<Unode>
iqubic: ok, then read a little about it. That's the tool that will likely help you get around having binaries (such as steam games) working on nixos
<Unode>
iqubic: yes but sorry, busy with other stuff.
<iqubic>
It's fine.
<iqubic>
How did you figure out that it was a mono issue?
<Unode>
iqubic: you might want to rethink nixos for a gaming platform. You are getting yourself into a lot of pain by using it. Specially if you don't (yet) have the know-how to workaround the most common issues.
<Unode>
iqubic: I poked in the game folder. Saw that it's mostly 32 bits and that it includes a bunch of MS-Windows files and a 'mono' binary.
garbas has quit [Client Quit]
<iqubic>
I know. I know.
<Unode>
from there I can only guess the problem has to do with /lib/ld-linux.so.2 which doesn't exist in nixos
garbas has joined #nixos
<iqubic>
Right. I know.
<Unode>
something that patchelf can help with.
<iqubic>
Should I read up on patchelf, or do you want to help me?
<Unode>
sorry I can't help right now. just giving you what I was able to identify from glancing at it
Guanin has quit [Remote host closed the connection]
jackdk has quit [Ping timeout: 245 seconds]
vk3wtf has quit [Ping timeout: 252 seconds]
aanderse_ has joined #nixos
aanderse has quit [Ping timeout: 250 seconds]
lassulus has quit [Ping timeout: 246 seconds]
lassulus has joined #nixos
teehemka_ is now known as teehemkay[away]
teehemkay[away] has quit [Quit: ZZZzzz…]
<gchristensen>
anyone been able to get nix's diff-hook to work?
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
jluttine has quit [Ping timeout: 250 seconds]
nikola_i has joined #nixos
vk3wtf has joined #nixos
rcshm_ has joined #nixos
blumenkranz has joined #nixos
<blumenkranz>
Hello. Do you happen to know how long does it take for an update to go through the whole Hydra process? I just had a Firefox update from 65.0 to 65.0, and I have seen the last update in the repos is from 4 days ago, and it just popped up now in my updates even though I have been using --upgrade the whole day.
<blumenkranz>
By the by, I am also having an issue downloading the latest SteamRuntime. It attempts to download from Valve, but it 404s, and then it attempts to download from abbradar.net, but it's down, so the rebuild process gets stalled. Is there any way to unstuck it?
<blumenkranz>
I am using "nixos-rebuild switch", but it still attempts to update.
<iqubic>
I've come to the conclusion that steam on NixOS is a lie. It doesn't work as well as you would expect. It's a giant bed of lies.
<iqubic>
Also, shouldn't I try the patchelf stuff first?
xkapastel has quit [Quit: Connection closed for inactivity]
<blumenkranz>
iqubic: That's pretty weird. Most of the native games I have tried worked well, except Borderlands 2, which needed some tweaks to skip the video cutscenes.
<Unode>
iqubic: not unless you know what you are doing. From what I checked it's probably easier to go with wine
<Unode>
blumenkranz: I think the game does some weird stuff to load libraries into mono.
ingenieroariel has joined #nixos
<blumenkranz>
Are you using firejail to run steam, by any chance?
<iqubic>
no.
<blumenkranz>
Mmm...
<iqubic>
Is there anyone I can get Mono 4.2.4 on Nixos?
<blumenkranz>
I mentioned it because the default profile includes seccomp, which breaks some instances of library linking.
<Unode>
iqubic: maybe an old nixpkgs commit
<iqubic>
How do I get firejail set-up?
<iqubic>
I don't have that right now.
<ingenieroariel>
hello channel, I packaged today Blender 2.8 beta for nixos, really recommended, in particular for kids (my daughter is going crazy making doll houses and exporting them to a VR app on the iPad). https://github.com/NixOS/nixpkgs/issues/55240
<{^_^}>
#55240 (by ingenieroariel, 1 hour ago, open): Provide Blender 2.8 beta as an option
<infinisil>
ingenieroariel: Phew, just thought I read Slender instead of Blender and was worried for a second
<blumenkranz>
iqubic: You need to "programs.firejail.enable = true", then add to "programs.firejail.wrappedBinaries" the line 'steam = "${pkgs.steam}/bin/steam";'
<ingenieroariel>
But I have no idea how to contribute it back to nixpkgs as it is a beta release, for now I am trying to package it in my nur repo
<blumenkranz>
But I warn you, it's an abstract kind of hell.
<blumenkranz>
Everything breaks twice as much.
jackdk has joined #nixos
<blumenkranz>
Only consider it if you are feeling particularly masochistic.
teehemkay[away] has joined #nixos
<iqubic>
blumenkranz: Does it make steam games work better?
<ingenieroariel>
infinisil: Slender Man, ¡corre por tu vida!
<blumenkranz>
iqubic: Nah, it just sandboxes them.
<blumenkranz>
You can configure it to block everything non-essential.
<iqubic>
So why did you suggest that as a potential fix to my issues?
<blumenkranz>
No, I mentioned it as a potential causation hahaha
<blumenkranz>
Just in case you were running it.
<iqubic>
Oh. I see.
<iqubic>
Unode: How do I force the use of 32 bit wine.
<Unode>
iqubic: google is your friend here
<blumenkranz>
iqubic: It appears in the NixOS wiki's article on Wine.
<Unode>
iqubic: and that link infinisil pointed you to back there is a really good read
<iqubic>
The one about how to ask questions?
teehemkay[away] has quit [Ping timeout: 250 seconds]
jluttine has joined #nixos
<blumenkranz>
infinisil: Should I open an issue on GitHub about the steam thingy, if only for bureaucratic purposes?
<infinisil>
blumenkranz: Yeah, and ping some people, like the ones that were involved in previous such problems
<infinisil>
Would be appreciated :)
<infinisil>
(and will get it fixed faster, if you can't fix it yourself)
<Unode>
iqubic: not just that section, the whole document.
<ingenieroariel>
I already know how to do it through nur repos (but that takes a day or two to become available)
<infinisil>
ingenieroariel: Ah yeah, that's an overlay, just put `nixpkgs.overlays = [(self: super: ...)];` into configuration.nix
teehemka_ has joined #nixos
<infinisil>
Or alternatively, put the `self: super: ...` thing into a file blender-overlay.nix, then use `nixpkgs.overlays = [ (import ./blender-overlay.nix) ]`
o1lo01ol1o has joined #nixos
<blumenkranz>
infinisil: oh wait, it seems like abbradar.net is back online
<blumenkranz>
Maybe it was just a temporary shortager
<blumenkranz>
It's slow as hell, but it seems to be advancing
<iqubic>
steam on wine is being a large pain.
<infinisil>
blumenkranz: Ah nice
<blumenkranz>
That said, is it normal it is downloading from abbradar and not Volvo's servers?
cransom has joined #nixos
Rusty1 has joined #nixos
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/fh9UZ
<iqubic>
So I can't get steam to work on wine on linux.
teehemka_ is now known as teehemkay[away]
teehemkay[away] is now known as teehemka_
Guest58298 has quit [Ping timeout: 250 seconds]
<ingenieroariel>
thanks infinisil
<ingenieroariel>
I ended up putting it directly in configuration.nix, for masochistic reasons I decided to not split my config until I reach the 1000 lines limit
silver has quit [Read error: Connection reset by peer]
blumenkranz has quit [Quit: WeeChat 2.3]
<iqubic>
So yeah. This mono game is a right pain.
sicklorkin has joined #nixos
<iqubic>
Because steam on wine crashed several times just now.
<aanderse>
does anyone have any thoughts on why @GrahamcOfBorg is reporting success on tests.redmine yet not really showing any meaningful log for this PR: https://github.com/NixOS/nixpkgs/pull/55234 ?
<gchristensen>
I keep a secrets.nix that I ust don't check in
<gchristensen>
I don't trust anything which involves mixing secrecy levels in the same repo
<bpye>
That is also a reasonable approach to take
upsaday_ has quit [Remote host closed the connection]
<bpye>
I do already manage a GPG keypair though
<gchristensen>
yeah, but it doesn't help if you accidentally forget to encrypt something
<DigitalKiwi>
also you can have private repos on github now
<bpye>
That's true :)
<gchristensen>
you might be smarter / more capable than I am, but I have personally accidentally pushed unencrypted private keys because I thought git-crypt had encrypted them.
<aanderse>
gchristensen: so you keep your secrets in the nix store? is it just you with access to the box?
<hodapp>
raaawr, don't put your secrets in version control in general
<bpye>
There is some value in versioning secrets though
<gchristensen>
use a different reop?
jasongrossman has joined #nixos
<gchristensen>
and don't push it
<bpye>
Maybe I'm best running gogs locally for that sort of thing, but then I'm in a bit of a catch 22 if I need to get that repo and my machine fails
<gchristensen>
use a different, private repo, which is encrypted?
<DigitalKiwi>
has bitbucket had incidents
hellrazo1 has quit [Ping timeout: 246 seconds]
<gchristensen>
will they have incidents? yes
<bpye>
I feel like it's less if and more when will an incident occur. Unfortunately that crap happens to everyone...
<bpye>
I'm glad we're on the same page
mibbitabc1234567 has joined #nixos
<DigitalKiwi>
just store everything on an encrypted drive and forget the password like i do
<bpye>
Managing private data, ensuring you have backups and ensuring it stays secret is definitely non-trivial
Supersonic has quit [Disconnected by services]
Supersonic112 has joined #nixos
Supersonic112 is now known as Supersonic
mibbitabc1234567 has quit [Client Quit]
lux1 has quit [Ping timeout: 244 seconds]
Ariakenom has quit [Quit: Leaving]
<bpye>
Sounds like a solid plan
<DigitalKiwi>
my secrets are safe with me even i don't know them!
countingsort has quit [Ping timeout: 250 seconds]
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
orivej has quit [Ping timeout: 240 seconds]
maximiliantagher has quit [Remote host closed the connection]
maximiliantagher has joined #nixos
sigmundv_ has quit [Ping timeout: 240 seconds]
sicklorkin has quit [Ping timeout: 272 seconds]
maximiliantagher has quit [Ping timeout: 245 seconds]
fresheyeball has joined #nixos
<fresheyeball>
so I just got around to upgrading to ghc863
<fresheyeball>
got all the deps builds
<fresheyeball>
it compilers in the nix-shell
<fresheyeball>
it compiles* in the nix-shell
<fresheyeball>
but fails on nix-build
<fresheyeball>
*** abort because of serious configure-time warning from Cabal
<fresheyeball>
what on earth is this fresh hell
<teto>
is it possible to merge https://github.com/NixOS/nixpkgs/pull/54948 ? the change is just version+hash but the package is long to build because rust builds everything (btw, I run this commit for the past few days without problem)
<{^_^}>
#54948 (by teto, 5 days ago, open): LanguageCLient-neovim: virttext support
teehemka_ is now known as teehemkay[away]
<ottidmes>
fresheyeball: how do you call nix-shell and nix-build?
<fresheyeball>
`nix-shell` and `nix-build` no args or nothin
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
<ottidmes>
then can you paste your shell.nix and your default.nix? I am asking because it might be that you be using different nixpkgs between them
<ottidmes>
or something else, but the more data, the better
<ottidmes>
personally I never run nix-build without arguments, either `nix-build '<nixpkgs>' --attr pkg` or `nix-build --expr 'with import <nixpkgs> { }; callPackage ./. { }'`
teehemkay[away] has quit [Quit: ZZZzzz…]
<ottidmes>
same for nix-shell, or my shell.nix has a { pkgs ? import <nixpkgs> { } } at the top, either way, I am always explicit about using <nixpkgs>, so I rule out it using some other mechanism
<ottidmes>
fresheyeball: nix/pkgs.nix is that somehow importing your nixpkgs repo?
<fresheyeball>
ottidmes: I don't have a shell.nix
<fresheyeball>
ottidmes: its a file that uses a pinned version, and does an overlay
<iqubic>
I can't understand why PlayOnLinux is giving me those messages, but then proceeding to work correctly?
<ottidmes>
fresheyeball: what then does nix-shell do for you?
<ottidmes>
fresheyeball: anyway, I guess the problem lies into your runtime system using something different than your pinned nixpkgs
<fresheyeball>
it setups up a an environment where I can build my software, and supplies additional developer tools and scripts
<fresheyeball>
but essentially its the same as cabal2nix ./. --shell
<ottidmes>
fresheyeball: which has: { nixpkgs ? import <nixpkgs> {}, compiler ? "default", doBenchmark ? false }: which is not your pinned repo
<iqubic>
so what should I do about this PlayOnLinux issue?
<ottidmes>
iqubic: not that I will be able to help you with your issue, but maybe paste the actual errors?
mizu_no_oto has joined #nixos
<iqubic>
there aren't any errors yet.
<ottidmes>
fresheyeball: how about: nix-shell --arg nixpkgs 'import nix/pkgs.nix' ?
<iqubic>
Just PlayOnLinux telling me that my games might not work.
<fresheyeball>
there isn't an arg because my default.nix is not a function
<iqubic>
PlayOnLinux is fine to continue without knowing where the library is.
<fresheyeball>
it statically expresses an env and does not have arguments beyond the ones you see in the file
<ottidmes>
fresheyeball: huh? nix-shell uses shell.nix, which you just said was basically what is being outputted by cabal2nix ./. --shell, or does --shell not just generate a shell wrapper around what it would normally generate for default.nix?
Dagger has quit [Excess Flood]
<fresheyeball>
what it generates does shell as well as build
Dagger has joined #nixos
<iqubic>
you should just post the generated shell.nix.
<ottidmes>
fresheyeball: I know, but I am quite sure that that is the cause of your problem, the thing I mentioned, about it mentioned <nixpkgs> while you want to use nix/pkgs.nix
<mdash>
ottidmes: do you still use bitwarden_rs?
<ottidmes>
fresheyeball: I am just not sure what you mean with "there isn't an arg because my default.nix is not a function"
<ottidmes>
mdash: yep
<ottidmes>
fresheyeball: that I got from your default.nix, but I was using nix-shell, there, so I am confused
<mdash>
ottidmes: is your gist up to date? ;-)
<ottidmes>
mdash: that is the gist I am running on my server, yes
<mdash>
i see it's not in nixpkgs yet
<mdash>
ottidmes: ok cool
<ottidmes>
mdash: correct, I do not want to publish it without the web frontend packaged as well, but I have tried a few times, but was unable to package it on NixOS (I got it working for myself by just going outside of Nix, and just using npm directly, since its only needed to generate the HTML files, nothing more)
<mdash>
ottidmes: i'll have a whack at it
<mdash>
any particular speed bumps you remember?
<ottidmes>
please do, I kept getting stuck mostly on one dependency not being supported
<ottidmes>
it uses the git:// protocol
<mdash>
joy :-[
<ottidmes>
which they do not like (node2nix, yarn2nix both variants)
<ottidmes>
mdash: I even tried an insane workaround, hijacking the calls to git and faking the responses for that dependency, but that did not work out
mizu_no_oto has quit [Quit: Computer has gone to sleep.]
<mdash>
ottidmes: first thing that comes to mind is the brute force approach of writing a derivation for it first and supplying it as an override
<ottidmes>
mdash: yeah maybe if it was supplied by an explicit fetchGit or some such it might work
<ottidmes>
mdash: seems like it already does
<ottidmes>
not sure which of my attempts this was, but I generated node-packages.nix with one of the tools, and it does use fetchgit properly for the git dependency
<ottidmes>
ah, node2nix did that (comment at top)
<ottidmes>
mdash: another problem case is opencollective
<ottidmes>
mdash: but that one the workaround worked properly for (I just create a /bin/opencollective that always returns 0)
iqubic has quit [Remote host closed the connection]
<ottidmes>
mdash: I can give you my attempt
<mdash>
mmm I get a different cargoSha256 than you did apparently
<gchristensen>
I did a thing as a quick two-evening hack, calculating how reproducible our minimal ISO is. Out of 1,244 build steps, 1,225 are bit-for-bit reproducible (98.473%)! check it out https://r13y.com/
<fresheyeball>
mdash: I am debugging this error, *** abort because of serious configure-time warning from Cabal
<fresheyeball>
from nix-build
<ottidmes>
mdash: I have to go now, to get some sleep, but I think your idea of prepackaging that problematic dependency seperately is probably the only way to make it work
<fresheyeball>
I can build in nix-shell just fine
<mdash>
ottidmes: i'll let you know
<ottidmes>
mdash: ok, that would be awesome, bye!
<mdash>
fresheyeball: -K will leave the build dir in /tmp so you can look at it
<fresheyeball>
mdash: just nix-build -K
jasongrossman has quit [Quit: ERC (IRC client for Emacs 26.1)]
<DigitalKiwi>
what does it mean when there are -unwrapped versions of things
<mdash>
DigitalKiwi: that the one that goes in the environment is a wrapper
<daniele->
hi there. has anyone installed nixos in a systemd-nspawn container? i just tarted with nixos and i have a problem regarding the /etc/os-release file. nspawn does not like the fact that it is a symlink
endorphin has quit [Client Quit]
<fresheyeball>
mdash: now that I found the error message
<rnhmjoj>
is it acceptable to "package" a software in nixpkgs by directly calling a nix espression in the source of the package itself?
<clever>
rnhmjoj: IFD isnt allowed within nixpkgs, it ruins the speed of nix-env -q
<fresheyeball>
is there a way I can just force my way past this?
endorphin has joined #nixos
<fresheyeball>
I just tried doJailbreak
<fresheyeball>
no love, same problem
endorphin has left #nixos [#nixos]
kleisli has joined #nixos
<clever>
fresheyeball: you have 2 different versions being mixed in, you need to ensure everything comes from the same haskellPackages and same override set
<clever>
fresheyeball: can you link the nix expressions?
<clever>
fresheyeball: callCabal2nix generates a nix file, in the /nix/store/
<clever>
fresheyeball: and then callPackages's that
<clever>
so you still have a nix file
lassulus has quit [Ping timeout: 268 seconds]
lassulus_ is now known as lassulus
<clever>
hlint also uses haskell-src-exts
countingsort has joined #nixos
<clever>
fresheyeball: oh, you can also use your overlays to just set haskell-src-exts_1_21_0 = throw "stop doing that";
<clever>
fresheyeball: then it will give a clear error, when something didnt stop doing it
pie__ has joined #nixos
<clever>
error: while evaluating the attribute 'propagatedBuildInputs' of the derivation 'hlint-2.1.14' at /nix/store/9s913lgsy4ravr2m0iqyb92646jlidnk-source/pkgs/stdenv/generic/make-derivation.nix:183:11:
hellrazo1 has joined #nixos
slack1256 has quit [Remote host closed the connection]
<teto>
anyone familiar with cmake enough to have an idea how to solve `Could NOT find Lua (missing: LUA_LIBRARIES) (found suitable version "5.1.4", minimum required is "5.1")` (full log https://paste.ubuntu.com/p/RbPq4P3Mpf/). Would writing a .pc file for luajit solve the problem for instance ?
__monty__ has joined #nixos
blumenkranz has joined #nixos
daniele- has quit [Quit: daniele-]
upsaday has quit [Ping timeout: 250 seconds]
<blumenkranz>
Hello. Yesterday I had some issues updating steam due to the steamrt mirror acting wonky. It is true there has been a steamrt update lately, but checking the nixos-unstable git ahows the last commit for the
<blumenkranz>
*the last commit shows its last update was 3 months ago. Can it be possible the package has been rebuilt due to some dependencies updating, and has to redownload aaaal of it again?
slack1256 has joined #nixos
<blumenkranz>
Like, for example, if python (which is a steam-runtime dependency) has been updated, it has to rebuild the steam-runtime package without reusing the cache.
<__monty__>
Maybe it depends on stdenv or something? That might cause a rebuild, yeah.
agander_ has joined #nixos
<__monty__>
It can't use the cached artifact because the identifying hash is based on the dependencies. So if the dependencies change the hash changes.
<blumenkranz>
Depends on python2, dpkg and bintools
<__monty__>
It should probably be a fixed output expression though.
civodul has joined #nixos
tzemanovic has quit [Remote host closed the connection]
<blumenkranz>
Guess an update to any of those would explain the rebuild, but I dunno why it doesn't happen more often.
tzemanovic has joined #nixos
agander has quit [Ping timeout: 240 seconds]
<__monty__>
Nixos-unstable's probably updated less frequently than nixpkgs-unstable?
<clever>
varies
<clever>
nixpkgs-unstable can be held back by darwin failures
<clever>
nixos-unstable can be held back by nixos levels failures (grub, key services)
<Avaq>
Hi! I'm hoping somebody can help me with this; I've tried getting into NixOps twice, but I ran into the same problem both times. The first time I figured I'd wait, because it's probably a bug in nixops on unstable; but now, half a year later, nothing changed.
<Avaq>
It then goes on to output the info, the way I expect. However, the 'nixops deploy' command in the next step gives the same error and immediately exists with 1.
<Avaq>
warning: evaluation of the deployment specification failed; status info may be incorrect
<Avaq>
error: value is null while a set was expected, at /nix/store/n8mg75pf8qchc8s982j5kiz45rzildpp-nixos-19.03pre168320.2d6f84c1090/nixos/pkgs/top-level/default.nix:63:5
<Avaq>
Like I mentioned; Last time I decided to give up and wait some time, but the situation hasn't changed in half a year. :\
<ottidmes>
mdash: did you already have a try, or did you not yet come around to try and package bitwarden web?
Dagger has quit [Excess Flood]
nikola_i has quit [Ping timeout: 250 seconds]
<{^_^}>
[nixpkgs] @teto opened pull request #55266 → Enable functionaltests for neovim → https://git.io/fh9tJ
Dagger has joined #nixos
<__monty__>
Avaq: What's the contents of that file? Sounds like an attribute is missing.
<ottidmes>
Avaq: I don't know enough about nixops to really help you with your problem, I have used it, but my use case was managing existing machines on NixOS, but have you tried --show-trace or --debug for more info
alex``` has joined #nixos
<{^_^}>
[nixpkgs] @teto closed pull request #54942 → lua: use lua environments for neovim/mpv → https://git.io/fhM3I
<{^_^}>
#54970 (by ElvishJerricco, 5 days ago, open): genericCallCabal2nix: init
<clever>
Avaq: line 63 of default.nix is dealing with localSystem, an arg to nixpkgs, i suspect your nixops version is too old to work on your nixpkgs version
nikola_i has joined #nixos
o1lo01ol1o has quit [Ping timeout: 244 seconds]
<__monty__>
clever: I found out about krops yesterday, have you evaluated it in comparison to nixops?
<clever>
__monty__: never heard of krops
<Taneb>
I've been looking into krops, since rougly yesterday
<{^_^}>
nixops#865 (by mogorman, 1 year ago, open): Key/Value backend
<wedens>
clever: so, after initial deployment, I upload my key and that's it?
<clever>
wedens: i would put your own ssh pubkey into the deployment files
<clever>
just users.users.root.openssh.authorizedPublickeys i think it was spelled
<Taneb>
"Warning: If you are using NixOps then don't use this option since it will replace the key required for deployment via ssh."
<clever>
Taneb: nixops appends to the list, it doesnt force the value
nikola_i has joined #nixos
lassulus has joined #nixos
<Taneb>
clever: I'm quoting the docs for users.users.<name?>.openssh.authorizedKeys.keys
<ottidmes>
Taneb: you can always export the keys from nixops and add them explicitly to your machine, then it does not matter if you remove nixops add-hoc adding of the keys
<clever>
Taneb: i would say those docs are wrong
<{^_^}>
[nix] @edolstra pushed to master « Revert "Restore parent mount namespace before executing a child process" »: https://git.io/fh9tn
sigmundv_ has joined #nixos
<Avaq>
> i suspect your nixops version is too old to work on your nixpkgs version -- @clever
<Avaq>
But my nixops is installed through environment.systemPackages, using nixos-unstable channel. It should be on the latest version, right? ('nixops --version' shows 1.6)
<{^_^}>
error: syntax error, unexpected '@', at (string):218:77
reinhardt has joined #nixos
johanot has joined #nixos
griff_ has joined #nixos
agander_ has joined #nixos
agander__ has joined #nixos
<Avaq>
clever: ^
agander has quit [Ping timeout: 246 seconds]
simukis has quit [Quit: simukis]
agander_ has quit [Ping timeout: 244 seconds]
<clever>
Avaq: there is a chance that the version in nixos-unstable, doesnt actually work on nixos-unstable
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<tommyangelo>
Does anyone have a config for a Dell XPS 15 5970 ? New company laptop, but it seems to be having issues with nouveau, and maybe the kernel too
klntsky has quit [Ping timeout: 256 seconds]
<Avaq>
clever: How does deployment.nix correspond to trivial.nix and trivial-vbox.nix in the manual?
<Taneb>
I do love it when nix-store --optimize frees 200GB
klntsky has joined #nixos
<Taneb>
I should set it to optimize automatically
<clever>
Avaq: replace deployment.nix with the list of files you originally gave to `nixops create`
<Avaq>
Thanks! :)
<Avaq>
clever: That seems to have done the trick. At least the error is gone from 'nixops info'! :)
<ottidmes>
Taneb: lol, that is more than 4x the total size of my NixOS installation on my desktop
<Avaq>
And it seems to be deploying as expected :D
<clever>
Avaq: that also confirms it is an issue with nixpkgs, but you can at least work around it now
<Taneb>
ottidmes: this is my work desktop, we use Nix very heavily. It's also a remote builder for our Hydra server, for example
<Avaq>
clever: So in effect, what I've done, is modify the nixpkgs version that's going to be used inside my deployments, because the unstable nixpkgs is not compatible with nixops?
<ottidmes>
Taneb: yeah makes sense, its that my install is old, 500G SSDs are cheap nowadays
<clever>
Avaq: yep
<clever>
Avaq: i also do this even when problems dont exist, so i can strictly control the version used in the deployment
<clever>
Avaq: so i dont have unplanned upgrades, caused by the host doing nix-channel --update
<Avaq>
clever.
<ottidmes>
clever++
<{^_^}>
clever's karma got increased to 87
<clever>
the -I nixpkgs=foo is one of the things nixops stores in the state file
<clever>
so that part will get lost if you move to another machine
<wedens>
why it can't be specified in expression?
<clever>
wedens: it can, but the current implementation has a heavy performance cost
<clever>
so i just leave a note to myself in a comment, on what to `nixops modify` to
<Avaq>
The deployment failed (virtualbox.service: Main process exited, code=exited, status=1/FAILURE), after some time; I'll try to figure out what happened there.
<clever>
Avaq: is virtualbox enabled on the host and already working?
<Avaq>
(I've been running your requested commands on the host)
<clever>
maybe check on both, just in case
<Avaq>
no entries on the host
<Avaq>
I've not had shell access to the vbox thus far
<clever>
not sure what else to check
<clever>
ive never tried using the vbox backend
<Avaq>
I'm just trying to get a basic example working to gain some understanding. :P
<Avaq>
The vbox example is the first in the nixops manual.
Jetien has joined #nixos
tzemanovic has quit [Ping timeout: 268 seconds]
<clever>
Avaq: if you have a working nixos install on a second machine, you can experiment with the none backend instead
<clever>
(and that second "machine" could just be a vbox copy you installed manually, from the ISO or pre-made vbox image)'
Guest65019 has quit [Ping timeout: 240 seconds]
<Avaq>
I have nixos on a raspberry pi, but not currently with me. I intended to experiment with it in my own time, but I figured I'd gain some experience with another backend first.
klDen[m] has joined #nixos
<Avaq>
clever: Should my deployed machine have a root user and password by default? I want to try logging in to it
Judson has joined #nixos
<clever>
Avaq: root pw can be used, but it is prefered to have an ssh key on root
agander_ has joined #nixos
elgoosy has joined #nixos
Judson is now known as Guest60340
<Avaq>
Ohh, I can ssh into it with 'nixops ssh' :)
<clever>
ah, then maybe it was deploying just fine, and some of the misc services are just broken
<clever>
and then it gives a warning when its done
<clever>
check the guest journal, and see whats up there
<Avaq>
systemctl there does show red
<clever>
and check the journal for the things in red
nikola_i has quit [Remote host closed the connection]
<Avaq>
# ls -l /run/{booted,current}-system/kernel
<Avaq>
^ I went back to exec that, and saw different version, which seemed to be what you were looking for. Still prying, one sec
agander_ is now known as agander
<clever>
Avaq: nixops created a vbox image, from a base disk image, running 4.4.24
<clever>
Avaq: nixops then deployed changes, that upgrade you to 4.14.97
<clever>
and it tried to load 4.14 kernel modules on 4.4
<clever>
rebooting the guest `nixops reboot` should solve it
hedning has quit [Quit: hedning]
goodwill has quit [Quit: ...]
reinhardt has quit [Quit: Leaving]
tzemanovic has joined #nixos
<Avaq>
okay, did that
<Avaq>
now the kernel versions are indeed the same
<Avaq>
and systemctl shows green
nikola_i has quit [Ping timeout: 244 seconds]
<Avaq>
I don't really understand what we just did. But thank you! :)
elgoosy has quit [Quit: Leaving]
<clever>
Avaq: when nixops first did a deploy, it upgraded the kernel, and then you had an old kernel and new modules, so it failed to load the modules
o1lo01ol1o has joined #nixos
vonfry has joined #nixos
o1lo01ol1o has quit [Ping timeout: 268 seconds]
Guanin has quit [Remote host closed the connection]
tzemanovic has quit [Remote host closed the connection]
Jackneill has quit [Ping timeout: 240 seconds]
Makaveli7 has joined #nixos
tzemanovic has joined #nixos
periklis has quit [Remote host closed the connection]
tzemanov_ has joined #nixos
phreedom has quit [Remote host closed the connection]
vonfry1 has joined #nixos
dave_uy has quit [Ping timeout: 245 seconds]
phreedom has joined #nixos
tzemanovic has quit [Ping timeout: 252 seconds]
LnL has joined #nixos
emacsomancer has quit [Ping timeout: 245 seconds]
aramiscd has quit [Ping timeout: 245 seconds]
ryantm has quit [Ping timeout: 245 seconds]
LnL is now known as Guest5079
vonfry has quit [Ping timeout: 268 seconds]
dave_uy has joined #nixos
aramiscd has joined #nixos
emacsomancer has joined #nixos
griff_ has quit [Quit: griff_]
Guest5079 has joined #nixos
Guest5079 has quit [Changing host]
Jetien has quit [Ping timeout: 245 seconds]
hyper_ch2 has joined #nixos
vonfry1 has quit [Quit: WeeChat 2.3]
<kai_w>
Is it possible to test packages (sandboxed) that depend on a graphics driver being available?
ryantm has joined #nixos
tzemanov_ has quit [Remote host closed the connection]
Jackneill has quit [Remote host closed the connection]
<Avaq>
What does it mean to supply multiple .nix files to nixops? Are the configurations merged?
<clever>
Avaq: yeah, they are all merged together
<clever>
if 2 files define different machines, then the deployment just contains both set of machines
<Twey>
kai_w: You can pass in LIBGL_DRIVERS_PATH to any libGL-using program explicitly, and point it to mesa_drivers or something
<clever>
if they define settings for both, then the options are merged at the nixos level, the same as using a list in imports
Jackneill has joined #nixos
<Twey>
kai_w: I don't think that tries to break the sandbox
<Avaq>
clever: Clear, thanks.
<kai_w>
Twey: In my case I need cuda, which happens to live in /run/opengl-driver/lib (and must match the loaded kernel module)
<kai_w>
that last bit is where I'm getting stuck, I don't think there's a way for nixpkgs to choose the 'right' cuda version automatically
<Avaq>
Another question; it seems that the directives in the .nix files for nixops have a mixed type. They can be a record, or a function. Is that correct?
<Avaq>
Mixed types confuse me :p
<clever>
Avaq: yeah, the main machinename = thing; stuff are just nixos modules
<clever>
Avaq: configuration.nix is a nixos module
<clever>
and they can either just be a raw set, or a function that takes inputs
<clever>
so you can just do machine = { pkgs, config, ... }: { ... };
<clever>
and paste in the contents of your configuration.nix file
<srhb>
And the set can be either of the module forms (eg, imports, options, config, or just the config, ... , to make things even better..
<Avaq>
ah
<{^_^}>
[nixpkgs] @christoph00 opened pull request #55269 → Steam: run update-runtime.py → https://git.io/fh9qh
<clever>
yayforj: which package are you overriding?
<yayforj>
no matter
<clever>
yayforj: the overlay must return a set of package = override;
<clever>
why are you using mapAttrs?
<yayforj>
i tried different
<clever>
ah, i see why
<yayforj>
clever
<yayforj>
yeah
rauno has quit [Ping timeout: 268 seconds]
<yayforj>
map name version -> map name package
<yayforj>
but when i use `callCabal2nix` it works
<clefru>
Is there a good place to publish/advertise my nix-overlay? It contains the amazing number of 6 packages, but I'd like to make it discoverable somehow that people don't repackage those things
<yayforj>
link please
<clever>
yayforj: i see the problem, you cant use callHackage or callCabal2nix when defining vector-algorithms
<yayforj>
clever: why not?
<clefru>
A web-based package search like Arch's AUR would be nice
<clever>
yayforj: because cabal2nix depends on vector-algorithms
agander has quit [Read error: Connection reset by peer]
<clever>
yayforj: and now you need cabal2nix, to define vector-algorithms, to build cabal2nix
<clever>
yayforj: you will need to manually run `cabal2nix cabal://vector-algorithms-0.8.0.1` (i think it was), stick that into a .nix file, and callPackage it
<yayforj>
clever: but funny enough, callCabal2nix works in that context
o1lo01ol1o has joined #nixos
nodyn has quit [Read error: Connection reset by peer]
Makaveli7 has joined #nixos
agander_ has joined #nixos
<yayforj>
clever: you're right, i tried to override `webdriver` package and all works
o1lo01ol1o has quit [Ping timeout: 240 seconds]
<clever>
yayforj: if you where not using mapAttrs, you could also make an override for vector-algorithms but call the attr my-vector-algorithms
<yayforj>
clever: how does `mapAttrs` change the situation?
<clever>
it allows you to decouple the attrname, and the cabal name
<clever>
so you could test the override itself, without impacting cabal2nix
schjetne has quit [Remote host closed the connection]
agander has joined #nixos
agander_ has quit [Ping timeout: 250 seconds]
ng0 has joined #nixos
genesis has quit [Ping timeout: 252 seconds]
genesis has joined #nixos
<Avaq>
Can I in-line a file in my nix configs? Something wants a file, which it will cat to an interpreter, I don't want to create extra files.
<tilpner>
And if you leave "foo" out of the path, you can even do nix-env -iA hello
realrokka_ has quit [Ping timeout: 245 seconds]
Avaq has joined #nixos
danielrf has quit [Quit: WeeChat 2.3]
realrokka has joined #nixos
<ottidmes>
any vscode users out here that have run into the problem where vscode default browser is that of a /nix/store path no longer existing? its very annoying, I cannot open links because it keeps pointing to an older no longer existant google chrome version
<clever>
ottidmes: try grepping ~/.cache/ for the hash of that chrome
<clever>
ottidmes: and ~/.share
<clever>
~/.local/share/applications for example has some related stuff
<Twey>
clever: But I'd assume nix-channel is the recommended way to set nixpkgs to a local nixpkgs checkout, as mpickering asked
o1lo01ol1o has quit [Ping timeout: 250 seconds]
<clever>
Twey: you would have to re-run nix-channel --update, every time you change the local nixpkgs
<clever>
Twey: but if you add a file to .nix-defexpr, it will parse that file every time you nix-env -iA
<clever>
so it wont need any update operations
goibhniu has quit [Ping timeout: 245 seconds]
<ottidmes>
clever: found some references to it, but for teamviewer, not vscode
griff_ has joined #nixos
<{^_^}>
[nixpkgs] @jtojnar merged pull request #54380 → SRT support for gstreamer → https://git.io/fhu9Y
<ottidmes>
clever: found the problem, xdg-open reports the old Google Chrome
<clever>
ottidmes: that uses some state somewhere in $HOME, strace might find it
<mdash>
ottidmes: I got optimistic and tried to build bitwarden_rs 1.6.1 with nightly from the mozilla overlay. it's telling me it can't find crate for `std`. :)
<ToxicFrog>
jboy: it's not meant to, but it's still possible for things to be packaged incorrectly
<ToxicFrog>
It looks like the Marp package¹ downloads the binaries and then uses patchelf on them, so if the invokation of patchelf is wrong, you have a bad time
<jboy>
yes, I just saw that as well, ToxicFrog. Maybe I'll try changing that to `autopatchelf`.
<tilpner>
mdash: Your cargoSha256 wasn't accepted here, try breaking it and building again
<{^_^}>
[nixpkgs] @7c6f434c merged pull request #55266 → Add checkPhase for neovim (disabled by default) → https://git.io/fh9tJ
<{^_^}>
[nixpkgs] @7c6f434c pushed commit from @teto to master « Add checkPhase for neovim (disabled by default) (#55266) »: https://git.io/fh93h
<tilpner>
And after I replaced it with the correct hash, the patch doesn't apply
<tilpner>
Ahh, yes, trailing newline
daniele- has joined #nixos
<mdash>
tilpner: boo
<tilpner>
mdash: Found your problem!
<mdash>
tilpner: alrighty
<tilpner>
mdash: You need to pass { rustc = rust_nightly.rust; inherit (rust_nightly) cargo; } to makeRustPlatform
<tilpner>
And you'll want to fix those "rev = master" lines eventually
<mdash>
of course. :-[
<tilpner>
Doesn't need recurseIntoAttrs here
<tilpner>
And might not need RUSTC_BOOTSTRAP=1;
<mpickering>
after I set `nix.nixPath` in my `configuration.nix` do I have to reboot to see those changes reflected in `NIX_PATH`?
<tilpner>
Unfortunately you have to at least login again
<tilpner>
You can set your NIX_PATH to point to symlinks to prevent future re-logins if you ever decide to move your nixpkgs repo somewhere else
<mpickering>
that seems to have done something thanks
<mpickering>
that sounds a bit complicated for me, perhaps in a few years I will do that
<mpickering>
So now my `NIX_PATH` has `/home/matt/.nix-defexpr/channels` first, that seems wrong to me?
<tilpner>
Depends. What did you set nix.nixPath to?
<tilpner>
Also, does your bashrc still alter your NIX_PATH? (or similar startup files)
<ottidmes>
clever: BTW it was XFCE that cached the google-chrome.desktop file causing xdg-open to use the wrong Exec path, fixed it by removing the cached .desktop file, but since I dont actually use XFCE, I will just remove it from my system the next reboot
<clever>
ottidmes: ah
<tilpner>
mdash: Confirmed to finish build without RUSTC_BOOTSTRAP
<mdash>
tilpner: thank you. I thought the overlay was similar enough to the rust derivation in nixpkgs to call makeRustPlatform the same way as in all-packages. :)
<mpickering>
but this time I have no channels installed so it's harmless
ng0 has joined #nixos
<tilpner>
mpickering: It's definitely possible to have a NIX_PATH without entries like that. I assume your .bashrc adds that
<mpickering>
I don't have a .bashrc
<tilpner>
Hmm, right, if you're on NixOS you wouldn't have gone through the Nix installation script
xkapastel has joined #nixos
Rusty1 has joined #nixos
ikitat has joined #nixos
waleee has joined #nixos
<mpickering>
I still don't get why it is there
<mpickering>
as it seems to override the default
<mpickering>
I set `nix.nixPath` and the changes are definitely in `NIX_PATH` but this additional entry
<mpickering>
the same thing broke my old installation as well
<tilpner>
Check your /etc/profile and read referenced files
<tilpner>
Especially the set-environment script called from there; it sets NIX_PATH
<tilpner>
Chec if it happens when you login as another user
<mpickering>
yes, it sets it with this additional directory
Makaveli7 has joined #nixos
<mpickering>
tilpner: Where is this file generated? I want to find the commit which introduced this change
<tilpner>
First run: nixos-option nix.nixPath
<mdash>
tilpner: yeah it's curious that the recommended config says master, Cargo.lock has a revision
<tilpner>
mpickering: Does the output of that have multiple paths in "Defined by"?
<tilpner>
Does the value include the unwanted path?
griff__ has joined #nixos
<mpickering>
tilpner: No, defined just in configuration.nix and doesn't include the unwanted path
<tilpner>
mdash: As I said earlier, your cargoHash wasn't accepted here. I blamed that on the usage of master, but if it's not that... you'll want to find out what is
Avaq has quit [Ping timeout: 245 seconds]
<mpickering>
if you look int the set-environment file, it checks to see if `.nix-defexpr/channels` exists and then puts that first in the NIX_PATH if it does
nikivi has quit [Remote host closed the connection]
<mpickering>
I am failing to set up a window manager, just getting a blank vty window with a blinking cursor
nikivi has joined #nixos
nschoe has joined #nixos
<{^_^}>
[nixos-homepage] @tilpner opened pull request #263 → Minify packages.json with jq → https://git.io/fh9nc
elgoosy has joined #nixos
yayforj has joined #nixos
<yayforj>
how can i build a haskell source with cabal slover?
glesica has quit [Quit: Peace out.]
<mpickering>
yayforj: Is that a nixos question?
<yayforj>
partially
<yayforj>
i ran cabal2nix in source
<yayforj>
but it breaks
<yayforj>
because there are bounds on everything
<yayforj>
which cabal2nix doesn't pick up
<yayforj>
and deps from nixpkgs don't satisfy it
<tilpner>
samueldr: pr above ^
<wedens>
yayforj: typical strategy is to override speicifc packages with required versions
<mpickering>
yayforj: You can use the haskell.lib.doJailbreak function
<samueldr>
tilpner: I'll look and approve later (obv. can't merge)
<mpickering>
which will ignore version bounds
<yayforj>
mpickering: i did so, but sources don't compile coz if breaking changes
<mpickering>
ok then you have to override the haskellPackages set with the right versions
<yayforj>
mpickering: i'm wondering if there's a way to make nix pull deps from hackage with cabal
danielrf has joined #nixos
<wedens>
you can just use cabal then :)
<yayforj>
mpickering: it's a lot of versions
<yayforj>
wedens: fair enough, but how do i put it into a store?
<aanderse>
i'm looking to add some default values to services.logrotate.config (so rsyslog files would be rotated), but i'd like the user to be able to override in case they want to do something different. i thought there might be some sort of way to allow this and make it easy for the end user, but i'm not sure what
<yayforj>
wedens: have another project depending on this one
<wedens>
you don't. you either manage dependencies via nix or cabal
<aanderse>
the idea is in the rsyslog module i'd add something to services.logrotate.config, but then the user could explicitly disable what the rsyslog module adds, and put their own in
<aanderse>
but the user wouldn't want to entirely override the contents of services.logrotate.config because other services could add stuff to that
<aanderse>
anyone have any ideas?
NemesisD has joined #nixos
civodul has quit [Quit: ERC (IRC client for Emacs 26.1)]
johanot has quit [Quit: WeeChat 2.2]
<{^_^}>
[nixpkgs] @KAction opened pull request #55279 → libcdb: init at 0.7.8 → https://git.io/fh9nH
<{^_^}>
[nixos-homepage] @edolstra pushed 2 commits to master: https://git.io/fh9cz
eadwu has joined #nixos
orivej has quit [Ping timeout: 250 seconds]
tzemanovic has quit [Remote host closed the connection]
fresheyeball has quit [Quit: WeeChat 2.2]
daniele- has quit [Quit: daniele-]
tzemanov_ has joined #nixos
orivej has joined #nixos
<tilpner>
clever: Have you seen error: a 'i686-linux' with features {} is required to build '/nix/store/...', but I am a 'x86_64-linux' with features {benchmark, big-parallel, kvm, nixos-test} anywhere when setting extra-platforms?
<clever>
tilpner: thats more about disabling the i686 support
<tilpner>
I think I might prefer it this way
<tilpner>
Let the user decide what platforms Nix should try to build for, instead of implementing a dozen detection mechanisms that might get it wrong
<tilpner>
(Not a dozen, just two for now)
tzemanov_ has quit [Remote host closed the connection]
<dhess>
So while I like NixOps, I've decided I'm not a fan of its "implicit" configuration. By this I mean the bits of the config that it creates/imports automatically, like the system closures, the deployment modules, etc.
<dhess>
Is anyone using NixOps more "explicitly" ?
<tilpner>
I wrote my own deploy script, like many other people
<dhess>
tilpner: a la Awake Security's nix-deploy?
<tilpner>
Yes. I was using nix-deploy, and then it broke somewhere around the Nix2 switch, and then I decided I could just write the 4 lines of code myself
<dhess>
tilpner: is your script using NixOps, or just nix-instantiate/nix-copy-closure etc like nix-deploy does?
<{^_^}>
[nixpkgs] @flokli pushed 2 commits to release-18.09: https://git.io/fh9CU
<tilpner>
It nix-instantiate/copy-closure with a module so that I can define the device IP/domain inside its config
<tilpner>
It can't create any devices
<tilpner>
And at the time I didn't know nixos-rebuild could deploy to remote machines as well
Ariakenom has joined #nixos
<dhess>
tilpner: right. I considered that myself, but I like the parts of NixOps that deploys secrets and manages the fleet automatically.
<dhess>
so I don't want to throw out the baby with the bathwater. I just want the config to be explicit, mainly so that I can build the system closures in Hydra.
<dhess>
(I also no longer use the NixOps provisioning, but that doesn't get in the way so I can just ignore it.)
<tilpner>
I also didn't like the state thing. Maybe someone can explain why it needs to manage its own ssh keys, when I have some already
<dhess>
tilpner: yes I agree with that as well. Especially given that the other state basically doesn't seem to matter at all.
<tilpner>
It makes some sense if you let it create your infrastructure, but not if you just hand it an ip address
simukis has quit [Quit: simukis]
<dhess>
exactly
<clever>
tilpner: when you have 100 machines that you basically never ssh into, and dont want to re-deploy every time somebody quits and you have to cut them off
<dhess>
clever: that's a good point.
<tilpner>
clever: So how do you know they're not holding on to a copy of the nixops state?
<dhess>
clever: on the other hand... yeah what tilpner said :)
<clever>
tilpner: that is a bug that needs to be fixed in nixops
<clever>
re-keying things
<tilpner>
I guess rekeying could be implemented without a system evaluation, so I guess you have a point
<dhess>
yeah it could be part of send-keys
Denommus has joined #nixos
<mpickering>
X server fails to start with "intel: waited 2020 ms for i915.ko driver to load", has anyone experience with this?
<clever>
dhess: nixops bakes its own public key into the nixos config it builds
<clever>
dhess: so send-keys cant really change things
<cransom>
i stopped nixops'ing partly due to everything i deploy is in an auto scaling group and worked out my own method for secret storage. needing to nixops when we scale up for load wasn't a feasible option
iqubic has joined #nixos
<dhess>
anyway I certainly don't want to rewrite key deployment or fleet management in my own wrapper, so I think the right move is to keep using NixOps.... I just don't want it to be so clever (pardon the pun)
<dhess>
clever: good point.
devalot has joined #nixos
<dhess>
tilpner: how do you deploy secrets with your wrapper?
dbmikus has joined #nixos
<tilpner>
dhess: I... don't :/
<tilpner>
cransom: Is your "own method for secret storage" viewable somewhere?
dermetfan has quit [Ping timeout: 240 seconds]
<dhess>
I'm also thinking about moving to something like AWS Parameter Store for secrets, so that is something I will probably eventually stop using in NixOps anyway.
<gchristensen>
nice
<dhess>
though for non-EC2 machines you still need to deploy the .aws credentials
<devalot>
I'm running nix-build in a Docker container. I'd like to cache /nix/store entries. Should I just mount /nix/store from a directory on the host?
Henson has joined #nixos
<dhess>
however that could be just a simple scp/rsync step as part of the wrapper
<cransom>
tilpner: it's not that exciting. i take some json files that dhall makes, use KMS to encrypt them and put them in the store. when the machine boots, it has permissions to ask kms for a key to decrypt them, and it sticks them in /run/keys for a service to parse them out for services to use them
<dhess>
so then it's really just the fleet management part of NixOps that provides value.
<dhess>
cransom: hey that sounds pretty exciting to me :)
<cransom>
but i should write that down somewhere.
<dhess>
I like every part of that (except maybe the part about the store)
reinhardt has quit [Ping timeout: 245 seconds]
<tilpner>
cransom: Is KMS an AWS thing?
<cransom>
tell me about the qualms about the store
<dhess>
cransom: I don't want encrypted secrets to be world-readable.
<cransom>
and yes, KMS is an AWS thing. but it's just an encryption service. you could do somethign similar with vault/et al
goibhniu has quit [Ping timeout: 245 seconds]
<cransom>
dhess: i don't know how to ship a file into the cloud where it isn't world readable.
ixxie has joined #nixos
<Henson>
what's the proper way to initialize a per-user profile? I have some symlinks in my home directory that point to /var/nix/var/profiles/per-user/username/foo but can't remember how I did it.
<gchristensen>
nice
<dhess>
cransom: in my imaginging, you don't ship it "into the cloud." You just scp it to the host.
<gchristensen>
nice oops
<tilpner>
Henson: On NixOS? users.users.foo.packages
<gchristensen>
Henson: it is automatic
<dhess>
host(s).
<gchristensen>
Henson: those symlinks get created as soon as your shell starts for the first time
<Henson>
just with the nix package manager, not on NixOS
<cransom>
the decrypted secrets are in memory only, and have permissions to keep them from most other things. but if you get into the box, you'll probably find a way to get them easily enough. and if we find that, well, we patch bugs, rotate secrets, and redeploy
<Henson>
I know there's "nix-env -p" but can't remember if I make that operate on the profiles in /nix/var/... do adjust profile generations, or if I can operate on the symlink in my home directory that points to the most current generation in /nix/var/...
goibhniu has joined #nixos
<Henson>
do adjust -> to adjust
<gchristensen>
Henson: it sounds like you have another question lurking below the questions you're asking
<dhess>
cransom: it's like shadow password files -- just another part of defense in depth.
<cransom>
dhess: that's a valid option. but one of our requirements was that there as few external dependencies as possible when a host comes up. we came from a world where chef had to run and download from github, update apt, and reboot before it was even ready. the only thing we wait on now is secrets decryption and the new hosts are up in about a minute
<dhess>
(though I realize that if you're doing immutable users on NixOS, you're already storing your crypted passwords in the store :\ )
<cransom>
if we have to have some signaling where a machine boots and waits for secrets from some other service we manage, it's just not appealing.
<Henson>
gchristensen: heh, the underlying question is how can I use a profile directory other than ~/.nix-profile. Using "nix-env -p" does all the symlinking in my home directory, instead of tucking everything away in /nix/var/... and having the symlink just point to the current generation.
<dhess>
cransom: let me be clear that I'm not being critical. As I said I like it (and it's better than what I have now). It's just the one bit that sounds not "great."
<dhess>
I would take it I think over what I'm doing now.
<Henson>
gchristensen: I used to know how to do it, but I've forgotten
<cransom>
our users don't have passwords set at all, just ssh public keys
<dhess>
cransom: I've also considered doing that. Are they just NOPASS in sudoers?
<gchristensen>
Henson: why do you want to? nix-env will leave everything in /nix/var/nix
<cransom>
dhess: trade offs, trade offs everywhere. but yes, the system now is an objective 10x improvement over the last thing we had.
<gchristensen>
but with one symlink in ~
rauno has quit [Remote host closed the connection]
<dhess>
cransom: is this on GitHub or similar?
<Henson>
gchristensen: this is for deploying an isolated piece of software. I don't want this software to be in ~/.nix-profile, but in its own separate profile directory.
nschoe has quit [Ping timeout: 245 seconds]
tzemanovic has joined #nixos
<gchristensen>
ok, what for? I'm trying to get to root motivation
<cransom>
dhess: rather than passwords, `security.pam.enableSSHAgentAuth`.
<dhess>
cransom: oh hmm, I've never used that
<gchristensen>
dhess, cransom: my job-2 had it setup that after any user logged in, it would "taint" the node and schedule it for termination 24hrs later
<clever>
cransom: sudo
<Henson>
gchristensen: just for the sake of conceptual cleanness.
<clever>
cransom: sudo requires a password, and cant accept a forwarded agent
<gchristensen>
Henson: what are you trying to d?
<gchristensen>
do
<gchristensen>
like, the very very end goal
<cransom>
(and also, wheel users don't need a password)
<dhess>
gchristensen: I have boxes with state, can't do that, and I need to shell in and poke around sometimes :)
<yayforj>
mpickering: wow! stack2nix! that nix file tho... does it replace the whole haskellPackage?
<cransom>
which is all the users that can log in, really
hellrazor has quit [Ping timeout: 272 seconds]
<cransom>
gchristensen: i like the taint notion. at least in my current environment, the people that can log in and have the ability to update something, know better.
tzemanovic has quit [Ping timeout: 244 seconds]
Denommus has quit [Remote host closed the connection]
<Henson>
gchristensen: I figured out how I did it. I specified the /nix/var/.../foo path with the "nix-env -p" command, and just made a symbolic link from ~/foo to /nix/var/.../foo
<clever>
cransom: and line 27 is to deal with the fact that the aws backend of nixops, DOES NOT bake its own key into the nixos config!
<dhess>
cransom: so let me get this straight; I use that and put the user in wheel and then the user can ssh in and sudo with no password being set?
<clever>
cransom: it relies on the file that was created on the first boot
<cransom>
my god so many hilights
<clever>
cransom: none, and aws, handle the nixops generated key in different ways
iqubic` has joined #nixos
<clever>
dhess: thats what it looks like to me
dermetfan has joined #nixos
<dhess>
good to know. I'll probably do that.
<cransom>
dhess: it was a while back, i don't remember if `wheelNeedsPassword` explicitly is required for ssh auth sock sudo auth, you may want to check yourself.
<gchristensen>
cransom: how does it feel
<dhess>
it does make it easier to `sudo foo` without thinking though :)
xkapastel has quit [Quit: Connection closed for inactivity]
<yayforj>
clever: should i put it in an overlay then? whould it be a 'correct' way of using it?
<clever>
yayforj: just import that file, and use its result, instead of haskellPackages
ubert has quit [Quit: Leaving]
<cransom>
clever: re obey services.openssh.authorizedKeysFiles, it doesn't. it would be nice if it did, there should be an issue for it. but changing authorizedkeys location isn't particularly common so it's not much of a pain point for the vast majority.
<{^_^}>
[nixpkgs] @veprbl pushed commit from @scalavision to master « singularity: 2.6.0 -> 2.6.1 »: https://git.io/fh9Co
<cransom>
but as you make choices that take you off the heavily beaten path, its not surprising that you find bugs like that. it's just not a place lots of people go
lux1 has joined #nixos
danielrf has quit [Ping timeout: 252 seconds]
alex_giusi_tiri has joined #nixos
iqubic` has left #nixos ["ERC (IRC client for Emacs 26.1)"]
<clever>
cransom: the same can be said about what nix has done to basically every package in the world
<cransom>
amen.
reinhardt has joined #nixos
<cransom>
some people do some really weird things to deliver software
xkapastel has joined #nixos
ng0 has quit [Quit: Alexa, when is the end of world?]
<dhess>
hehehe
sigmundv_ has quit [Ping timeout: 240 seconds]
ThatDocsLady has quit [Ping timeout: 250 seconds]
<alex_giusi_tiri>
Hi! When Nix is compiling a package, where would its build dir be? I ran, in preBuild=''pwd'', but the path is absolute, such as '/build/<pkg-name>/build/'. I have this error where a './configure' file starts with '#!/usr/bin/env bash', but '/usr/bin/env' would not be found. I tried sed substituting it with just 'env', 'bash', and with '/run/current-system/sw/bin/bash', and now, I am trying with '${bash}', but it seems to me
<alex_giusi_tiri>
a chroot (because there is no actual '/build' folder in the root), and cannot access outside files.
<alex_giusi_tiri>
with ${bash} I got /nix/store/a9i0a06gcs8w9fj9nghsl0b6vvqpzpi4-bash-4.4-p23/bin/bash: ./configure: /nix/store/a9i0a06gcs8w9fj9nghsl0b6vvqpzpi4-bash-4.4-p23: bad interpreter: Permission denied
<elgoosy>
however, doing nix-shell --attr env default.nix, i get an error error: cannot auto-call a function that has an argument without a default value ('mkDerivation')
ikitat has joined #nixos
<srhb>
alex_giusi_tiri: You should not rely on knowing the temporary path anyway. You can use patchShebangs to fixup any configure scripts in preConfigure.
_kwstas has joined #nixos
_kwstas has quit [Client Quit]
<{^_^}>
[nixpkgs] @Infinisil merged pull request #51980 → nixos/munin: New options (and some bugfixes) for service configuration → https://git.io/fpFYI
<srhb>
alex_giusi_tiri: You were close to the solution though, however, ${bash} is just the path to the base directory of bash in the nix store. With /bin/bash appended, it would have worked. However, prefer patchShebangs to get the bash from stdenv :)
<clever>
alex_giusi_tiri: you can also run it on a directory, and all files in the directory will be patched
<clever>
alex_giusi_tiri: if the script is ran during the build, then patchPhase is a good place, if its at runtime, then any phase can work
<alex_giusi_tiri>
oh, that's very convenient!
<alex_giusi_tiri>
thank you
Denommus has quit [Remote host closed the connection]
danielrf has joined #nixos
agander has quit [Ping timeout: 250 seconds]
bpa has joined #nixos
ixxie has quit [Ping timeout: 240 seconds]
agander has joined #nixos
ixxie has joined #nixos
peacememories has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
dermetfan has quit [Ping timeout: 250 seconds]
Church_ has joined #nixos
yayforj has quit [Ping timeout: 245 seconds]
<Henson>
what if I want to use Nix in single-user mode as the root user? The documentation says that the single-user mode doesn't support it, but I don't need the functionality or complexity of the multi-user setup, I just want to use single-user nix as root. Can I just install it as another user and then chown /nix to root and make a nix-profile for root? Or will that somehow cause problems?
<Henson>
Nix package manager, that is
<clever>
Henson: if nix has root when doing a build, it assumes the nixbld group exists, and has build users within it
<clever>
i'm not sure if that has an off switch
<Henson>
clever: ok
<mpickering>
So now I have a window manager, it just loads xterm rather than xmonad
<Henson>
clever: ok, so how does one install nix as root in the first place, in order to run the nix daemon to make multi-user mode work?
<clever>
mpickering: the default desktop manager is xterm, you need to enable another one, and then either select it at the login screen, or disable xterm
<mpickering>
window manager I set on line 79 to xmonad
<mpickering>
enable xmonad on line 77
<clever>
mpickering: it may also help to set desktopManager.xterm.enable = false;
<clever>
and nixos-rebuild tries to avoid restarting X, because it interupts everything, so you need to `systemctl restart display-manager.service` manually after these kinds of changes
<mpickering>
I have just been rebooting before
<mpickering>
Will try that
<Church_>
I really need to get my lobsters package mainstreamed,
<mpickering>
after logging in, I now have a screen with a cursor
<mpickering>
I'm not sure though how the logic screen part is configured
dermetfan has joined #nixos
<clever>
mpickering: if you ssh in from another box, and run `ps -eH x`, is xmonad running, anything else near it?
<gchristensen>
it'd be nice if nixops made GC roots for its deployment builds
<clever>
gchristensen: it does, when you turn on rollback
<Avaq>
What happens if I use nixops to deploy to a targetHost that runs NixOS? Will that machine create a new main generation? Is it possible, and what does it mean to deploy multiple NixOps machine specs to the same targetHost?
<clever>
Avaq: when you do a `nixops deploy`, it will completely replace the old system, by just copying the new stuff into /nix/store and updating /run/current-system
<clever>
Avaq: and setup a new generation, as-if you had done a nixos-rebuild
<lucus16>
Hmm, how would I go about creating a new cross compiling stdenv, specifically for riscv32 embedded?
<clever>
Avaq: if you missed anything vital, it will behave the same as if you had simply removed that from configurationnix
<lucus16>
shlevy: Do you know something about that?
<Avaq>
clever: I see. So the configuration.nix inside that machine becomes irrelevant
<clever>
Avaq: yep
<clever>
Avaq: one critical thing when migrating from nixos to nixops, is to ensure you setup fileSystems. and boot. correctly, since it ignores configuration.nix and hardware-configuration.nix
aswanson has quit [Quit: WeeChat 2.2]
<Avaq>
Right. That makes sense.
<{^_^}>
[nixos-hardware] @nocent closed pull request #91 → apple: Add support for MacBook 3,X → https://git.io/fhtYi
<digitalgrease>
can nixos do running kernel updates, similar to ksplice?
<clever>
behind the scenes, nixos-rebuild always obeys <nixos-config>, which defaults to /etc/nixos/configuration.nix
<clever>
digitalgrease: not currently, and i havent seen any plans to add it
<clever>
Avaq: but, the above module, will remap <nixos-config> to a file that just throws an error
<clever>
Avaq: so, nixos-rebuild will fail with a clear error, rather then undoing everything nixops did
<clever>
Avaq: it will also remap <nixpkgs> to whatever nixpkgs was last used to deploy
ottidmes has quit [Quit: WeeChat 2.2]
<Avaq>
Ah, so if me or someone would be using that machine, knowing it's NixOS, this would prevent them from making the mistake of doing a nixos-rebuild to say, update their software
<clever>
lucus16: cross-compile package sets are auto-generated for nearly all platforms (maybe all?) and you can then use that to cross-compile almost any package
<clever>
collision between `/nix/store/bfw27ldvy66v6mhwkzillzz1ddfz5lwf-python3.7-msgpack-0.5.6/lib/python3.7/site-packages/msgpack/__pycache__/exceptions.cpython-37.pyc' and `/nix/store/9dkp2r0hsr2jkd44qinw6m7fk8lsx2xy-python3.7-msgpack-python-0.5.6/lib/python3.7/site-packages/msgpack/__pycache__/exceptions.cpython-37.pyc'
Thra11 has quit [Quit: WeeChat 1.9.1]
<pie_>
laas: have you found anything existing
<clever>
builder for '/nix/store/4y42l6irgbdszk6npn1a9mz5hr34l7jw-python3-3.7.2-env.drv' failed with exit code 25
<clever>
jonreeve: you have conflicting files in a python thing your installing
<clever>
jonreeve: something you put into systemPackages
migy has quit [Quit: migy]
migy has joined #nixos
aswanson has joined #nixos
<laas>
pie_: no I haven't
<laas>
that's why I asked
<jonreeve>
clever: How do I resolve the conflict?
<clever>
jonreeve: it looks like you have 2 different variants of msgpack in that python env
<laas>
it seems like I'll have to make my own, but I'm still a noob to nix
<tilpner>
Or... help whoever works on that, I guess
<jonreeve>
clever: Is there a command I can run that will fix it?
<clever>
jonreeve: you need to edit your configuration.nix, to not put 2 msgpack's into the python env
<{^_^}>
[nixpkgs] @vcanadi opened pull request #55284 → pythonPackages.notedown: init at 1.5.1 → https://git.io/fh9lb
<jonreeve>
clever: I don't even have one msgpack
moosefish has joined #nixos
<clever>
jonreeve: do you have anything python based in your systemPackages?
<jonreeve>
Yep. Lots of stuff: pandas, matplotlib, jupyter, nltk, spacy...
<clever>
jonreeve: try commenting one out, and see if `nixos-rebuild build` passes or fails
nokomprendo has quit [Quit: Leaving]
<moosefish>
I've just installed nix multi-user on MacOS (Mojave) and most interesting nix commands (eg: 'nix-env -i nix-info') die before finishing with "error: unexpected end-of-file". Can't seem to find any information searching around, is there somewhere I can look to start debugging this?
<tilpner>
moosefish: Pass multiple -v to them
<tilpner>
-vvv (etc.) works too
<pie_>
laas: if youre not in a rush i think i have something with some patches at home, ill be home in an hour or twi
Czen has quit [Ping timeout: 240 seconds]
tzemanov_ has joined #nixos
<jonreeve>
clever: That seems to work, but I need all of those packages to work together for a python project
<clever>
jonreeve: that will at least narrow down which one is broken, and needs more attention on it
<jonreeve>
LnL: Can you get that gist to work, if you replace `pyyaml` with `spacy`?
<jonreeve>
I can't install this library for the life of me
<LnL>
simplest things you can do is do a single-user install or nix 2.1.3
<moosefish>
let's give 2.1.3 a go then, danke
<aszlig>
laas: only a default.nix for lutris itself, but it was for the work on ripping apart the lutris runners/installers so that it uses Nix, so it's not in a working state (yet)
<aszlig>
laas: and i've been packaging games for nix since a while, so i thought lutris could be a good fit if it handles all the steam, gog, itch, whatnot-integration well
<pie_>
clever: I've run into something quite strange.
Mateon1 has quit [Ping timeout: 250 seconds]
Mateon3 is now known as Mateon1
<aszlig>
although itch.io is one of the less annoying platforms because you can get the exact revision you ask, while for example for gog* and HB you always get the latest revision and thus the sha won't match anymore
<infinisil>
slabity: cransom: I wish people would pay a bit more attention to backwards compatibilty and deprecation..
<infinisil>
It's annoying when your code suddenly stops working without any indication why
nschoe has quit [Ping timeout: 252 seconds]
<aszlig>
* -> only applies for gnu/linux, because for win/osx there is the gog galaxy api
<pie_>
clever: https://bpaste.net/show/ae0b64b6586e The first rebuild works, the second rebuild errors as expected, but I dont understand how it capable of building anything at all.
<zupo>
Hey all! I'm stuck with a Raspberry PI with Nixos installed. The Pi is configured to have a static IP (10.3.9.10) when I am home, but I away for some time, on a different network, where it needs the IP to be 192.168.8.10. I change the ip in configuration.nix, but then nixos-rebuild switch fails because it cannot download cache.nixos.org. Can I tell nixos-rebuild not to use the Internet? Or can I tell the system to use a different IP temporarily via the terminal?
moosefish_ has joined #nixos
<aszlig>
so if lutris handles the ugly parts, we can do the nice parts with nix
<pie_>
clever: basically im guessing the function that is configuration.nix has to get its pkgs argument from a hardcoded version of import <nixpkgs>
<clever>
pie_: changes to nix.nixPath dont take effect until afetr the build has finished
<pie_>
clever: and i dont have a nixpkgs because i wanted to name it nixos
<infinisil>
zupo: Try disabling substituters with `--option substitute false`
<pie_>
clever: but yes it is
<__monty__>
infinisil: You might like ranger. We're sporting python compatibility back to 2.6 and we try to never introduce backwards incompatible configuration changes. #shamelessplug ; )
<infinisil>
__monty__: I mean, I don't mind getting rid of old stuff, but there at least could be something like an error "This is not supported anymore" instead of just removing it without a trace
<clever>
pie_: nixos-rebuild will first try to build -A nix, to get the new nix from your (new) nixpkgs
<LnL>
jonreeve: it works outside of python.withPackages, which is pretty weird
<clever>
pie_: if that fails for any reson, it thinks your old nix is too old
jonreeve has quit [Ping timeout: 246 seconds]
<clever>
pie_: so it grabs a hard-coded nix build directly from the binary cache, bypassing all nix expressions
upsaday has joined #nixos
upsaday has quit [Client Quit]
<vaibhavsagar>
slabity: can you use openjdk?
<__monty__>
infinisil: That doesn't result in such an error?
<aszlig>
laas: while this works great for some single "nix-build <somegame>" because it somewhat abuses fixed-output derivations, it will get nasty with lots of parallel requests, because you need to solve a captcha then for every single download (all that within a timeout they have, which i don't remember)
Czen has quit [Ping timeout: 272 seconds]
<infinisil>
__monty__: Am confused, I thought we're talking about the jdk9 vanishing
<aszlig>
laas: so the idea was, that lutris would handle all the api login, captcha handling and whatnot and just supply the api-token to nix and you get a collection of games readily built without fetching random binaries from lutris.net (which it does right now)
_kwstas has joined #nixos
<__monty__>
Oh, I thought we were talking APIs, nvm, carry on.
<slabity>
vaibhavsagar: Openjdk8 has a bug in it for the particular app I'm using, and openjdk11 doesn't even run. Crashes with a 'NoClassDefFoundError'
<slabity>
Because I guess it got rid of swing.plaf.windows.WindowsLookAndFeel? Which is for some reason required in the app I'm using
<aszlig>
laas: so by default, the nix store in lutris would live in a chroot within the homedir, but optionally you could just use the system store, which should also make it very easy to package lutris on nix(os)
<{^_^}>
[nixpkgs] @benley pushed commit from @r-ryantm to master « latte-dock: 0.8.4 -> 0.8.5 (#55254) »: https://git.io/fh98u
<spacekitteh[m]>
how do i set a derivation argument when invoking nix-shell? is it e.g. --arg myparam 'true'
moosefish_ has quit [Quit: Page closed]
civodul has joined #nixos
<pie_>
clever: oh huh. thats kind of weird but makes sense I guess, you looked up the source_
<pie_>
?
<clever>
spacekitteh[m]: --arg myparam will only work if shell.nix starts with { myparam }: or { myparam ? def }:
<clever>
pie_: i memorized the source....
<pie_>
clever: ok so you just had to think for a second xD
<clever>
pie_: i had to wait for chrome to load, so i could link the exact line :P
<spacekitteh[m]>
clever: it does. if that's the case then it must be stack passing the parameter non-verbatim then
<pie_>
clever: oh lol i didnt see the link
<pie_>
clever: aah found i
Czen has joined #nixos
<pie_>
clever: thanks... /me still scratching his head :P
<clever>
spacekitteh[m]: `strace -f -e execve ...` will show the exact commands being ran
<spacekitteh[m]>
oh cheers
agander has quit [Quit: Leaving]
<{^_^}>
[nixpkgs] @scalavision opened pull request #55286 → singularity: rewrite path to cp in bootstrap script → https://git.io/fh98V
tzemanovic has joined #nixos
Glider_IRC has joined #nixos
tzemanov_ has quit [Ping timeout: 240 seconds]
<zupo>
@infinisil: nope, didn't help
<infinisil>
zupo: error?
<zupo>
it tries to download something from tarballs.nixos.org
<zupo>
then says dependencies couldn't be built
<zupo>
I have the configuration in git and I confirmed I did not change anything else besides "address" , "defaultGateway" and "nameservers" values in https://github.com/zupo/nix/blob/master/tv.nix#L44
<infinisil>
zupo: Try also giving it `--option fallback true`
<zupo>
sec
<infinisil>
(Btw, you can look what these options do in `man nix.conf`)
o1lo01ol1o has quit [Remote host closed the connection]
<symphorien>
dev ${insert the device name of your network card}
<zupo>
got it
<symphorien>
the list of devince is in ip link
<spacekitteh[m]>
anyone know how to do this correctly in Stack? i get an this error: error: undefined variable 'arg' at (string):1:3. the command i actually run is: stack build --nix-shell-options --arg testDeps 'false'
phizzz has joined #nixos
<{^_^}>
[nix] @matthewbauer opened pull request #2667 → Add armv6l-linux & armv7l-linux as cross jobs → https://git.io/fh98d
<aminechikhaoui>
Avaq gchristensen not sure actually why rollback isn't enabled by default, maybe because that will keep a gc root for each deployment which takes up space so it was more of an opt-in option
<{^_^}>
[nixpkgs] @dywedir pushed commit from @markuskowa to master « molden: 5.8.2 -> 5.9.3 (#55044) »: https://git.io/fh94f
<zupo>
@infinisil, @symphorien: solved it with "ifconfig eth0 192.168.8.10 netmask 255.255.255.0 up" and "route add default gw 192.168.8.1". Thanks for the help!
<phizzz>
i'm struggling with getting a basic nix-shell going with a test package, w3m for example. is there a way to do something like this in default.nix: import <nixpkgs> { w3m }; and then do `nix-shell` and have the w3m package installed there?
<phizzz>
infinisil: ok i will try that, but i want to understand the default.nix syntax
sigmundv_ has joined #nixos
<symphorien>
the syntax for shell.nix is: it must evaluate to a derivation whose buildInputs will be available in the PATH of the shell
<symphorien>
a handy way to get such a derivation is to use mkShell as above
<infinisil>
symphorien: That's more the semantics, not syntax though
<symphorien>
right
<symphorien>
well for the syntax, see my example above :þ
<Avaq>
Trying out NixOps with a targetHost (my rbpi already running nixos) for the first time (tried vbox earlier). When I 'nixops deploy', I get the following error message (after "closures copied successfully"):
<Avaq>
File "/nix/store/s2hqrcnqlny1vp8r88kkjml9asmaqyfp-nixops-1.6/lib/python2.7/site-packages/nixops/deployment.py", line 731, in worker
<Avaq>
pi> /nix/var/nix/profiles/system/bin/switch-to-configuration: line 8: `use Sys::Syslog qw(:standard :macros);'
<Avaq>
pi> /nix/var/nix/profiles/system/bin/switch-to-configuration: line 8: syntax error near unexpected token `('
<Avaq>
pi> /nix/var/nix/profiles/system/bin/switch-to-configuration: line 7: use: command not found
<Avaq>
raise Exception("unable to activate new configuration")
<Avaq>
I had a feeling it might be related to using unstable channel on the targetHost (pi), so I changed it to a stable channel (nixos-18.09) and rebuilt. However, 'nixos-version' still reports "19.03pre167858.f2a1a4e93be (Koi)". Does anyone have an idea what might be going wrong?
<lucus16>
clever: How do pkgsCross and e.g. https://github.com/riscv/riscv-gnu-toolchain relate? It doesn't seem like nixpkgs knows about that repo, but it does seem like something it would need
<jonreeve>
But it's not working, complains of collisions
<jonreeve>
I just pasted the collision error in that gist
<matthewbauer[m]>
lucus16: they both have the same upstream - gcc & binutils
<matthewbauer[m]>
otherwise they use their own build processes that might not be very close
<Avaq>
What is the output of 'nixos-version' based on?
<symphorien>
the revision of nixpkgs used to build the current generation
erobl[m] has joined #nixos
<Avaq>
I expect it to give me 18.09something when I set my nixos channel to nixos-18.09; but it's still giving me the previous version I had; 19.03pre..., from when I was using unstable channel.
<srhb>
Avaq: It is in fact hardcoded into the script.
<srhb>
Avaq: Then you probably still have nixos-version from that old channel.
<symphorien>
Avaq: you must first nixos-rebuild for it to have effect
<Avaq>
I did nixos rebuild
<Avaq>
Many times, even rebooted
<symphorien>
can you run nix-info ?
<symphorien>
you might have changed your user's channel, and not root's
<typetetris>
Hi there! Is clangd already in nixpkgs?
<typetetris>
When is the next "upgrade" of the emacs melpa packages in nixpkgs?
knupfer has quit [Remote host closed the connection]
<srhb>
typetetris: I'm not aware of any schedule.
<srhb>
typetetris: If you think we should bump it, you can open an issue and ping the usual suspects? Or maybe just PR the bump :)
<srhb>
instructions are in the -packages exprs.
knupfer has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
<typetetris>
I look into it. (lsp-clangd is outdated now and included in lsp-mode, but that "step" isn't in nixpkgs yet.)
<srhb>
Ah. :)
o1lo01ol1o has joined #nixos
jonreeve has quit [Ping timeout: 250 seconds]
o1lo01ol1o has quit [Ping timeout: 268 seconds]
knupfer has quit [Remote host closed the connection]
knupfer has joined #nixos
mkoenig has quit [Ping timeout: 268 seconds]
mkoenig has joined #nixos
dermetfan has quit [Ping timeout: 245 seconds]
Ariakenom has quit [Quit: Leaving]
civodul has quit [Quit: ERC (IRC client for Emacs 26.1)]
simukis has joined #nixos
eadwu has quit [Quit: WeeChat 2.3]
maximiliantagher has quit [Remote host closed the connection]
bpa has quit [Remote host closed the connection]
maximiliantagher has joined #nixos
itorres has quit [Remote host closed the connection]
tzemanovic has quit [Remote host closed the connection]
tzemanovic has joined #nixos
maximiliantagher has quit [Ping timeout: 268 seconds]
tzemanov_ has joined #nixos
tzemanov_ has quit [Read error: Connection reset by peer]
iqubic has joined #nixos
tzemanov_ has joined #nixos
tzemanov_ has quit [Remote host closed the connection]
tzemanov_ has joined #nixos
tzemanov_ has quit [Read error: Connection reset by peer]
tzemanov_ has joined #nixos
tzemanov_ has quit [Remote host closed the connection]
tzemanov_ has joined #nixos
tzemanov_ has quit [Read error: Connection reset by peer]
tzemanovic has quit [Ping timeout: 252 seconds]
<typetetris>
emacs2nix doesn't build for me at the moment ...
tzemanov_ has joined #nixos
tzemanov_ has quit [Remote host closed the connection]
<iqubic>
what is emacs2nix?
<typetetris>
the tool to update melpaPackages and stuff
gagbo has joined #nixos
devalot has left #nixos ["ERC (IRC client for Emacs 26.1)"]
<typetetris>
nice he totally missed pointing out fetching the submodules ...
<pie__>
laas, still here?
tzemanovic has joined #nixos
tzemanovic has quit [Remote host closed the connection]
c0ffee152 has joined #nixos
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]
ottidmes has quit [Ping timeout: 250 seconds]
tzemanovic has joined #nixos
tzemanov_ has joined #nixos
tzemanov_ has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @flokli pushed commit from @bachp to release-18.09 « gitlab-runner: 11.6.0 -> 11.7.0 »: https://git.io/fh9B0
Dagger has quit [Excess Flood]
troydm has joined #nixos
<delan>
g’day 👋 anyone use virtualisation.libvirtd.qemuOvmf? I can’t seem to get UEFI guests to boot from CD on 18.09 nor unstable, they always drop into the UEFI shell