined #nixos
xoxoboii has joined #nixos
xoxoboii has left #nixos [#nixos]
orivej_ has joined #nixos
orivej has quit [Ping timeout: 244 seconds]
Morfio has quit [Quit: Leaving]
fendor has quit [Remote host closed the connection]
Rusty1 has joined #nixos
fendor has joined #nixos
xoxoboii has joined #nixos
xoxoboii has left #nixos [#nixos]
xoxoboii has joined #nixos
xoxoboii has left #nixos [#nixos]
turion has quit [Ping timeout: 256 seconds]
waleee-cl has joined #nixos
orivej_ has quit [Read error: Connection reset by peer]
proofofkeags has joined #nixos
orivej has joined #nixos
mananamenos has joined #nixos
<mananamenos>
hi, i've booted usb drive with nixos. Im in plasma5. The user is nixos@nixos. How do I become root here to start formatting disk and installing nixos?
nixuser has quit [Read error: Connection reset by peer]
<mananamenos>
su -, and then i dont know the password. Hitting enter does not work
<mananamenos>
anyway it seems using sudo works without password so i can do it this way. thanks obadz
orivej has joined #nixos
<mananamenos>
is there a something wrong if my swap partition (8G) is sda1 while the main partition is sd2. I mean swap is in the beginning of sectors, while main partition follows
NeoCron has quit [Read error: Connection reset by peer]
janneke has quit [Remote host closed the connection]
janneke has joined #nixos
stiell has joined #nixos
doublex has quit [Read error: Connection reset by peer]
doublex has joined #nixos
cosimone has quit [Quit: Quit.]
orivej_ has joined #nixos
<lukego>
I'm running NixOS master and Blender doesn't give me GPU/OpenCL acceleration despite my running X with an AMD Radeon card (5700 XT). Any tips to troubleshoot? I already added 'config.cudaSupport = true;' to ~/.config/nixpkgs/config.nix and reinstalled blender 2.83 but no effect.
justanotheruser has quit [Ping timeout: 244 seconds]
orivej has quit [Read error: Connection reset by peer]
orivej has joined #nixos
wfranzini has quit [Quit: wfranzini]
<scoates>
anyone know if there's a reason nix-idea (https://github.com/NixOS/nix-idea) is stagnant/incompatible with 2020.*? just lack of interest? or maybe there's a fundamental change that needs a bunch of work?
<__monty__>
I'm writing a nix expression for a simple shell script and I'm getting a permission denied "cannot move $src/script.sh to $out", what am I doing wrong? installPhase = "mkdir -p $out; mv ${src}/script.sh $out/"
<FRidh>
$src points to the source in the store and is read-only. When you passed in a `src = ..;` it is already copied/unpacked into your local build folder
<colemickens>
Imagine a CI server and a nix cache. I have all of my system closure cached. I change the hostname. The CI box needs to redownload the entire system to build the updated toplevel for the hostname, even though it doesn't actually really need most of those store paths available (AFAIK). What if you could FUSE mount a nar cache over nix store for such a CI scenario. Or is this a bad idea?
<gchristensen>
so it wouldn't actually fetch those paths unless they were actually accessed?
<colemickens>
Hopefully, in theory.
codygman has quit [Read error: Connection reset by peer]
<gchristensen>
I've wondered about making the builder lazier myself
plp has joined #nixos
proofofkeags has quit [Remote host closed the connection]
codygman has joined #nixos
<colemickens>
mic92 has nix-build-uncached which works for my overlay pkgs, but I think my system closure will have the problem I describe above. I realize that some of the paths that need to be built will need to reach into *some* of the paths (and thus hit FUSE -> net -> cache), but I am wondering if it's a small enough amount for a win for a CI scenario.
<gchristensen>
like start the build before the closure is present and intercept store path accesses, block until they're actually substituted, and then realize them
dbe1 has quit [Ping timeout: 240 seconds]
proofofkeags has joined #nixos
<gchristensen>
but it would require very very very careful writing of drvs because they would be immediately accessed in most cases just by being in the PATH
<gchristensen>
though if we're dreaming, nix coud retain metadata about what was in those and fake out the replies
fendor_ has joined #nixos
proofofkeags has quit [Ping timeout: 240 seconds]
* colemickens
needs to read up on nix store internals and what metadata it tracks/stores.
emmanuel_erc has joined #nixos
<emmanuel_erc>
Hello there. I am currently using Arch Linux and I'm thinking of switching over to use NixOS on my personal laptop (or maybe a new laptop if it comes to that)?
<colemickens>
I *think* I was imagining something like that. I was hoping that, if a dependent derivation doesn't actually need the contents of the output, that it could be "faked" out, either at the FUSE layer or nix, yeah.
<emmanuel_erc>
Is the switchover process seamless?
<Twey>
emmanuel_erc: Not sure what you mean by that. You can play with Nix on Arch to get a feel for it before going full NixOS! But NixOS is a separate distribution you'll have to install separately (and manually — like Arch, we have no fully-automated installer yet, though NixOS does take care of more of the setup process for you once you've got your partitions and things in place).
<nabataeus>
Hello, is there a way I can set default display settings when I start an X session?
<nabataeus>
From the `configuration.nix' file?
proofofkeags has joined #nixos
ixxie has quit [Ping timeout: 256 seconds]
ckauhaus has quit [Quit: WeeChat 2.7.1]
<Twey>
nabataeus: What kind of display settings? You can set many options with the `services.xserver` config set (see `nixos-option services.xserver`, or the manual: https://nixos.org/nixos/manual/options.html#opt-services.xserver
<nabataeus>
So, I have an Xrander script that I want to start everytime xserver is loaded, how can I load that script?
<{^_^}>
[nixpkgs] @Ma27 pushed commit from @r-ryantm to release-20.03 « graylog: 3.3.1 -> 3.3.2 »: https://git.io/JJTAB
proofofkeags has quit [Ping timeout: 240 seconds]
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/JJTAz
<tpw_rules>
how can i start a nix-shell using a checkout of nixpkgs?
ixxie has joined #nixos
turion has quit [Ping timeout: 246 seconds]
ingenieroariel has joined #nixos
<Twey>
nabataeus: Probably the best way is to configure it in your Xorg config using services.xserver.extraDisplaySettings or services.xserver.extraConfig
turion has joined #nixos
<ingenieroariel>
Hello, I am new to cross compilation and while trying mobile-nixos I got the following error:
<Twey>
tpw_rules: (if you're not making local changes, you can also put the URL of a tarball there)
<tpw_rules>
if i can say nix-shell '<nixpkgs>' though, why can't i just say nix-shell /path/to/nixpkgs? i thought that's what <nixpkgs> was sugar for
<{^_^}>
[nixpkgs] @blaggacao opened pull request #92197 → WIP: kubespy: init at 0.5.1 → https://git.io/JJTAd
<Twey>
tpw_rules: You can in general, but passing the path to a derivation means something specific to nixpkgs (‘start a shell in the build environment of this derivation’) — so you can't combine it with things like -p
<nabataeus>
Twey: I'll do as you said, thank you.
orivej has quit [Ping timeout: 256 seconds]
zupo has joined #nixos
<nabataeus>
Though, I think I can use `.xinitrc' for this right?
orivej has joined #nixos
<Twey>
nabataeus: You can use `-A` but that's a little different
<nabataeus>
-A?
<Twey>
Oops, sorry
<Twey>
tpw_rules: ↑
<nabataeus>
Oh.. :D
nabataeus has quit [Quit: ERC (IRC client for Emacs 26.3)]
<tpw_rules>
ah i see. i was indeed trying to use -p
<Twey>
tpw_rules: `nix-shell /path/to/nixpkgs -A some.derivation` means ‘drop me into the build environment for some.derivation’, whereas `NIX_PATH=nixpkgs=/path/to/nixpkgs nix-shell -p some.derivation` means ‘drop me into a shell that has some.derivation available’
<typetetris>
fresheyeball: I work in a project, which uses node2nix, but a coworker has set it up. Ask your question I look if I can figure it out with our working code base.
<fresheyeball>
typetetris: so I have 2 projects that are node projects
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
zupo has joined #nixos
<typetetris>
fresheyeball: Shouldn't it be sufficient to add both of your projects to a single `npm-deps.json` file? You can specify a git repo or a tar ball there.
<pax-12>
how can I edit default.pa on nixos (I want to do load-module module-echo-cancel)
dingenskirchen has quit [Quit: dingenskirchen]
dingenskirchen has joined #nixos
<hazel[m]>
AmandaC: yeah, I noticed that -- it doesn't actually seem to matter
<hazel[m]>
I don't use that module so it might
<hazel[m]>
my apologies for that
dingenskirchen has quit [Client Quit]
dingenskirchen has joined #nixos
<AmandaC>
hazel[m]: the yaml it's generating is invalid, so the i3status-rust refuses to start due to it. I'm not that familiar with funcitonal programmign so I'm not sure how I'd go about fixing it
<AmandaC>
er, toml not yaml
dingenskirchen has quit [Client Quit]
<hazel[m]>
rip, I'll play with it -- one moment
dingenskirchen has joined #nixos
<hazel[m]>
the only module I use with that is maildirs so
<pax-12>
quinn, can I have an example, it is not in the documentation and I am new to nix and nixos
<hazel[m]>
since an extra " would be *extra* problematic
<AmandaC>
Nope, the extra " -- the three "'s in the output I sent are what's actualy in the file
<hazel[m]>
one moment let me check the mention again
MidHotaru has quit [Quit: Connection closed for inactivity]
<AmandaC>
Ending up with """, "prev", "next"
<hazel[m]>
yep I found it
dingenskirchen has quit [Client Quit]
<AmandaC>
s/prev/play/
<hazel[m]>
hm ok let me play with this for a few.
<hazel[m]>
I'll get back to you, just got home
<AmandaC>
sure. :)
dermetfan has joined #nixos
<AmandaC>
no problem, I commented out the offending option for now, thanks for taking a look at itanyway. :)
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « all-cabal-hashes: update to Hackage at 2020-07-02T22:46:43Z »: https://git.io/JJkeZ
eoli3n__ has joined #nixos
<hazel[m]>
oh I don't actually need a fold here
<hazel[m]>
I didn't know `concatStringsSep` existed at this time
<{^_^}>
[cabal2nix] @peti pushed to master « doc: update live-stream-notes.org before the merge »: https://git.io/JJkel
eoli3n_ has quit [Ping timeout: 256 seconds]
<typetetris>
fresheyeball: Otherwise my only idea would be to write a `nodeEnv.buildNodePackage` expression manually and add it to the generated node-packages.nix. Not nice, as one would have to patch every times its regenerated.
<{^_^}>
[nixpkgs] @peti pushed 2 commits to haskell-updates: https://git.io/JJkez
<{^_^}>
[nixpkgs] @peti merged pull request #91759 → Update Haskell package set to LTS 16.3 (plus other fixes) → https://git.io/JJvAc
drakonis has quit [Read error: Connection reset by peer]
<quinn>
pax-12: you set it in configuration.nix
arjen-jonathan has quit [Ping timeout: 256 seconds]
veehaitch has quit [Quit: leaving]
<{^_^}>
[nixpkgs] @jonringer merged pull request #87618 → [20.03] nixos/test-driver: Specify /bin/sh shell when running a bourne shell script as the user → https://git.io/JflX4
<{^_^}>
[nixpkgs] @jonringer pushed commit from @chkno to release-20.03 « nixos/test-driver: Specify /bin/sh shell when running a bourne shell script as the user »: https://git.io/JJkvY
<quinn>
pax-12: oh i saw your old message. just add whatever you want to add to default.pa to it as a string. do hardware.pulseaudio.extraConfig = "load-module module-echo-cancel"; in your configuration.nix and then nixos-rebuild switch
<quinn>
it should restart pulseaudio and everything for you and it'll load that module
<{^_^}>
[mobile-nixos] @samueldr merged pull request #177 → Fix Mobile NixOS regressions after the pkg-config changes in Nixpkgs → https://git.io/JJTWg
<{^_^}>
[mobile-nixos] @samueldr pushed 3 commits to master: https://git.io/JJkJw
orivej has quit [Ping timeout: 246 seconds]
KarlJoad has joined #nixos
orivej has joined #nixos
knupfer has quit [Remote host closed the connection]
knupfer has joined #nixos
<KarlJoad>
Has anyone managed to run 2 monitors off 2 separate nVidia cards? I believe I have the nVidia drivers installed, and am using the nVidia X server configuration tool, but cannot seem to get outputs correct.
KarlJoad has quit [Quit: ERC (IRC client for Emacs 26.3)]
emilsp has quit [Quit: bye]
emilsp has joined #nixos
tobi_ has joined #nixos
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
Kritnich has quit [Quit: Bye bye.]
bbb1 has joined #nixos
cybrian has joined #nixos
bbb1 has quit [Client Quit]
karetsu has quit [Quit: WeeChat 2.8]
Thra11 has quit [Quit: WeeChat 2.8]
zeromorphism has joined #nixos
drakonis has joined #nixos
<tobi_>
hello, is it possible to use install the emacs addon org-roam without using the emacs overlay? in the nixpkgs pull request 80280 it is mentioned that the package will arrive in nixpkgs "through the normal melpa workflow" once it is in melpa but I can't seem to install it
<hexo>
howdy, I still cannot get rid of slovak translations
<hexo>
something is still messing with $LANGUAGE var
<hexo>
$ echo $LANGUAGE
<hexo>
sk:en
<hexo>
wtf :(
<hexo>
i've even tried extraLocaleSettings = { LANGUAGE = "en"; LC_ALL = "en_US.UTF-8"; };
<hexo>
with partial luck
<hexo>
how is this even possible
<hexo>
?
<hexo>
that my nixos switched languages without being told to do so
<gchristensen>
that is pretty bizarre I've never heardof this
<gchristensen>
hexo: what is your preferred terminal?
<gchristensen>
erm, shell
<hexo>
bash
orivej_ has quit [Ping timeout: 256 seconds]
<neonfuz2>
I'm writing my first security bugfix to nixpkgs and I'm wondering how to submit so it will get merged into older channels
orivej has joined #nixos
<samueldr>
is it happening in a desktop environment? if so, is it also happening via SSH? a virtual console?
<neonfuz2>
it's relevant for 19.03 and up for sure
knupfer has quit [Quit: knupfer]
<neonfuz2>
is that the oldest supported version?
knupfer has joined #nixos
<neonfuz2>
(IE 18.09 is not supported)
<gchristensen>
hexo: I wouldrun: bash --login -x -i, then type `exit`, then comb through the logs bash printed to see if you can see where it is being set
<samueldr>
neonfuz2: 19.09 is kind of out of support now
<samueldr>
neonfuz2: though depending on the security issue it might warrant it
<aleph->
Yeah I should go update.
<aleph->
What's the latest revision?
<samueldr>
you don't want the "latest revision" ;)
<neonfuz2>
it's a remote buffer overflow so...
<aleph->
Okay, latest stable :p
<neonfuz2>
it really should be updated
<samueldr>
the latest stable version is 20.03 aleph-
<aleph->
Ya ya, thought so. kk
<neonfuz2>
does nixos only support the latest stable?
gnidorah has quit [Quit: Connection closed for inactivity]
<samueldr>
"kind of"
<neonfuz2>
lol
<samueldr>
neonfuz2: we might want to get you in touch with the security team before you publish anything
<samueldr>
(to coordinate, and better evaluate)
<neonfuz2>
its a publically known bug I'm just preparing to add patches to nixpkgs
<samueldr>
ah
<samueldr>
the way it was phrased I assumed it was an issue specific to Nixpkgs
<CRTified[m]>
Is there any example configuration.nix for a "family-friendly" setup? Basically a simple config with stuff that is helpful for non-powerusers?
<samueldr>
neonfuz2: the way it's done is to open a PR for nixpkgs-unstable/master, and depending on the skillset of the contributor, either ask that it is backported (be insistent for security isses), or open the PRs backporting the change themselves
<samueldr>
neonfuz2: the backport is basically using `git cherry-pick -x` on the commit, on the stable branches
<samueldr>
neonfuz2: the `-x` is important machinery so the commit is identified properly by git as being cherry-picjed
<quinn>
CRTified[m]: there is a "configuration collection" that is mostly advanced stuff on the wiki, but the one that comes by default is probably the simplest and best documented you will get. if you have specific things you want examples of i might be able to help though
orivej_ has joined #nixos
orivej has quit [Ping timeout: 246 seconds]
<tobi_>
systemctl fails to start redshift but when I run the command in ExecStart manually it works. the error messages in journalctl are "`RANDR Query Version' returned error -1", "Initialization of randr failed.", "X request failed: XOpenDisplay" and "Initialization of vidmode failed". I have a hunch that this might be because I'm running displayManager.defaultSession = "none+xmonad" but I have no idea how I could
<quinn>
tobi_: if it's a normal unit, not a user unit, i don't think it'll work at all. if it's a user unit, i think you want to make sure $DISPLAY is in its environment
<CRTified[m]>
quinn, ma27 : I can handle nixos myself (having contributed already). It's more or less the problem that I'm already too much of a poweruser so that I miss most stuff people expect from a normal operating system :D
<tobi_>
quinn: I'm not that familiar with systemd but I'm quite sure it is a user unit (I have to run "systemctl --user status redshift" to see see it). how would I add this variable to its environment declaratively?
knupfer1 is now known as knupfer
kenran has quit [Ping timeout: 256 seconds]
catern has quit [Excess Flood]
<quinn>
tobi_: i think you want to read the service file? if systemd has it, you want to use passenvironment, but i haven't used X in long enough that i can't offer much more
catern has joined #nixos
user_0x58 has quit [Ping timeout: 244 seconds]
NightTrain has joined #nixos
<tobi_>
quinn: thanks for your help, I will try to fix it
<quinn>
CRTified[m]: i mean i don't really know what you want. "normal" OS functionality for the average user is installing packages so if you just change systemPackages you're already there
knupfer has quit [Ping timeout: 256 seconds]
ddellacosta has quit [Quit: WeeChat 2.8]
simba2 has quit [Ping timeout: 260 seconds]
meh` has quit [Ping timeout: 264 seconds]
<aleph->
Hmm, does something seem off with this golang deriv? https://paste.rs/1nh Invoking via `nix-build -E 'with import <nixpkgs> {}; callPackage ./default.nix {}'` and I'm consistently getting `builder for '/nix/store/r65wcckv8gaa155xg91ijj80wink74d7-chihaya-1.0.0.drv' failed with exit code 1` type errors. Also been getting that in general when building golang derivations.
ddellacosta has joined #nixos
fendor_ has quit [Remote host closed the connection]
<jneplokh>
Hey, so yesterday I got the advice to store password files in `/var/` and change the permissions. To be specific, I am using Nextcloud. I made the db pass file and admin pass file, and it worked when everything was in `/etc/nixos/configuration.nix`. However, when I switched it to `/etc/nixos/nextcloud.nix` and imported it, when I go to the domain it says the db pass file does not exist. Why could this be?
<tobi_>
quinn: I've got it to work via systemd.user.services.redshift.environment. I had to define DISPLAY as well as XAUTHORITY which points at my .Xauthority file. thanks again for your help!
<jneplokh>
What's stranger is the admin pass file works just fine, I tested this by temporarily changing the location of the db pass file to a world-readable location, but the admin pass file is still in the same directory as where the db pass file used to be . Once I switch it back, it doesn't work
gustavderdrache has quit [Quit: Leaving.]
red[evilred] has quit [Quit: Idle timeout reached: 10800s]
<jneplokh>
Could it be something with the database user and permissions?
<jneplokh>
Ok this is weird, so I did further testing and even though there is a file at `/var/nextcloud-db-pass` when I go to my nextcloud domain it says the file does not exist. I have the permissions set where only root can access it, should I change that?
proofofkeags has quit [Ping timeout: 246 seconds]
tobi_ has quit [Quit: leaving]
<jneplokh>
And it only affects that file strangely, `/var/nextcloud-admin-pass` has the exact same permissions, yet no issue with it
o1lo01ol1o has quit [Ping timeout: 264 seconds]
demize has quit [Quit: Do Androids Dream of Electric Sheep?]
demize has joined #nixos
robogoat has quit [Remote host closed the connection]
<jneplokh>
What should I set the permissions of the file to?
<{^_^}>
[nixpkgs] @jtojnar pushed commit from @dasj19 to master « searhorse: 3.36 -> 3.36.2 »: https://git.io/JJkmJ
orivej has quit [Ping timeout: 246 seconds]
<quinn>
has anyone done any kernel work? kind of lost packaging amdgpu-pro
orivej has joined #nixos
<Ashy>
isnt amdgpu-pro effectivly obsolete now?
<Ashy>
i've got rx580s in all my linux desktops and just use the mainline drivers
<wrl>
Ashy: 5700xt still doesn't have mainline openCL or rocm support :c
<{^_^}>
[nixpkgs] @veehaitch opened pull request #92237 → vimPlugins.vim-monokai: init at 2020-06-17 → https://git.io/JJkmq
<quinn>
Ashy: i mean they dropped a new version on june 12. i *think* i need pro for 32bit opengl too
<quinn>
and of course i need 32 bit opengl for wine...you get the picture
<quinn>
wrl: i'm pretty close to having this working i think so if you have any experience with kernel packaging you might benefit yourself
<Ashy>
wrl: ah ok, that's good to know
Masoud has joined #nixos
codygman has quit [Ping timeout: 244 seconds]
codygman has joined #nixos
<wrl>
quinn: oh?
<Twey>
quinn: I just packaged some modules yesterday, I'm not an expert but it's fresh in my mind!
veehaitch has joined #nixos
<quinn>
wrl: i've done all of the packaging but the kernel module afaict. only issue is there is a #include in the module pointing to something missing from include/
<quinn>
so i interpreted that as it needing 5.6.x and this being minor version 13 of the patch? but i don't really understand kernel patch/module versioning so that could be important