<jlv>
I'm trying to get NixOS up and running on a GPD Pocket 2, and I'm running into some weird issues with closing the lid. I want to turn off the display but not suspend, when the lid is closed. I tried setting `services.logind.lidSwitch = "ignore"`, but now when I close the lid, the display turns black but the backlight increases to max, and when I
<jlv>
open the lid, the display is stuck like that. Anyone ever run into an issue like this or have any suggestions?
maddo has quit [Quit: See ya]
kleisli has quit [Remote host closed the connection]
<{^_^}>
#37010 (by mguentner, 2 years ago, open): buildRustPackage does not work with private/ssh dependencies
<avn>
risson: in theory -- yes, although you probably need to supply NIX_PATH manually. And not sure how it practically useful
<risson>
I want to deploy a django application whose dependencies are specified with pipenv
<risson>
Now, I do have a mkShell expression to get in a shell with all those dependencies, and I don't want to package all of them, because it would take me all night
<risson>
And I'd like to finish this so I can sleep
oida has joined #nixos
<avn>
better postpone it to morning ;)
<adisbladis>
"Friends dont let friends use Pipenv" - adisbladis
<risson>
Yes I know, but I want it to be deployed by morning
<jlv>
Are there any options other than `services.logind.lidSwitch`, and its variants, that control what happens when the lid of a laptop closes? Setting logind to `"ignore"` does not actually result in nothing happening when the lid closes. Something is still trying to (badly) disable the display.
<b42>
jlv: there's services.upower.ignoreLid but not sure whether that's releavnt
<jlv>
b42: I don't have `services.upower` enabled, so I assume that won't affect anything.
pbogdan has quit [Ping timeout: 246 seconds]
globin has joined #nixos
justanotheruser has quit [Ping timeout: 272 seconds]
<rajivr___>
I tried to do `passExtensions.pass-import = super.passExtensions.pass-import.overrideAttrs (import ./pass-import);` in an overlay, but that doesn't seem to work.
<evils>
anyone know how to get hydra to build the dependencies of a package without building the package itself?
<glittershark>
not hydra, but iirc you can build the build environment of a `mkDerivation` with `-A env`
<cole-h>
If you know the dependencies by attr, you can `@GrahamcOfBorg build <attr1> <attr2> <attr3>`, assuming this is on a PR
<evils>
it's for the kicad package, i want kicad.base and kicad.libraries (minus packages3d) to be built, and if a user wants kicad, they get all the components hydra can build, and "build" the packages3d part locally
<evils>
well, for kicad specifically, there's kicad-small that has everything minus packages3d; i'm trying to avoid adding a kicad-unstable-small to get kicad-unstable to work
captjakk has quit [Remote host closed the connection]
<pjt_014>
Okay, I'm in the relevant path in my copy of the git repo. Am I just adding the relevant line somewhere?
<pjt_014>
Also how do I test this? nix-shell somehow?
<thefloweringash>
gchristensen: I think channels.nix.gsc.io is slightly broken, the /latest-url for nixos-unstable is pointing to https://channels.nixos.org/nixos-unstable. It used to point to a particular revision of nixpkgs
<{^_^}>
[nixpkgs] @delroth opened pull request #83745 → jadx: init at 1.1.0 → https://git.io/Jv7WQ
reallymemorable has joined #nixos
Supersonic has quit [Disconnected by services]
Supersonic112 has joined #nixos
Supersonic112 is now known as Supersonic
<cole-h>
You can modify the derivation you want to build (dvtm) and then use `nix-build -A pkgsCross.aarch64-multiplatform.dvtm` and see if that works
sjkelly1 has quit [Remote host closed the connection]
<pjt_014>
cole-h: I added "patchShebangs --build default.nix" to the postPatch section and ran that build cmd
<pjt_014>
got "error: cannot auto-call a function that has an argument without a default value ('callPackage')"
<pjt_014>
same with "--build configure"
<{^_^}>
[nixpkgs] @veprbl opened pull request #83747 → pythonPackages.pyarrow: remove dead code in preCheck, use pytestCheckHook → https://git.io/Jv7lL
<pjt_014>
same error when I call it without the patch line
<cole-h>
You should be running patchShebangs on the scripts from dvtm, not the Nix files :P
<hyper_ch>
pjt_014: you clone the nixos-repository, create a new branch, add your patches/fixes/changes to it, commit, push it to your github repo and then make a pull request
<pjt_014>
That much I know. It's just the way cole-h described the fix made it seem like I needed to patch individual scripts in the store, which wouldn't make sense.
<pjt_014>
I'm browsing around for an example of this kind of fix right now cause I still can't tell how it's supposed to work.
<hyper_ch>
you have a seperate patch file?
ddellacosta has quit [Ping timeout: 252 seconds]
<pjt_014>
No, I'm still figuring out how to actually fix this specific thing. Haven't done it before.
<srhb>
pjt_014: If you intend to PR, it's indeed easier to work with a local checkout, in which case from the root of it you can use nix-build . -A yourpackage
<srhb>
pjt_014: (That is, after modifying the appropriate file)
fragamus has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<pjt_014>
srhb: thanks, that was the cmd I was looking for. I assume it uses the default.nix of the dir I'm in?
<srhb>
pjt_014: Yes. In fact, even . can be ommitted, and it will still do so.
wavirc22 has quit [Read error: Connection reset by peer]
<clever>
mica[m]: line 173, it just does `docker run --rm ...` against the name, which i think it pulls from dockerhub
<clever>
mica[m]: but once in docker, it wont pull again
<srhb>
mica[m]: It's just regular docker semantics, as far as I can see. If you change the name (say, to a newer tag) that one will be grabbed.
<srhb>
Sorry, the *image* in the option terminology.
<srhb>
It's up to docker to resolve that image into some version (and I think not specifying a tag implies "latest" -- but I don't recall the semantics. All I do remember is that not specifying a specific, versioned tag is scary, because you'll be running "some version god knows which")
<mica[m]>
thanks to both of you
<mica[m]>
looks like every time the unit starts, it fetches the image? since running docker with `-rm` will remove the container image after the service stops
<mica[m]>
does that sound correct?
<srhb>
mica[m]: I don't think removes the _image_, only the container
<clever>
mica[m]: the --rm says to delete the container, but not the image
<clever>
mica[m]: so all state will be lost at stop, and it will start fresh from the image
<srhb>
But I don't recall whether docker will try and re-resolve the "latest" tag each time you try to run a container from that image.
<srhb>
You'll have to test that.
raingloom has joined #nixos
<pjt_014>
srhb: I'm in my local checkout in nixpkgs/pkgs/tools/misc/dvtm, and running nix-build . -A dvtm gives an error: "cannot auto-call a function without a default value ('callPackage')"
<srhb>
pjt_014: Stand at the root of the nixpkgs repo
<pjt_014>
ah
<srhb>
pjt_014: The individual package expressions are meant to be called via callPackage (which provides all the dependency arguments) -- the top level nixpkgs expression takes care of that.
zaeph1 is now known as zaeph
<srhb>
pjt_014: To call an individual package with your system nixpkgs, you'd stand in the package expression dir and do something like `nix-build -E 'with import <nixpkgs> {}; callPackage ./. {}'`
<srhb>
But I usually do the former, since that ensures you're using the checkout version of nixpkgs to build the package.
<mica[m]>
it seems there is a strong trend to not use the "latest" tag in docker if you're shipping an image for production. So one would need to update the nix config file every time a new docker image comes out.
<srhb>
mica[m]: Yeah.
raingloom has quit [Read error: Connection reset by peer]
<srhb>
mica[m]: Silently self-updating packages is generally quite scary.
raingloom has joined #nixos
<srhb>
mica[m]: It's usually better to automate the update process (say, finding the new tag and automatically writing it to some file readable by nix)
<mica[m]>
srhb: I've been experimenting with some non-critical software about how lazy I can actually be.
<pjt_014>
Well I can replicate the error with the aarch64 version. Just need to figure out how to fix it now
<mica[m]>
apparently the problem with the latest tag is how docker doesn't try and pull when starting.
<srhb>
mica[m]: This is of course just my Opinion^{tm} but my rule of thumb is that any update of any package needs action from me. That action may be as simple as ./update-all-the-things.sh followed by a git commit and a rebuild. I may get a lot of new versions, but it corresponds to a commit that I can revert whenever I want to.
<mica[m]>
sorry, trying to learn about docker inside nix.
<srhb>
mica[m]: There's probably some force pull mechanism you can feed to extra options for the run command, but I strongly advise against this approach. :)
<mica[m]>
srhb: yes I understand
<mica[m]>
I've been using the snap of nextcloud on ubuntu, and it has worked pretty well. I was wondering how I could do it similarly in Nix :)
<srhb>
Apparently it's called --pull=always
<srhb>
But I definitely prefer the two-phased approach that update the tag, then deploys a reproducible tag.
<mica[m]>
srhb: I understand, thank you for taking the time to help me.
<srhb>
Sure thing :)
lordcirth_ has quit [Remote host closed the connection]
lordcirth_ has joined #nixos
<pjt_014>
srhb: After looking thru some examples, I tried "patchShebangs ." in dvtm.nix. The error is unchanged.
opthomasprime has joined #nixos
<pjt_014>
I don't need to do something else to make nix see it, do I?
opthomasprime has left #nixos [#nixos]
<srhb>
pjt_014: If you're unsure whether your changes are taking effect, you can put in a big fat exit 1 or echo "pjt_014 debug message" somewhere in the build and confirm for yourself :)
<pjt_014>
good ol' "echo AAAAAAAAAAAAAAAAAAAAAAA" it is then
<srhb>
pjt_014: OK, where did you make the changes and what command did you run?
<pjt_014>
In checkout_root/pkgs/tools/misc/dvtm/dvtm.nix, I added a line to the postPatch block that reads 'patchShebangs .' Then in checkout_root I ran nix-build -A pkgsCross.aarch64-multiplatform.dvtm
<srhb>
pjt_014: Notice that the postPatch phase is guarded by `optionalString`
<pjt_014>
conditional on customConfig, yeah
<srhb>
Did you set that somewhere?
<pjt_014>
set what? it's only an import, right?
<srhb>
pjt_014: What I'm trying to say is: If you put it in the string that's guarded by customConfig, and didn't append it unconditionally, you'll need to set customConfig in order for it to take effect.
<srhb>
pjt_014: I *think* what you want is to either unconditionally append something to postPatch, or conditionally append based on architecture. For now, you could just do the unconditional append to start debugging.
<pjt_014>
ah. I take it it's a ++ or something to that effect?
<srhb>
pjt_014: In which case patchShebangs can't help you.
<pjt_014>
stdout just says 'installing terminfo description'. I'd have to poke around unless --verbose can tell me.
<srhb>
pjt_014: Or grep might help. But try simply calling tic --help in prePatch or something
vandenoever has joined #nixos
<srhb>
pjt_014: Maybe tic isn't included in the aarch version of ncurses
<pjt_014>
oooh a subtle bug
<pjt_014>
spicy
<bqv[m]>
Or just, look in ncurses…
<srhb>
pjt_014: I'm just spitballing, hopefully gives you something to go on.
<pjt_014>
It def does.
<srhb>
(The error message implies to me that it's not a shebang, that it's trying to call tic from $PATH, and it's not there)
m0rphism has joined #nixos
<pjt_014>
On a quasi-whim, I'm gonna put echo $PATH in the postPatch block
<pjt_014>
oh yeah
<pjt_014>
nix paths are hUge
<pjt_014>
well then
<bqv[m]>
Pipe it into tr ':' '\n'
<pjt_014>
its R E G E X T I M E
<pjt_014>
or that
<bqv[m]>
Its unusable otherwise
cr4y1 has joined #nixos
charukiewicz has joined #nixos
raingloom has quit [Ping timeout: 256 seconds]
FRidh has quit [Ping timeout: 256 seconds]
<{^_^}>
[nixpkgs] @jtojnar opened pull request #83755 → libjcat: init at 0.1.0 → https://git.io/Jv7BB
<charukiewicz>
I ran into an issue on an EC2 instance where my disk became full while running nix-copy-closure and after reboot I can't SSH into the machine. I tried mounting the volume onto another NixOS machine, and I can do so successfully, but I'm not sure how I can repair the system. Is there some command I can run on the temporarily mounted nix-store to repair it?
raingloom has joined #nixos
<pjt_014>
Do you have nixos-enter available?
<charukiewicz>
Yes
<pjt_014>
If a lack of disk space is the problem, a single nix-collect-garbage would free it up
<charukiewicz>
Well this is the error that nix-copy-closure failed at:
<charukiewicz>
error: creating symlink from '/nix/store/24r4sj37nd72bwgsfjwfxp76j8aq4ggh-system-path/sbin/grub-mkimage' to '/nix/store/z5h7hcvbzvy9sjazplmkpq59q8iyhh7k-grub-2.04/sbin/grub-mkimage': No space left on device
<charukiewicz>
So I assume there may be a grub issue
<pjt_014>
Well that might need fixing too, but freeing up that space would give you some spare room to work with
<{^_^}>
[nixpkgs] @nomeata opened pull request #83756 → ghc-8.4.4.nix: Do not use git.haskell.org → https://git.io/Jv7BD
<srhb>
You need more flags to nix-collect-garbage to start deleting older kernels.
<pjt_014>
nix optimise-store couldn't hurt either, once you have your system ssh-able and want to try the copy again
<srhb>
The problem in this case looks like the boot partition
lsix has joined #nixos
Acou_Bass has quit [Ping timeout: 256 seconds]
turion has joined #nixos
<srhb>
So just collecting garbage in the store won't make a difference.
<pjt_014>
wouldn't it with -d?
<srhb>
pjt_014: Yes.
<pjt_014>
the ol' nuclear option.
<srhb>
That will delete older system generations and free up space for rebuild boot.
<srhb>
(Well, --delete-older-than 30d or something might be the less nuclear option)
FRidh has joined #nixos
<pjt_014>
charukiewicz: you might also try nix-env --delete-generations +2 or something to that effect
<charukiewicz>
okay
<pjt_014>
and then the garb. collect
<charukiewicz>
How can I list generations?
<pjt_014>
nix-env --list-generation
<pjt_014>
--list-generations
<pjt_014>
escuse the typo
<charukiewicz>
no problem
nschoe has joined #nixos
Acou_Bass has joined #nixos
<charukiewicz>
So it looks like the method I've been using for deployments hasn't been creating new generations, so there were only two on the system, back from Nov 2019. I've deployed several times since then.
<srhb>
charukiewicz: You may need to point it at --profile /nix/var/nix/profiles/system manually
<charukiewicz>
Ah yes, that did the trick
<charukiewicz>
149 generations
<bqv[m]>
,locate log4e
<{^_^}>
Couldn't find in any packages
pnobre has joined #nixos
<charukiewicz>
okay so I tried to remount the volume as the root volume on the original instance and it looks like it's still not booting
<charukiewicz>
can I do something like nixos-enter followed by nixos-rebuild to force it to rebuild grub, etc?
<pjt_014>
You should be able to, yes.
<pjt_014>
or nixos-enter /mnt/mountpoint to give you an interactive shell to try it in
cizra has quit [Ping timeout: 260 seconds]
coco has quit [Ping timeout: 256 seconds]
chagra has joined #nixos
cizra has joined #nixos
kleisli has quit [Remote host closed the connection]
kleisli has joined #nixos
MichaelRaskin has quit [Ping timeout: 264 seconds]
lordcirth__ has quit [Remote host closed the connection]
lordcirth__ has joined #nixos
<charukiewicz>
Would the nixos-enter chroot have limited network connectivity in some way? I try to nixos-rebuild and I get
<{^_^}>
[nixpkgs] @cdepillabout pushed 2 commits to haskell-updates: https://git.io/Jv70u
<pjt_014>
what are you using with it? switch?
<charukiewicz>
Yes
<pjt_014>
try boot, I doubt it'd try to fetch anything
<charukiewicz>
Same warning related to not being able to resolve host name
<pjt_014>
hm. try --keep-going, it might still be able to build the boot files anyway
<charukiewicz>
Looks like the issue is related to the way that the chroot starts. Running top/htop shows that nothing is actually running. The process list is empty other than my bash shell
<xfix>
on https://hydra.nixos.org/machines, machine e05744a2 seems to be stuck, all its jobs are executing for at least a day and 8 hours
<pjt_014>
charukiewicz: Is that with the interactive version of the shell too?
<xfix>
hm, so there is a bug somewhere that causes jobsets to block
<charukiewicz>
pjt_014: I found a GH issue that said you could copy /etc/resolv.conf to /mnt/etc/ and that would fix the issue. I did that and it seemed to work.
<xfix>
it doesn't affect 19.09, apparently
<charukiewicz>
But after the nixos-rebuild boot, I unmounted and remounted to my original EC2 instance, still not booting correctly apparently. Can't SSH into it
<pjt_014>
The thing you mentioned previously about the rebuild not doing anything according to top--was that incorrect then?
orivej has quit [Ping timeout: 264 seconds]
<charukiewicz>
I just meant that no other processes were running in the chroot, I figured something was necessary in order for DNS to get resolved. Looks like it was just missing a /etc/resolv.conf
<charukiewicz>
I'm kind of bummed that this happened. It's really not clear to me what it is that caused the machine to enter this state. As I said before, it looks like the disk full error occurred during a grub-related symlink, but I figured the nixos-rebuild would fix that
<pjt_014>
Yeah, nixos doesn't always handle full disks well. I've had it where my disk was completely full and I learned that sqlite transactions take a few bytes of disk space to complete :/
<pjt_014>
Then again, many things don't as they should (or at all) on a full disk
<pjt_014>
like bash autocompletions
<benny>
how do I manipulate emacsWithPackages (or otherwise) to provide the emacs with an environment that contains libvterm?
cfricke has joined #nixos
pnobre has quit [Ping timeout: 252 seconds]
zeenk has joined #nixos
lsix has quit [Quit: WeeChat 2.7.1]
<{^_^}>
[nixpkgs] @vcunat pushed commit from @nomeata to release-19.09 « ghc-8.4.4.nix: Do not use git.haskell.org »: https://git.io/Jv7E6
<{^_^}>
[nixpkgs] @vcunat merged pull request #83756 → ghc-8.4.4.nix: Do not use git.haskell.org → https://git.io/Jv7BD
zupo has joined #nixos
lsix has joined #nixos
<{^_^}>
[nixpkgs] @vcunat pushed commit from @nomeata to release-20.03 « ghc-8.4.4.nix: Do not use git.haskell.org »: https://git.io/Jv7EP
<benny>
adisbladis: thanks, I'm using DOOM Emacs and it wants to build it again using straight.el and that fails
<benny>
make[2]: *** No rule to make target 'libvterm-prefix/src/libvterm/.libs/libvterm.a', needed by '../vterm-module.so'. Stop.
<adisbladis>
benny: That's a completely different issue then.
linarcx has joined #nixos
<adisbladis>
Is there some way you can tell straight.el _not_ to build a dependency?
<benny>
that's new after getting the vterm environment set up...
<benny>
what I've done before with sqlite is that I set up the environment and just manually build it and placed it in the place it expects the .so
<charukiewicz>
Well. After fiddling with this for over an hour, I gave up and deployed a new fresh NixOS machine from the community AMI, updated the IP address, and ran my deploy command. New server up and running in under 5 minutes.
<charukiewicz>
pjt_014: thanks for your help. I didn't know about nixos-enter, that's very helpful.
Makaveli7 has joined #nixos
<pjt_014>
no problem :D
<adisbladis>
Sounds very hacky :/
<pjt_014>
Do be sure to watch your disk space though
<benny>
yeah I'm not happy with it, but I would just have it rather work than not ;-)
chagra has quit [Ping timeout: 265 seconds]
<charukiewicz>
I guess one question: I was running df -h and it said that I had several gigs free
<adisbladis>
I don't have a good solution for you.. You're sadly on your own with doom/spacemacs
<pjt_014>
How big was that closure?
lordcirth__ has quit [Read error: Connection reset by peer]
<charukiewicz>
Is this "hidden" disk usage related to profiles in /nix/var/nix/profiles/system?
lordcirth__ has joined #nixos
<charukiewicz>
Uh, not sure
hio has joined #nixos
<pjt_014>
benny: open a github issue maybe?
<pjt_014>
charukiewicz: well what was it?
<charukiewicz>
It's a compiled Haskell web server. I assume what actually gets copied isn't that large.
<charukiewicz>
I had a 20GB ssd
<benny>
what I need is antithetical to nix I find. I need to prove the build environment needed for libvterm into the emacs package for it to work
<pjt_014>
If it included ghc then it would be quite large.
raingloom has quit [Ping timeout: 250 seconds]
<adisbladis>
benny: You're gonna have a bad time with that. The vterm build is not exactly straight forward.
<benny>
epkgs: emacsql-sqlite worked to build it while running. (magit-forge needs it)
<adisbladis>
benny: My problem is the lack of taste in package management.
<pjt_014>
It redirects a command's output to a file and gives you back your terminal. You can then queue up multiple things and they'll run in sequence.
<charukiewicz>
Ah I see
<pjt_014>
for debugging It's really nice because you dont have to rely on screen backscroll
<adisbladis>
They are making it next to impossible for distro packagers
<benny>
adisbladis: fair enough, but I think having pinned dependencies like straight.el does is already a bit better than just melpa which is time sensitive
<pjt_014>
If you were doing a one-off version of that command and did like nq timeout 1m do-thing, you could look at the log file and see how much stuff it's pulling in
Makaveli7 has quit [Quit: WeeChat 2.7.1]
<charukiewicz>
makes sense
<parsnip>
did someone say emacs?
<pjt_014>
I wouldn't be surprised if that was eating up all the space. GHC is an absolute unit to work with. You ever try to compile it for aarch64?
<charukiewicz>
I have not
<charukiewicz>
but on that note, I need to head off. thanks again for the help
<pjt_014>
one more ting
<pjt_014>
consider setting nix.autoOptimiseStore
cr4y1 has quit [Read error: Connection reset by peer]
<charukiewicz>
Ah, interesting
<pjt_014>
On a home machine that's a bit rough on a disc for my taste but on a server it's free space reclaiming with no downside really
<bqv[m]>
Like, literally porting it from elisp to nix
linarcx has quit [Quit: WeeChat 2.8]
<adisbladis>
parsnip: But it's not good enough to go into upstream nixpkgs
<benny>
parsnip: back then doom was quite a bit smaller (and sounding like a hipster) I was just coming off of an attempt to use org-babel to organize my .emacs.el
<bqv[m]>
Ill post a sample in #freenode_#nixos-emacs:matrix.org later
<adisbladis>
We should start our own "community-distribution" ;)
<adisbladis>
The tagline can be that stuff actually works out of the box
linarcx has joined #nixos
<benny>
I'm recently quite re-excited by emacs with native emacs lisp compilation for example
<bqv[m]>
Lol
<bqv[m]>
Nix-straight.el looks neat
<sphalerite>
Does anyone have a nice way to run gdb on wrapped programs?
<bqv[m]>
Unwrap them :p
<adisbladis>
I just don't see the appeal in straight
<adisbladis>
It's like nix--
<bqv[m]>
I did, before nix
<bqv[m]>
Now I've replaced it with nix
<bqv[m]>
But on gentoo it was way nicer than the emacs overlay
<benny>
to be fair, I think nix is a guix-- ;-)
<pjt_014>
sphalerite: I think this is what patchelf is for (?)
<pjt_014>
maybe
<pjt_014>
the dev side of nixos is not my strong suit
<bqv[m]>
benny: only if you're ok with being gnud without consent
<benny>
I wouldn't want to use it because I'm pragmatic but it's a lisp and therefor already better in my world and I thought I grasped the definitions while I still struggle with nix's
<parsnip>
"the best emacs package manager is neither package.el, nor straight.el. it's nix"
<parsnip>
the best editor is neither emacs nor vim, it's emacs+evil.el
<bqv[m]>
benny: tbh I'm coming around to nix as a language
<parsnip>
the best distro is neither spacemacs nor doom. it's ubuntu.
<bqv[m]>
parsnip++
<{^_^}>
parsnip's karma got increased to 4
<adisbladis>
Lol
<benny>
man what a revelation evil.el was
<bqv[m]>
parsnip--
<parsnip>
lol
<adisbladis>
Let's start a new emacs distribution called ubuntu-emacs (of course based on nix)
<bqv[m]>
emacs, for humans
<pjt_014>
sphalerite: you may also want to poke around nixos.wiki, the subreddit (maybe post the question there), and/org ask here again at a later time. Someone who knows may not be here right now because timezones.
<adisbladis>
pjt_014: sphalerite is not exactly new to this community ;)
<pjt_014>
Sorry, I know like 5 1/2 people here by name
<bqv[m]>
sphalerite: for what its worth I've found using gdb on store binaries is an exercise in pain any way you slice it
<adisbladis>
pjt_014: It's ok :) You're doing great! It was not at all a criticism
<adisbladis>
bqv[m]: Omg I _hate_ that tagline so much
<avn>
bqv[m]: nice, does anyone befriuend Nix-straight.el with doom-emacs?
<sphalerite>
zupo: not in a nice typed way, but you can use networking.firewall.extraCommands
<zupo>
ah, nice!
<zupo>
thanks!
<bqv[m]>
rycee: hey, have you had any thoughts to how you'd like to support flakes in home-manager?
<bqv[m]>
cc timokau
globin has quit [Quit: o/]
<bqv[m]>
i've been maintaining my fork at bqv/home-manager/flakes, and it works ok, but it's not compatible
turion has quit [Ping timeout: 272 seconds]
globin has joined #nixos
<rycee>
bqv: No, I haven't looked at flakes at all yet beside what I saw in Brno. I might have some time in June-July but my schedule is starting to fill up again so it might not be until the fall :-/
<rycee>
But I have a long list of other things to work on so flakes is pretty far down the list.
kleisli has quit [Remote host closed the connection]
kleisli has joined #nixos
<bqv[m]>
essentially all that needs to change is that <nixpkgs> can't be explicitly used anywhere, and there needs to be a flake.nix, it's not too significant but might require a structural change in iirc 3 or so files?
<bqv[m]>
understandable otherwise, though.
<bqv[m]>
rycee: would you like to pull my fork into a branch so others can use it?
<bqv[m]>
at least leave the maintenance of that branch up to the community
<DigitalKiwi>
can you delete /old-root/ after using nixos-infect on a digital ocean host or does it need it for something
<adisbladis>
The RFC is not even accepted yet. I'd be wary of putting flakes into too many things until then.
<rycee>
Would it be possible to put the flake part in NUR?
<benny>
just a bit out left-field: I'm unable to build nixpkgs.graalvm8 and I think (it takes hours to build and then my system crashes) it's because the build process runs out of memory. I'm basically asking how you guys would investigate this
<benny>
slow iteration process and system crash is an annoying combination to just try to play around ;-)
<bqv[m]>
not really. not without putting the whole flake in
<bqv[m]>
adisbladis: it's fairly nonintrusive, in fairness. and i've seen little to no bugs in nixFlakes, besides the one where GC_free causes a segfault when i try and build emacs
<bqv[m]>
i don't forsee flakes being abandoned (especially since they're versioned)
chloekek has joined #nixos
<rycee>
I can put it in a separate branch and people can make flake related PRs against that one. bqv where is your branch? I can have a look at it in the evening after work.
<turion>
Or even create a new function that uses runCommand or something to create a Python package, and then pass its result as src for buildPythonPackage?
linarcx has joined #nixos
linarcx has quit [Client Quit]
laduke has quit [Ping timeout: 240 seconds]
linarcx has joined #nixos
<turion>
I wonder what's the better approach... probably the latter
Darkmatter66 has quit [Ping timeout: 246 seconds]
<typetetris>
Hi there! I can't find a way to just provide a nixos machine with a handwritten resolv.conf. Could somebody hand me a pointer to docs, how it is done? Or just write the nixos option name?
<chloekek>
To whoever made the packages webpage load in under five seconds, great job! :)
<typetetris>
nameservers .. okay
<eyJhb>
Does NixOS actually support Microsoft Teams well?
Darkmatter66 has joined #nixos
<Taneb>
typetetris: there's networking.resolveconf.extraCOnfig which gets appended to the end of resolv.conf
<typetetris>
Taneb, thanks. Can I not use resolveconf somehow?
<Taneb>
Hmm, I don't really know
mexisme has quit [Ping timeout: 240 seconds]
<ottidmes>
bqv[m]: does it still work without flakes with those changes, should it not default to <nixpkgs> in some places? Also some seem just styling changes / whitespace, probably better to keep them out
<bqv[m]>
ottidmes: flakes are a total replacement for channels, so <nixpkgs> doesn't resolve, afaik
<ottidmes>
bqv[m]: I know, but the codebase should work for both flakes / non-flakes until flakes become the standard everything uses
<bqv[m]>
it doesn't :p
<bqv[m]>
oh, sorry, i misread your message
<bqv[m]>
i'm not sure it works without flakes with that patch, but feel free to test
Mateon3 has joined #nixos
fusion809 has quit [Remote host closed the connection]
<ottidmes>
bqv[m]: but is there any reason to have it in a seperate branch if it can be made to work for both
<bqv[m]>
specifically the part that's changed is the nixos module part, the package part should be absolutely fine
<ottidmes>
bqv[m]: not going to do that, not a HM user myself
<bqv[m]>
the point is that i'm not sure it can be easily made to work for both without structural change, hence why i pinged the creatro
<ottidmes>
turion: not that I am aware of, you have to do something yourself in runCommand. I needed something else recently, to basically overlay multiple paths together with symlinks, so say I want to only overwrite nixos/default.nix, but keep the rest of nixpkgs, it would create symlinks to build such a structure. If that is what you need, I can share it with you
zupo has joined #nixos
<turion>
sphalerite: I think not...
<turion>
ottidmes: Oh yes, that would be nice! I'll try and see whether I can make PR to nixpkgs with that eventually if that's fine with you
mallox has quit [Remote host closed the connection]
zupo has quit [Ping timeout: 256 seconds]
<bqv[m]>
i had a system to do that to add PRs that weren't merged yet, before i used flakes
<bqv[m]>
because some PRs were waiting for months and it was silly
<turion>
Ok, even stupider question: How do I make a local file into a derivation. Like writeFile, but I don't want to specify the text verbatim.
<bqv[m]>
${file}
<bqv[m]>
you shouldn't specifically need to though.
<turion>
thanks!
<bqv[m]>
what are you after, the file's contents?
<bqv[m]>
builtins.readFile
<turion>
Ah ok
globin_ has joined #nixos
<turion>
My long plan is to write a function that takes a .py file, Python-packages it in a trivial way, and then calls buildPythonPackage on it
sigmundv_ has quit [Remote host closed the connection]
oida has joined #nixos
kenran has quit [Quit: leaving]
ambroisie has joined #nixos
<turion>
I'm still stuck at how to make a local source file into a derivation... ${path} does not what I want
sigmundv_ has joined #nixos
<bqv[m]>
you want to make a directory into a derivaiton, not a file, don't you
<turion>
I basically have a local source file ./path/to/file.py, and I want to have a derivation containing /different/path/to/other_filename.py
eeva has quit [Remote host closed the connection]
eeva has joined #nixos
eeva has joined #nixos
eeva has quit [Changing host]
<turion>
bqv[m]: No, just a file
<turion>
Unfortunately it's in a directory with other stuff I _don't_ want to put in the derivation
<bqv[m]>
use stdenv.mkDerivation, set { src = ./path/to; }, then set { installPhase = '' mkdir -p $out/different/path/to; cp file.py $out/different/path/to/; ''; }
<turion>
(And I'm not allowed to move that stuff, and I can't predict what else is there)
<bqv[m]>
note that that's kinda impure and you should just find a way to have a hashable src
<turion>
How is it impure, or in other words, why can't this be hashed?
<bqv[m]>
because your derivation there uses the local filesystem as an input
<typetetris>
hmm I am trying to do an nixos install over debian, using the install from chroot method (it is a vServer and I only have remote access). On `nixos-install` I get an `writing to file: Operation not permitted` strace revealed, that some child process wants to write `/proc/<pid>/uid_map` and that fails. What can I try to get that working?
<ocharles>
How does concurrency work with remote builders? If I try and nix-store /nix/store/...drv on my local machine, presumably the entire closure gets uploaded to the remote builder, and is it then just the same as doing nix-store -r on that machine?
<bqv[m]>
it actually uses the current working directory as an input!
phreedom has quit [Remote host closed the connection]
phreedom has joined #nixos
<turion>
So by hashable src you mean a git commit of some repo?
<bqv[m]>
doesn't have to be git, just has to be fetchable from anywhere and easily reproducible
<turion>
If I change something in ./path/to/file.py, the hash of the derivation will change, and nix knows to rebuild this, right?
lsix1 has joined #nixos
<bqv[m]>
alternatively, bundle that file directly with your derivation, like most people do with package.jsons in nixpkgs
<bqv[m]>
turion: currently, no
pbogdan has joined #nixos
globin_ has quit [Ping timeout: 272 seconds]
<turion>
bqv[m]: Ah, but if I specify a whole directory, hashing works? In that case, I can create one derivation of that directory, and a second one where I've just moved out one file
<turion>
It's a bit tedious because it would trigger some unnecessary extra rebuilds
chagra has joined #nixos
globin_ has joined #nixos
<turion>
What do you mean by "directly with your derivation"? As in, "put the file in the same folder like the derivation"?
<bqv[m]>
yeah, because a derivation shouldn't generally be sourcing a random file somewhere on your filesystem (it's actually disabled in restricted mode nix)
chagra_ has joined #nixos
<turion>
Ok, so the directory .. is treated differently than .
<turion>
That makes sense. I guess I can just put all my .nix files in the same directory as the python files
bvdw has quit [Read error: Connection reset by peer]
globin_ is now known as globin
<bqv[m]>
can i ask what you're building?
globin has joined #nixos
globin has quit [Changing host]
bvdw has joined #nixos
xkapastel has joined #nixos
<turion>
bqv[m]: Gladly :) My long plan is to write a function that takes a .py file, Python-packages it in a trivial way, and then calls buildPythonPackage on it
mananamenos has joined #nixos
<turion>
My colleagues have a lot of .py files floating about, but they aren't packaged
MichaelRaskin has joined #nixos
<turion>
So I'm trying to package them
<turion>
But I want the packaging to happen in nix, in order to leave the source repo unchanged
<mananamenos>
hi, services.postgresql.inititialScript does not succeed. Is it because I had already started postgresql before editing this line? How can I remove/rebuild the whole postgresql service from zero?
<turion>
My function will have to pass those along, I guess
<turion>
bqv[m]: Maybe making a derivation out of the whole repo is fine
<turion>
You mean pinning a commit of that repo, and then getting the files out from there?
<bqv[m]>
i feel the most sensible way of handling that situation is to have a nix file that takes that repo as it's src, and builds several bespoke derivations for each python script, perhaps
<bqv[m]>
if you're against pinning a specific commit, you can use fetchTarball, but I'd just use a specific ref if you can
Naptra has quit [Remote host closed the connection]
<turion>
Yeah I guess it's fine to do that. I'm trying to change the original source as little as possible
<turion>
bqv[m] ++
<bqv[m]>
:)
<turion>
bqv[m]++
<{^_^}>
bqv[m]'s karma got increased to 3
arstneio has joined #nixos
reallymemorable has joined #nixos
<turion>
Oh no... how do I pass credentials to fetchgit?
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<adisbladis>
turion: You don't
<adisbladis>
turion: Use builtins.fetchGit
<adisbladis>
That runs in the evaluator, so it has access to your users keys.
teto has joined #nixos
<turion>
adisbladis: Ah :) thanks!
mananamenos has quit [Remote host closed the connection]
<arstneio>
hey, I have `shell.nix` generated from `cabal2nix`. I'm trying to use a different GHC version. When I change `compiler ? "default"`to something else, it tries to build GHC locally. Shouldn't there be a way to get the binary directly from cache? If so, what do I need to change in my `shell.nix`?
<{^_^}>
[nixpkgs] @dadada opened pull request #83769 → WIP: Nixos/dokuwiki multi-site support and additional plugins and templates → https://git.io/Jv7rj
mallox has joined #nixos
<bqv[m]>
arstneio: chances are, hydra isn't cacheing packages for those versions
hmpffff_ has quit [Ping timeout: 272 seconds]
<bqv[m]>
which is reasonable. it shouldn't have to cache packages for every haskell version, the vast majority of people will only ever use the latest
<Yaniel>
and the latest may simply not be built yet if it was just updated
<bqv[m]>
at least you have some caching... emacs packages don't get cached at all...
<bqv[m]>
i've got conflicting definitions in <unknown-file> again
<mudri>
I opened an issue on cabal2nix a week ago and have got no response. The issue affects nixpkgs users, so should I open an issue on the nixpkgs repository too? https://github.com/NixOS/cabal2nix/issues/447
<{^_^}>
cabal2nix#447 (by laMudri, 1 week ago, open): Agda not using caches for primitive modules
<arstneio>
ah ok, so I will have to look at a different version of nixpkgs? What is the best way to check if a version of nixpkgs satisfies all my package version constraints (other than manually checking the nixpkgs github)?
<bqv[m]>
arstneio: no, the derivations you're after have been deleted
zaeph1 has quit [Ping timeout: 258 seconds]
<bqv[m]>
you have to build them
<bqv[m]>
unless you can find them on cachix or someone else has them
Florian[m]4 has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
<kenran>
I am writing a Haskell application with nix integration and I can use `nix-build` to build my binary. The app has some runtime dependencies (e.g. FreeTDS to access the database) which are present inside my nix-shell.
MichaelRaskin has quit [Read error: No route to host]
<kenran>
Now I'm missing one step: make the binary work on the target (NixOS) machine so that I don't have to be inside a nix-shell to get the runtime dependencies. Can you give me a hint on how to do that?
Rusty1 has joined #nixos
<kenran>
I might have taken a wrong turn earlier already with my train of thought, so feel free to correct me on anything I might be doing in a dumb way!
<tilpner>
kenran: How does your application use/find freetds?
<kenran>
tilpner: phew, I'm not sure I understand what you mean. In my shell.nix I have pkgs.freetds in my buildInputs, as well as a nix expression to create an ODBC ini file that contains the FreeTDS driver configuration.
<mananamenos>
hi, pgadmin4 is not on nixos packages, so i use the docker image. It works fine, however once i create some servers in pgadmin, nothing persists after the docker process restarts. Is this normal?
<kenran>
My application then just uses a connection string that in turn says it uses FreeTDS.
<adisbladis>
mananamenos: Presumably you have to map a volume to wherever pgadmin4 saves it's state.
<bqv[m]>
i never liked docker
<mananamenos>
adisbladis: thank you, i'll google for that. This would go somewhere with `docker run ....`, right?
globin has quit [Ping timeout: 240 seconds]
<tilpner>
kenran: I'm not familiar with freetds. You need to figure out who loads freetds how. Presumably you use some Haskell library, that does the loading for you, right?
<tilpner>
kenran: Have you looked at the option environment.freetds?
<tilpner>
" When at least one entry is configured the global environment variables FREETDSCONF,
<tilpner>
FREETDS and SYBASE will be configured to allow the programs that use freetds to find the library and config.
<tilpner>
", this seems like your library might expect these variables in the environment
<tilpner>
kenran: pymssql on the other hand, uses propagatedBuildInputs = [freetds];
cosimone has quit [Remote host closed the connection]
cosimone has joined #nixos
<kenran>
tilpner: I think I'm misunderstanding you (not a native English speaker here). Yes, the Haskell odbc library does the loading, and it knows it has to use FreeTDS since the connection string says so. When I enter a nix-shell, a nix expression I wrote is evaluated that creates an odbc ini file which contains the location of FreeTDS (I can post this soon)
<kenran>
Also, I set an environment variable ODBCSYSINI to the path to that file, so it's used by odbc.
<kenran>
I wrote this, but it was some time ago and I don't really understand the "magic" that makes me able to use ${freetds} there :)
<pie_[bnc]>
kenran: freetds comes in scope from `with pkgs;`
<kenran>
When I nix-build my application, it has nothing to do with my shell.nix file (maybe that's wrong?), so it doesn't know about FreeTDS in my system. I now wonder whether I should just add freetds to my NixOS system globally, but feel like that's wrong, or how I can "teach the binary" that's built that it needs freetds with the above config expression and such
<pie_[bnc]>
kenran: when you use ${} to substitute a variable in a string, it coerces the first output derivation to a path
<kenran>
The expression I posted is not used in conjunction with my `nix-build` at all and I don't know if and how I can or should use it
Naptra has quit [Remote host closed the connection]
Mbarak has joined #nixos
<pie_[bnc]>
> pkgs.hello.outputs
<{^_^}>
[ "out" ]
<pie_[bnc]>
ok not the best example but yeah
Mbarak has left #nixos [#nixos]
<pie_[bnc]>
some packages have their outputs in a different order so the default coercion is not always .out
Mbarak has joined #nixos
<tilpner>
kenran: Okay, so let's just set ODBCSYSINI to that file, just for your application
<tilpner>
kenran: You can either do that with a wrapper (as shown earlier), or when deploying (how are you deploying? systemd service?)
<mananamenos>
adisbladis: Sorry maybe you also know what should the target be for pgadmin `--mount source=my-vol,target=?? \`?
<kenran>
I'm not sure how I want to deploy it (partly because I haven't figured out how to best run the app yet :D). systemd would be an option, yes. for this app, it might be feasible to just have an open tmux session where it runs in one tab though...
<tilpner>
kenran: Well, those two questions need to be answered at the same time, unfortunately
captjakk has joined #nixos
<tilpner>
(Unless you wrap it, then it works everywhere)
<tilpner>
kenran: With a systemd service, you can set environment.ODBCSYSINI = callPackage ./odbcinst.nix {};
<tilpner>
kenran: And then, when you start the application inside that service, it should find freetds
<tilpner>
That option refers to systemd.services.<name>.environment
<mananamenos>
adisbladis: thank you!
domogled has quit [Remote host closed the connection]
domogled has joined #nixos
rauno has joined #nixos
<kenran>
tilpner: the wrapping sounds like something I could try first and maybe be content with
captjakk has quit [Ping timeout: 246 seconds]
<kenran>
but systemd looks better because I could then write an expression for my service and have that in the description of my server
<kenran>
tilpner: thank you very much! I will try around with both possibilities for a bit and ask follow-up questions in a couple of days when I got to trying it out :)
<{^_^}>
[nixpkgs] @adisbladis opened pull request #83775 → pulseaudio: Enable bluetooth support by default → https://git.io/Jv76v
cfricke has quit [Quit: WeeChat 2.7.1]
jeromelanteri has joined #nixos
bennofs1 has quit [Ping timeout: 265 seconds]
dermetfan has quit [Ping timeout: 246 seconds]
<jeromelanteri>
hi, new install on qemu VM. I'm trying to setup keyboard layout on first opened tty... how to do please ?
<jeromelanteri>
doesn't find any manual page around this. Only find for Xserver keyboard layout.
captjakk has joined #nixos
<jeromelanteri>
first login after to install is nixos@nixos shell login.
<tilpner>
jeromelanteri: loadkeys foo, to do it manually
<tilpner>
(where foo could be "de")
<tilpner>
And console.keyMap to do it permanently
<jeromelanteri>
permission denied (i tryed it allready)
<tilpner>
As root?
<jeromelanteri>
tilpner, no becuase su command ask for a password i'm ignoring.
globin has joined #nixos
<jeromelanteri>
tilpner, ok, i do a sudo call...
<jeromelanteri>
on first install what is the root password ?
<tilpner>
You set it during installation
<jeromelanteri>
no
<tilpner>
There is no default password
<jeromelanteri>
ask me nothing during install
<tilpner>
Did you set it in configuration.nix?
<jeromelanteri>
no
<tilpner>
Then it should have made you set one
<jeromelanteri>
ok
captjakk has quit [Ping timeout: 256 seconds]
<jeromelanteri>
it is the very first install pretty new there on my VM. No question was asked.
<tilpner>
Did you already install, or are you still installing?
<jeromelanteri>
find configuration.nix
cr4y1 has quit [Read error: Connection reset by peer]
<the_pumpkin_man[>
Hello, I am actually on nixos-unstable, and I don't understand why when I launch lutris, there's no icon. I also tried on nixos-stable channel
<pie_[bnc]>
jeromelanteri: have you installed already, or are you in the installer live environment?
CcxWrk_ has joined #nixos
CcxWrk has quit [Killed (weber.freenode.net (Nickname regained by services))]
CcxWrk_ is now known as CcxWrk
<jeromelanteri>
pie_[bnc], ok, i just realized that i'm on a kind of tty installe rlive environment. it is not installed allready.
LysergicDreams has joined #nixos
<pie_[bnc]>
(i dont think it should have asked you for a password then, but idk)
knupfer has quit [Ping timeout: 240 seconds]
alanz has quit [Ping timeout: 240 seconds]
r0bby has quit [Ping timeout: 240 seconds]
teozkr has quit [Ping timeout: 240 seconds]
{^_^} has quit [Read error: Connection reset by peer]
alexfmpe[m] has quit [Ping timeout: 240 seconds]
psiperator[m] has quit [Ping timeout: 240 seconds]
petersjt014[m] has quit [Ping timeout: 240 seconds]
bendlas has quit [Ping timeout: 240 seconds]
bridge[evilred] has quit [Ping timeout: 240 seconds]
joshmeredith has quit [Ping timeout: 240 seconds]
wucke13 has quit [Ping timeout: 240 seconds]
boogiewoogie[m] has quit [Ping timeout: 240 seconds]
akapaka[m] has quit [Ping timeout: 240 seconds]
abbradar[m] has quit [Ping timeout: 240 seconds]
sephii has quit [Ping timeout: 240 seconds]
ongy[m] has quit [Ping timeout: 240 seconds]
quidel[m] has quit [Ping timeout: 240 seconds]
pantomath[m] has quit [Ping timeout: 240 seconds]
simbergm has quit [Ping timeout: 240 seconds]
l_inus[m] has quit [Ping timeout: 240 seconds]
dreadward[m] has quit [Ping timeout: 240 seconds]
theduke[m] has quit [Ping timeout: 240 seconds]
mindtree[m] has quit [Ping timeout: 240 seconds]
gcoakes[m] has quit [Ping timeout: 240 seconds]
Cadey has quit [Ping timeout: 240 seconds]
Leira[m] has quit [Ping timeout: 240 seconds]
tek[m] has quit [Ping timeout: 240 seconds]
ericsagnes has quit [Ping timeout: 240 seconds]
noj has quit [Ping timeout: 240 seconds]
zgrep has quit [Ping timeout: 240 seconds]
VaNilLa[m] has quit [Ping timeout: 240 seconds]
chris|1 has quit [Ping timeout: 240 seconds]
timokau[m] has quit [Ping timeout: 240 seconds]
LeshaInc has quit [Ping timeout: 240 seconds]
rot13 has quit [Ping timeout: 240 seconds]
tghume has quit [Ping timeout: 240 seconds]
hax404 has quit [Ping timeout: 240 seconds]
leah2 has quit [Ping timeout: 240 seconds]
kyren has quit [Ping timeout: 240 seconds]
yorick has quit [Ping timeout: 240 seconds]
ardumont has quit [Ping timeout: 240 seconds]
sputny[m] has quit [Ping timeout: 240 seconds]
gudea[m] has quit [Ping timeout: 240 seconds]
Mic92 has quit [Ping timeout: 240 seconds]
drewr has quit [Ping timeout: 240 seconds]
fasd has quit [Ping timeout: 240 seconds]
index-0[m] has quit [Ping timeout: 240 seconds]
davidak[m] has quit [Ping timeout: 240 seconds]
bohan[m] has quit [Ping timeout: 240 seconds]
slabity has quit [Ping timeout: 240 seconds]
atemu12[m] has quit [Ping timeout: 240 seconds]
srid[m] has quit [Ping timeout: 240 seconds]
imj[m] has quit [Ping timeout: 240 seconds]
unclechu has quit [Ping timeout: 240 seconds]
thefloweringash has quit [Ping timeout: 240 seconds]
aniketd[m] has quit [Ping timeout: 240 seconds]
yoho[m]1 has quit [Ping timeout: 240 seconds]
cocreature has quit [Ping timeout: 240 seconds]
`slikts has quit [Ping timeout: 240 seconds]
janneke has quit [Ping timeout: 240 seconds]
nikola1 has quit [Ping timeout: 240 seconds]
awygle has quit [Ping timeout: 240 seconds]
fooker has quit [Ping timeout: 240 seconds]
farn has quit [Ping timeout: 240 seconds]
ShaRose has quit [Ping timeout: 240 seconds]
cosimone has quit [Read error: Connection reset by peer]
talyz has quit [Ping timeout: 240 seconds]
aranea has quit [Ping timeout: 240 seconds]
ShaRose has joined #nixos
knupfer1 is now known as knupfer
janneke has joined #nixos
LeshaInc has joined #nixos
bendlas1 has joined #nixos
sephii has joined #nixos
teozkr_ has joined #nixos
alanz_ has joined #nixos
hax404 has joined #nixos
Mic92 has joined #nixos
tghume has joined #nixos
noj has joined #nixos
janneke has joined #nixos
janneke has quit [Changing host]
nikola1 has joined #nixos
waleee-cl has joined #nixos
ardumont has joined #nixos
cosimone_ has joined #nixos
talyz has joined #nixos
`slikts has joined #nixos
tokudan_ has joined #nixos
cocreature has joined #nixos
`slikts has joined #nixos
`slikts has quit [Changing host]
bridge[evilred] has joined #nixos
aranea has joined #nixos
Nixola has quit [Remote host closed the connection]
wucke13 has joined #nixos
r0bby_ has joined #nixos
yorick has joined #nixos
joshmeredith has joined #nixos
Nixola has joined #nixos
ericsagnes has joined #nixos
fooker has joined #nixos
awygle has joined #nixos
Nixola is now known as Guest42810
<jeromelanteri>
pie_[bnc], did asked for nothing at all... just a kind of boot menu for installer mode... just the next step.
Cadey has joined #nixos
farn has joined #nixos
srid[m] has joined #nixos
zgrep has joined #nixos
<tilpner>
No, it asks for a password while installing
<alexgood>
Hey folks, I'm getting an error running gpg. Specifically I'm getting a `no pinentry` error. I have `services.gnupg.agent.enable = true` in my config, anything I'm missing?
<tilpner>
If you haven't done that, it hasn't asked yet
tokudan has quit [Ping timeout: 240 seconds]
<jeromelanteri>
tilpner, i did remove and restart the process for be 100% sure...
<jared-w>
linux 5.6 is out now. Yey
petersjt014[m] has joined #nixos
<jeromelanteri>
tilpner, so yes, the first step is to see a kind of boot menu to choose install mode
<jeromelanteri>
pie_[bnc], no, just see your link now. I'm surprise because there is something indicate on first tty opened, but if i strictely follow what they said, it failed. Or user have to know something who is not writed...
<jeromelanteri>
ALT-F3 doesn't change anything because of the autologin on nixos.
<pie_[bnc]>
tilpner: i guess i also kind of lost track of what the problem is
<pie_[bnc]>
jeromelanteri: you fixed the keyboard, so what is the problem now?
<jeromelanteri>
what is poorly documented and push first time user down is this sentence: "You then must set a password for either "root" and "nixos" with "passwd" to be able to login.
ghasshee has quit [Ping timeout: 240 seconds]
<jeromelanteri>
well...
cosimone has quit [Remote host closed the connection]
<pie_[bnc]>
jeromelanteri: that is for the installed system
<jeromelanteri>
no, sure not, write passwd command just change the nixos user passwd.
cosimone has joined #nixos
<jeromelanteri>
no root user passwd has been changed yet.
<jeromelanteri>
so what ?
cosimone has quit [Client Quit]
<pie_[bnc]>
i mean, thats for if you want to log in to the installed system, when you boot into the installed sytem
<jeromelanteri>
ok
cosimone has joined #nixos
<pie_[bnc]>
possibly involving a chroot and passwd before you reboot
<jeromelanteri>
ok, understand better.
<pie_[bnc]>
but it's been a while since the last time i did this
cosimone has quit [Client Quit]
cosimone has joined #nixos
<jeromelanteri>
It should be indicate to clarify the situation. First time user can't know.
<pie_[bnc]>
if i understood tilpner correctly, nixos-install or whatever will ask you for a password to set?
<pie_[bnc]>
jeromelanteri: have you actually gone through the whole nixos-install process?
<pie_[bnc]>
jeromelanteri: see the part in the manual that says "Do the installation: "
<pie_[bnc]>
thats where the password stuff is
<jeromelanteri>
pie_[bnc], i should take time to read your link first or maybe all the nixos-help document and write the first step usable information from this doc to be able to understand what to do at first step to install correctly...
drewr has joined #nixos
<jeromelanteri>
pie_[bnc], i gone no where more than just start iso fresh downloaded image on qemu VM and click on first link boot loader like screen to get this nixos automatic login on tty
<pie_[bnc]>
jeromelanteri: right, so you probably want to start reading/going through the instalation section of the manual
<pie_[bnc]>
feel free to ask questions if you have any problems, and welcome to nixos \o/ :P
<jeromelanteri>
and God didn't tell me nothing more than what i read on tty top text poorly documented, the next is to read the big documentation and find what i have to do next.
globin_ is now known as globin
<pie_[bnc]>
nixos is not the most user friendly thing
globin has joined #nixos
globin has quit [Changing host]
<jeromelanteri>
yes pie_[bnc] no other choice to stop now and going back when i will have a full day time to investigate to learn something usuall before to go ahead. The doc is good i see, but i have to read a lot for find the first next step to first install NixOS.
frhun has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
<pie_[bnc]>
jeromelanteri: have you ever installed linux befre
<jeromelanteri>
pie_[bnc], yes, i can see that, but the idea of NixOS is still very good. I think later someone would increase sentences at first connection and make it user friendly.
<pie_[bnc]>
yeah, people want to, theres just always more work than people :P
<jeromelanteri>
pie_[bnc], i'm on archlinux, i use also FreeBSD on VPS servers, and i did install any Debian like distro on ARM hardware, etc... so yes, i have a little idea of what is Unix and Linux systems since 20 years more now.
<jeromelanteri>
pie_[bnc], sure.
<pie_[bnc]>
the main steps for nix are not completely alien to linux setup, you have to do partitioning (pay attention to GPT/MBR UEFI bootloader compatibility stuff), mount the needed partitions, generate the initial config, edit the config, and run nixos-install
<nix-build>
[nix] @edolstra pushed to master « Remove global -I flags »: https://git.io/Jv7iF
<pie_[bnc]>
its not hard once you;ve done it a few times x)
<jeromelanteri>
i'll just do BIOS MBR easy way on Qemu VM.
<pie_[bnc]>
should be fine
<alexgood>
Hey folks, apologies for repeating request, not sure if it got missed. Anyone know how to gpg running under gnome?
gustavderdrache has joined #nixos
<glitters`>
alexgood: what's special about gnome?
kriteus[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
glitters` is now known as glittershark
<pie_[bnc]>
alexgood: idk, butwhat problem are you running into
<jeromelanteri>
pie_[bnc], when i will have time, maybe i would like to create a kind of very first beginner video to show how to first step install to help people.
<pie_[bnc]>
jeromelanteri: sure
<alexgood>
I've got `programs.gnupg.agent.enable = true` set but running `gpg` gives me `no pinentry` errors
<alexgood>
Not necessarily anythung special, it's just the environment I'm in
<glittershark>
oh hmm
<pie_[bnc]>
jeromelanteri: looks like youtube already has some actually
<glittershark>
alexgood: is that *all* you have enabled wrt gpg?
<jeromelanteri>
pie_[bnc], ho... good. In French langage ?
<alexgood>
Specifically, `gpg --import` gives me a bunch of `error readiong <secret>: No pinentry`
<alexgood>
I've also added `gnupg` to `systemPackages`
<glittershark>
are you using home-manager?
<alexgood>
Yup
cosimone has quit [Remote host closed the connection]
<glittershark>
so I use `services.gpg-agent.enable = true;` in my home.nix
<glittershark>
and that works for me
<pie_[bnc]>
jeromelanteri: hmm ok i thought they exsted but the first few i looked at werent great
<alexgood>
I'll give it a shot
cosimone has joined #nixos
<pie_[bnc]>
so nevermind i guess, make the video :p
<pie_[bnc]>
i think someone here wrote an article on nix for a french magazine once but i dont remember who it was
<pie_[bnc]>
or maybe it was just on the discourse
rauno has quit [Remote host closed the connection]
<jeromelanteri>
pie_[bnc], no pb, i will do it. I see i just need real time to look closer first to preparate something not too long to show and also something clear and simple. It is not easy to do things clear and simple.
<alexgood>
There should be a gpg agent running under systemctl right? I can't see one
<alexgood>
Yeah `no pinentry`
<glittershark>
alexgood: it's in `systemctl --user`, if you use home-manager
<glittershark>
you may need to `systemctl --user start gpg-agent` after running home-manager switch
<alexgood>
Ah nice, it is running
<glittershark>
cool
<pie_[bnc]>
mbarak: it might be reasonable to ask once in #nixos-dev
mvnetbiz_ has joined #nixos
<glittershark>
try tweaking the values of services.gpg-agent.pinentryFlavor in home.nix
<Mbarak>
pie: thanks! i'll try that.
<glittershark>
Type: null or one of "curses", "tty", "gtk2", "qt", "gnome3", "emacs"
<jeromelanteri>
pie_[bnc], thank you.
Mbarak has left #nixos [#nixos]
Mbarak has joined #nixos
<glittershark>
gtk2 is the default
<alexgood>
Will do. There are a bunch of logs in the service filesaying "IPC connect call failed"
<alexgood>
So that sounds promising
<glittershark>
yeah
<alexgood>
*service journal
<glittershark>
tty will almost definitely work, if you can stand having to type your password into a terminal
atopuzov[m] has quit [Quit: Idle for 30+ days]
<glittershark>
I've run into this exact issue on macos, fwiw
<glittershark>
not running nix at all
Criena has quit [Quit: Idle for 30+ days]
<glittershark>
maybe try disabling `programs.gnupg.agent` in your configuration.nix too?
flacks_ has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
MerlinGttlinger[ has quit [Quit: Idle for 30+ days]
nocoolnametom has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
marijan[m] has quit [Quit: Idle for 30+ days]
<alexgood>
So I'm getting a weird error trying to build the pinentryFlavor option, it's complaining that `home-manager.users.alex.services.gpg-agent.pinentryFlavor` doesn't exist
aveltras has joined #nixos
<glittershark>
is it nested under users.alex?
<glittershark>
because it shouldn't be
captjakk has joined #nixos
<rycee>
alex: You are using the master branch of Home Manager with the 19.09 version of Nixpkgs. You need to use the 19.09 branch of Home Manager.
mvnetbiz_ has quit [Quit: Bye!]
<aveltras>
having the following configuration.nix, is it possible to only resolve the path to ssh keys on the remote machine ? im building it on local machine and it complains that it can't access the root directory but i don't mean
<avn>
adisbladis: backing to node packages... I checked node2nix and napalm, but not sure if possible to avoid 500kb package-lock? If I undersrtand right -- instalign dev tools is recursive, and bring half on npmjs
<aveltras>
turion: the thing is i'd like to build the remote config completely on local machine
captjakk has quit [Ping timeout: 252 seconds]
<glittershark>
oh, I didn't know that was possible! that's cool. No, you're probably doing the right thing.
<glittershark>
<bqv[m]>
<glittershark "because it shouldn't be"> yes, it should
<glittershark>
every time I sit down at someone else's computer or ssh into a production machine and my aliases aren't there I have a very hard time
<ottidmes>
jakobrs: that would have been an implementation detail and they could have implemented it as (meta.maintainers or []), so no, I am sure they mean, it should have a maintainer + I have seen reviewers ask for this and maintainers being added to those that did not have any but were already in nixpkgs
alexgood has left #nixos ["User left"]
<ottidmes>
bqv[m]: The biggest downside I notice from aliases+functions is that I sometimes want to share a one liner, which uses those, and it takes some effort to then rewrite it back to the underlying commands, especially for some of the functions I use
<mac66>
I have a newbie question. I have a late 2019 Razer Blade Stealth with an Intel iGPU and an NVIDIA GTX1650 discrete GPU. NixOS 19.09 and unstable work fine with the iGPU but I cannot get the discrete GPU working. Latest attempt is the nvidia-offload configuration which seems to work fine in Arch. I have no idea what the difference is.
<bqv[m]>
exactly
<bqv[m]>
i prefer to just get faster at typing the original
<bqv[m]>
that way the alias is in my head, not tied to a machine
<mac66>
tilpner: works fine with iGPU, get the error below with nvidia. I suspect I borked the PCI id maybe, or something like that.
<mac66>
nvidia-offload glxspheres64Polygons in scene: 62464 (61 spheres * 1024 polys/spheres)GLX FB config ID of window: 0x8c (8/8/8/0)Visual ID of window: 0x11fX Error of failed request: BadValue (integer parameter out of range for operation) Major opcode of failed request: 152 (GLX) Minor opcode of failed request: 3 (X_GLXCreateContext) Value in
<mac66>
failed request: 0x0 Serial number of failed request: 27 Current serial number in output stream: 28
linarcx has joined #nixos
<adisbladis>
oznt: What do you want to do? Are you sure you cant express what you want in Nix?
<tilpner>
Oh, it binds nix, nice
<oznt>
I would like to add elements in a nix config without installing nix os
<oznt>
a python lib would be cool because I give the users a script to manipulate nix file without having to install nix
<glittershark>
like structural editing of nix scripts
<tilpner>
mac66: And you checked the PCI ids yourself?
julm has quit [Quit: leaving]
<bqv[m]>
i think nix even has encryption support
<nix-build>
[nix] @domenkozar pushed to fix-transient-errors « downloader: when retry but can't resume, retry the whole file »: https://git.io/Jv71H
<bqv[m]>
rfc 5
quidel[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
julm has joined #nixos
fendor has quit [Ping timeout: 246 seconds]
<oznt>
adisbladis, when you write Nix, what exactly do you mean?
<mac66>
tilpner: Yes. But maybe I got it wrong? lspci says the nvidia card is 57:00.0 and I put it in as PCI:87:0:0 in my configuration.nix
<mac66>
Intel GPU is 0:2:0
<tilpner>
Is that a typo, or did you actually use different digits?
<adisbladis>
oznt: I mean use `lib.importJSON` to read the values from your JSON file
<mac66>
Not a typo, the wiki said nvidia driver wants decimal
<pie_[bnc]>
jeromelanteri: i looked at the message you were talking about, thats saying you have to set a password to be able to ssh into the installer image
<tilpner>
mac66: Well, 57 != 87
<oznt>
sounds fine, ok, can you give a link to a working example?
<nix-build>
[nix] @domenkozar opened pull request #3457 → downloader: when retry but can't resume, retry the whole file → https://git.io/Jv71F
chloekek has quit [Ping timeout: 246 seconds]
<mac66>
is 57 hex not 87 decimal
<mac66>
?
<tilpner>
mac66: Huh, are you sure it needs to be encoded? My ids are small enough to be the same in both encodings
<adisbladis>
oznt: Something like `{ lib, ... }: let userJSON = lib.importJSON ./jsonfile.json; in { users.users = userJSON; }`
<nix-build>
[nix] @domenkozar pushed to fix-transient-errors « downloader: when retry but can't resume, retry the whole file »: https://git.io/Jv71h
<mac66>
But I guess I could easily try with 57
<teto>
what does the delta represnet in this lib.debug.traceSeq output ? "CSApprox" = <δ:vimplugin-csapprox-2013-07-27>
<mac66>
tilpner: This is puzzling. Something else must be missing because I am using the same desktop environment on the same machine and the same NV environment variables and it worked as when it worked on Arch.
magnetophon has quit [Ping timeout: 256 seconds]
<mac66>
lspci also says the nvidia driver is loaded for the dGPU
<oznt>
adisbladis, looks like a lot of effort ... I would appreciate a more complete example.
<mac66>
*as when it worked on Arch
<adisbladis>
oznt: Without a more specific question I can't provide that
<tilpner>
mac66: Sorry, I probably can't help you much :/
cosimone has quit [Quit: Terminated!]
<mac66>
tilpner: Thanks for trying!
LysergicDreams has quit [Ping timeout: 256 seconds]
<michaelpj>
anyone using plasma on recent unstable? I'm scared of updating because of https://github.com/NixOS/nixpkgs/issues/79713 as I do kind of need to be able to configure displays...
<nix-build>
#79713 (by y-usuzumi, 7 weeks ago, open): Theme is broken on several tabs of systemsettings5
captjakk has quit [Ping timeout: 256 seconds]
<adisbladis>
And it's really no less work than something that parses Nix and emits new Nix
<adisbladis>
s/no less/no more/
<adisbladis>
In fact its very likely far less work
LysergicDreams has joined #nixos
mac66 has quit [Ping timeout: 240 seconds]
<nix-build>
[nixpkgs] @mkg20001 opened pull request #83782 → krita: disable parallel building #35359 → https://git.io/Jv7Mo
<the_pumpkin_man[>
Hello, I am actually on nixos-unstable, and I don't understand why when I launch lutris, there's no icon. I also tried on nixos-stable channel
captjakk has joined #nixos
justanotheruser has quit [Ping timeout: 252 seconds]
fenedor has quit [Read error: Connection reset by peer]
noonereally[m] has quit [Ping timeout: 240 seconds]
M4cheese[m] has quit [Ping timeout: 240 seconds]
sleibrock[m] has quit [Ping timeout: 240 seconds]
hiroshi[m] has quit [Ping timeout: 240 seconds]
risebell[m] has quit [Ping timeout: 240 seconds]
HailSagan[m] has quit [Ping timeout: 240 seconds]
kraem[m] has quit [Ping timeout: 240 seconds]
ryanmaclean[m] has quit [Ping timeout: 240 seconds]
matthewbauer has quit [Ping timeout: 240 seconds]
Irenes[m] has quit [Ping timeout: 240 seconds]
worldofpeace has quit [Ping timeout: 240 seconds]
mica[m] has quit [Ping timeout: 240 seconds]
Notkea[m] has quit [Ping timeout: 240 seconds]
killercup[m] has quit [Ping timeout: 240 seconds]
ecou[m] has quit [Ping timeout: 240 seconds]
maljub01[m] has quit [Ping timeout: 240 seconds]
DenisLehmann[m] has quit [Ping timeout: 240 seconds]
layus[m] has quit [Ping timeout: 240 seconds]
jamiemagee has quit [Ping timeout: 240 seconds]
nrdxp[m] has quit [Ping timeout: 240 seconds]
truby has quit [Ping timeout: 240 seconds]
isgy[m] has quit [Ping timeout: 240 seconds]
fstamour[m] has quit [Ping timeout: 240 seconds]
TroelsAgergaardJ has quit [Ping timeout: 240 seconds]
wedens[m] has quit [Ping timeout: 240 seconds]
thequux[m] has quit [Ping timeout: 240 seconds]
st3r4g[m] has quit [Ping timeout: 240 seconds]
nolan_d has quit [Ping timeout: 240 seconds]
MilkManzJourDadd has quit [Ping timeout: 240 seconds]
colemickens has quit [Ping timeout: 240 seconds]
kdedev[m] has quit [Ping timeout: 240 seconds]
puzzlewolf has quit [Ping timeout: 240 seconds]
turion has quit [Quit: Leaving.]
Fulgen has joined #nixos
zaeph1 has joined #nixos
gulplante has joined #nixos
aquarial has joined #nixos
Niall has joined #nixos
risebell[m] has joined #nixos
gulplante has quit [Client Quit]
HailSagan[m] has joined #nixos
Makaveli7 has quit [Quit: WeeChat 2.7.1]
kraem[m] has joined #nixos
Pidgeotto has quit [Remote host closed the connection]
<teto>
the_pumpkin_man[: maybe the icon is not installed by the package. Or do you mean game icons ?
Pidgeotto has joined #nixos
sleibrock[m] has joined #nixos
justanotheruser has joined #nixos
linarcx has quit [Quit: WeeChat 2.8]
Asmadeus has joined #nixos
justanotheruser is now known as Guest32855
<bqv[m]>
uh hey, curl is segfaulting
<bqv[m]>
that's not normal right?
zaeph has quit [Ping timeout: 258 seconds]
<bqv[m]>
nevermind, was using it wrong...
mt[m] has joined #nixos
jamiemagee has joined #nixos
wedens[m] has joined #nixos
noonereally[m] has joined #nixos
Irenes[m] has joined #nixos
kdedev[m] has joined #nixos
matthewbauer has joined #nixos
colemickens has joined #nixos
<bqv[m]>
https://0x0.st/iMMC.png : left=been using rebases and tags to keep track of my nix configurations, right=the nix elisp i was talking about earlier
M4cheese[m] has joined #nixos
worldofpeace has joined #nixos
hiroshi[m] has joined #nixos
thequux[m] has joined #nixos
nrdxp[m] has joined #nixos
TroelsAgergaardJ has joined #nixos
isgy[m] has joined #nixos
Mbarak has quit [Remote host closed the connection]
DenisLehmann[m] has joined #nixos
truby has joined #nixos
sepi``` has joined #nixos
layus[m] has joined #nixos
linarcx has joined #nixos
MichaelRaskin has joined #nixos
boegel has joined #nixos
<ottidmes>
bqv[m]: That is one of the reasons I try to avoid emacs, what if I really start to like it. I would likely waste too much time improving my emacs setup, I already do that plenty enough with NixOS :P
gulplante has joined #nixos
nolan_d has joined #nixos
ryanmaclean[m] has joined #nixos
st3r4g[m] has joined #nixos
Notkea[m] has joined #nixos
MilkManzJourDadd has joined #nixos
killercup[m] has joined #nixos
bohan has joined #nixos
drozdziak1 has joined #nixos
linarcx has quit [Client Quit]
KennosiS has joined #nixos
fstamour[m] has joined #nixos
mica[m] has joined #nixos
captjakk has joined #nixos
ecou[m] has joined #nixos
puzzlewolf has joined #nixos
maljub01[m] has joined #nixos
Shados has joined #nixos
Shados_ has quit [Ping timeout: 246 seconds]
Fulgen has quit [Ping timeout: 246 seconds]
greizgh has quit [Ping timeout: 240 seconds]
<bqv[m]>
:D
ar has quit [Ping timeout: 260 seconds]
<the_pumpkin_man[>
teto: I mean the icon on the top bar. For example, the cross you click to make the window to exit, or to minimize the window. Also other buttons from lutris, like add game/import game, etc
<bqv[m]>
by the way, is there a `nixos-option search` or something
<bqv[m]>
i'm so sick of having to open a browser to do that
raingloom has quit [Ping timeout: 256 seconds]
<infinisil>
gchristensen: Where did {^_^} go?
<srk>
bqv[m]: man configuration.nix and / ?
<hodapp>
TIL: there is a 'man configuration.nix'
ddellacosta has joined #nixos
<bqv[m]>
i'm assuming it's nix-build
nschoe has quit [Ping timeout: 256 seconds]
<hodapp>
sometimes I forget about man pages because certain programs don't provide them well :[
<gchristensen>
infinisil: for now its nix-build, I'll look in a bit
<bqv[m]>
srk: oh that's neat.
<srk>
that's documentation.nixos.enable :)
<bqv[m]>
i'm guessing that's just nixpkgs though
cole-h has joined #nixos
luelista has joined #nixos
<bqv[m]>
hodapp: smae
<bqv[m]>
see: "man nix"
<hodapp>
bqv[m]: see above :P
Mbarak has joined #nixos
<srk>
bqv[m]: I think it works for import-ed modules as well
Mbarak has quit [Remote host closed the connection]
<NobbZ[m]>
I have a `shell.nix` which uses `python35Packages.poetry` as its `buildInputs` and it seems as if this causes a lot of cerivations to be built from scratch, including running the tests. Is there a way to disable tests globally (for this `shell.nix`)?
<srk>
hm, not sure. but I also have 'man home-configuration.nix' provided by home-manager
SuppenGeist has joined #nixos
edef__ has joined #nixos
edef__ is now known as edef
<bqv[m]>
srk: doesn't seem to, for me, but i'm using flakes
<bqv[m]>
i have documentation.nixos.enable set to true
<nix-build>
[nixpkgs] @wmertens pushed 4 commits to staging-19.09: https://git.io/Jv7yP
zaeph2 has joined #nixos
<SuppenGeist>
Hey, I'm currently trying to write my first nix module. I created a basic default.nix and currently trying to evaluate it with the command 'nix-shell $NIXPKGS -A packagename'. But nix-shell doesnt seem to recognize my default.nix. I also set $NIXPKGS to my local nixpkgs repository.
<srk>
NobbZ[m]: not that easy. you might be able to do that by using overlay to override python35Packages.callPackage but that sets it for all python35Packages
zaeph2 is now known as zaeph
<bqv[m]>
oh, cool
<bqv[m]>
guess that'll do
<srk>
bqv[m]: no idea bout that, would be nice if that would work
raingloom has joined #nixos
<infinisil>
SuppenGeist: Where is the default.nix?
luelista has quit [Ping timeout: 260 seconds]
<bqv[m]>
srk: that's totally a great idea though, a manpage generated from all the modules you have available
fragamus has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<bqv[m]>
hey nix gurus, is that possible?
<SuppenGeist>
infinisil I located it in pkgs/applications/version-management/openproject
<bqv[m]>
i mean, i'm sure it's possible, but would it be easy to do?
<qyliss>
bqv[m]: configuration.nix(7)
<infinisil>
SuppenGeist: Then you also need to add a definition in pkgs/top-level/all-packages.nix
<qyliss>
configuration.nix(5), sorry
zaeph1 has quit [Ping timeout: 256 seconds]
Nixola has joined #nixos
liori_ has quit [Ping timeout: 260 seconds]
statusfailed has quit [Ping timeout: 260 seconds]
<SuppenGeist>
infinisil I see, thank you!
Nixola is now known as Guest36206
<bqv[m]>
i do not have a (7)
statusfailed has joined #nixos
<bqv[m]>
i have a (5), it is not all my modules
liori has joined #nixos
fenedor is now known as fendor
zfnmxt has joined #nixos
<NobbZ[m]>
srk: I don't care. I can put that overlay into that `shell.nix` only. Do you have an example how I have to write that override?
fragamus has joined #nixos
sb0_ has quit [Quit: Leaving]
<bqv[m]>
yep, i grepped through, pretty sure every option in configuration.nix(5) is declared by nixpkgs
<bqv[m]>
not a single one by any of the other modules i have
puck has quit [Read error: Connection reset by peer]
<mjrosenb>
so, given that there *should* be macos specific code for idea-ultimate, how can I find out why it installed the linux binary of idea-ultimate?
<pie_[bnc]>
what was the man page with nixos options?
<pie_[bnc]>
ok man configuration.nix
zebrag has joined #nixos
inkbottle has quit [Ping timeout: 252 seconds]
linarcx has quit [Client Quit]
<cole-h>
`nixos-option` might help as well
linarcx has joined #nixos
linarcx has quit [Client Quit]
linarcx has joined #nixos
<nix-build>
[nix] @edolstra pushed to fetchers « Backport libfetchers from the flakes branch »: https://git.io/Jv7Hd
<pie_[bnc]>
cole-h: yeah my nixos-option is broken right now because my NIX_PATH is set up weird
<nix-build>
[nix] @edolstra opened pull request #3459 → Backport libfetchers from the flakes branch → https://git.io/Jv7HA
mrCyborg has joined #nixos
mzan has joined #nixos
<nix-build>
[nixpkgs] @juliendehos opened pull request #83787 → pythonPackages.fcmaes: init at 0.9.5.7 → https://git.io/Jv7Hx
linarcx has quit [Client Quit]
<nix-build>
[nixpkgs] @ryneeverett opened pull request #83788 → vimPlugins: Automate git commits when updating. → https://git.io/Jv7Hj
linarcx has joined #nixos
Rusty1 has quit [Remote host closed the connection]
<gchristensen>
niksnut: oh cool, looking forward to reading that :D
<benny>
can I use a vm or cgroups resource limited sandbox to build a package? my system crashes (I think due to OOM) when I try to build a certain nix package (graalvm8)
<benny>
the former of course somewhat automagically, because I know I can just spin up a VM ;-)
<tilpner>
You can set MemoryHigh instead of MemoryMax and systemd will supposedly take memory away more aggressively, and slow the process down. No idea if and how well that works though
<benny>
tilpner: I'm grasping at straws, because graalvm8 is a dependency for some interesting tools and having system crashes whenever I try to build something is not optimal (tm) ;-)
<benny>
I would rather the build process dies so I can confirm it's a memory situation
<tilpner>
Well, it's not just the build process
<tilpner>
I would expect the entire daemon to die, along with all build processes it had
mehlon has joined #nixos
<tilpner>
Which may be equivalent in your case, but the distinction can be important in others
<nix-build>
[nixpkgs] @Ericson2314 pushed 6 commits to master: https://git.io/Jv75L
naivesheepx has quit [Quit: WeeChat 2.7.1]
zaeph has quit [Ping timeout: 240 seconds]
<mdash>
nixos-generators looks like a great project, but i seem to be missing something important about how to use it to generate a 32bit image from a 64bit host
pbogdan has joined #nixos
zaeph has joined #nixos
<mdash>
i'm getting "can't open fsimg nixos.raw: Value too large for defined data type"
reallymemorable has quit [Quit: reallymemorable]
<nix-build>
[nixpkgs] @Ericson2314 opened pull request #83794 → tensorflow-estimator-1.x: Put back comment → https://git.io/Jv75s
<mdash>
maybe it's trying to make a >4GB image?
<nix-build>
[nixpkgs] @risicle opened pull request #83795 → nghttp2: add python bindings as pythonPackages.nghttp2 → https://git.io/Jv75n
<nix-build>
[nixpkgs] @Ericson2314 merged pull request #83794 → tensorflow-estimator-1.x: Put back comment → https://git.io/Jv75s
<nix-build>
[nixpkgs] @Ericson2314 pushed 2 commits to master: https://git.io/Jv75c
<lassulus>
mdash: hmm, I never tried generating a 32bit image, what command are you trying?
reallymemorable has joined #nixos
<nix-build>
[nixpkgs] @Ericson2314 opened pull request #83796 → Backport #83518 for 20.03 → https://git.io/Jv750
chagra_ has joined #nixos
opthomasprime has joined #nixos
nschoe has joined #nixos
zupo_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<__monty__>
With modules like `services.borgbackup.jobs` is it possible to add to the systemd service definition? I'd like to add Restart= and RestartSec= directives.
knupfer has joined #nixos
pnobre has joined #nixos
<nix-build>
[nixpkgs] @bjornfor pushed commit from @r-ryantm to master « plantuml: 1.2020.2 -> 1.2020.5 »: https://git.io/Jv75S
<cole-h>
__monty__: If it's possible to set Restart and RestartSec in a user-defined service, you should just be able to override the borgbackup service with those things
<cole-h>
e.g. `systemd.user.services.borgbackup.<however you can set Restart> = whatever;` is how I'm imagining it
<cole-h>
(but I don't know for sure, sorry)
pbb has quit [Remote host closed the connection]
lovesegfault has joined #nixos
pbb has joined #nixos
<mdash>
lassulus: I was seeing "booted via startup_32" in the console when trying to load a 64bit image and no messages from linux kernel, so I assumed it was an arch issue
<lassulus>
hmm, could be. afair the openstack image builder started up a qemu-vm to do stuff and then saved the container contents to an archive, sounds like it can't open the container image for some reason
<lassulus>
I will try to debug that
pbogdan has quit [Remote host closed the connection]
<mdash>
lassulus: on the other hand, this 'debian 10 cloud' image the openstack host already has is 64bit...
<lassulus>
mdash: ah, why do you want to use 32bit then?
pbb has quit [Excess Flood]
<__monty__>
cole-h: But they're not user services and no "borgbackup" service is actually created, there's a service for every job you configure.
kleisli has quit [Remote host closed the connection]
<__monty__>
cole-h: I'm just wondering whether there's a general mechanism or whether I need to patch the nixos module.
<mdash>
lassulus: 'cause I didn't notice and was trying things randomly :)
<cole-h>
I meant that as an example, since I don't use NixOS. I'd work through how the module creates the service and then set Restart= if its possible
<nix-build>
[nixpkgs] @ryneeverett opened pull request #83798 → vimPlugins: Add --proc flag to update script. → https://git.io/Jv7d3
<mdash>
lassulus: it boots fine running qemu myself
<nix-build>
[nixpkgs] @ryneeverett closed pull request #83008 → vim-plugins: Add --debug flag to update.py. → https://git.io/Jv7dG
<mdash>
and all the other images on this host use bios boot, so that's what I set up for
<lassulus>
mdash: ah cool, there were also some issues with the openstack image builder and I never really tried it. I just use the nixpkgs functions to generate the image
<mdash>
lassulus: okay. my hunt for the guilty parties continues!
<mdash>
;-)
pbb has joined #nixos
lsix1 has joined #nixos
<nix-build>
[nixpkgs] @FRidh pushed 0 commits to python-unstable: https://git.io/Jv7dC
<evanjs>
Gah. Trying to use rust bindgen with a shell.nix. got stdint.h but still missing cstddef :/
nschoe has quit [Remote host closed the connection]
kloenk has joined #nixos
pbogdan has joined #nixos
domogled has joined #nixos
mexisme has joined #nixos
<evanjs>
mmkay, think it was just clang-tools that I needed
<evanjs>
or removing the explicit addition of nix_cflags... hrm
noudle has quit []
<nix-build>
[nixpkgs] @cole-h opened pull request #83799 → cantata: make qtmultimedia and vlc optional → https://git.io/Jv7di
tsrt^ has joined #nixos
jlv12 has joined #nixos
reallymemorable has quit [Quit: reallymemorable]
<jlv12>
Does anyone know how to run `xset` as root, without hard coding a particular user to switch to, such as in `services.acpid.lidEventCommands`?
mehlon has quit [Quit: Leaving]
<jlv12>
I'm trying to turn off my laptop display when the lid closes. I got it working by hard coding my primary username, but that isn't an ideal solution.
<chrisaw>
I thought I could append to the existing postStart script for a service using lib.mkAfter -- i.e. system.services.mysql.postStart = lib.mkAfter '' echo "Example" '';
<chrisaw>
What am I doing wrong? :S
<tilpner>
chrisaw: What isn't working?
<chrisaw>
I'm just getting: The option `systemd.services.mysql.serviceConfig.ExecStartPost' has conflicting definitions. I know I could force overwrite it with Force but I just want to append a few commands on to the end of the existing script.
<tilpner>
It depends on who set the current value, and how
<nix-build>
[hydra] @samueldr opened pull request #729 → hydra-eval-jobs: Mirror eval errors in STDERR → https://git.io/Jv7Fk
<tilpner>
Do you have another definition?
<chrisaw>
Nope, it's just pulling the package default definition.
<jlv12>
tilpner: Thanks. I actually found both of those links, but the wiki article wasn't really helpful, and I couldn't get the script working. Your one liner appears to work though.
o1lo01ol1o has joined #nixos
<tilpner>
chrisaw: That's not going to work
<tilpner>
chrisaw: Are those lines independent of the script that's currently running?
<chrisaw>
tilpner: Yes they are - just a bunch of commands I want to run once mysql starts.
<tilpner>
Then let's try to add another ExecStartPost, instead of altering the existing one
<chrisaw>
tilpner: Yeah that would work equally as well - if only I knew how. :)
<dhess>
hi all, if I want to install both Windows 10 and NixOS on the same box, I assume I should install Windows 10 first?
<jlv12>
dhess: If they're installed on different drives. I don't think it matters.
jassob has quit [Quit: WeeChat 2.5-rc1]
<jlv12>
If they're on the same drive, Windows might override the boot sector, so you probably want to install that first.
o1lo01ol1o has quit [Remote host closed the connection]
<dhess>
they'll be on different drives. But will Win 10 install itself nicely into a grub menu, for example?
erictapen has quit [Quit: leaving]
chagra has joined #nixos
<jlv12>
dhess: Hmm, I'm not sure. I always switch OS from bios. I think you could update the grub menu after installing Windows, to add it, but it would be easier to just install Windows first.
erictapen has joined #nixos
cosimone has quit [Quit: Quit.]
<dhess>
Oh yeah no way I'm doing that :)
<dhess>
anyway figured it was easier to install Windows first, thanks for confirming.
<jlv12>
np
SuppenGeist has quit [Remote host closed the connection]
<qy[m]>
i mean as i understand it from reading an old pr, the issue is that nix changed it's method for generating manpages at some point, and home-manager just hasn't fully caught up entirely
<qy[m]>
that's not home-manager's fault, but it is a situation that centers on home-manager
<qy[m]>
hence that comment
<qy[m]>
the blame is entirely on nix, but the problem centers on home-manager
selfsymmetric-mu has quit [Remote host closed the connection]
<gchristensen>
I mean,
<gchristensen>
nixos' modules -> docbook thing hasn't really ever been an API
<gchristensen>
so depending on it is sort of a "keep both pieces" scenario
evanjs has joined #nixos
<chrisaw>
tilpner: I wonder if it's possible to use lib.mkForce ${some_ref_to_original_poststart} ++ ' new_stuff_here '; kind of syntax?
<tilpner>
chrisaw: That's what I'm trying to do, but it's not pretty
jlv12 has quit [Remote host closed the connection]
captjakk has quit [Remote host closed the connection]
* cole-h
wonders if it's a spurious failure to eval master
captjakk has joined #nixos
<gchristensen>
cole-h: is this an ofborg thing?
<cole-h>
Eval is failing on my PR saying that the branch it will merge into fails to cleanly eval
<nix-build>
[nixpkgs] @burke opened pull request #83804 → solargraph: Document the original bundle, not the bundlerApp's: → https://git.io/Jv7Fj
<rycee>
bqv: Home Manager doesn't use the NixOS documentation system so there might be some incompatibilities.
wiml has joined #nixos
<qy[m]>
right, so it's completely incompatible with documentation.includeAllModules
<qy[m]>
that makes sense then
<rycee>
Well, in this case I imagine it is because any references to other options in a HM option description will refer to the options by their name as seen from a standalone HM installation while `includeAllModules` will see the name with a `home-manager.users.<name>` prefix so the generated docbook code will be all wrong.
<rycee>
It may be fixable by the NixOS documentation system keeping track of submodules and giving the option attribute prefix as an argument to the XSLT code that generates the manual, which then could prepend the prefix to all references.
elorm has joined #nixos
<rycee>
I think it wouldn't be simple, though :-/
elorm has joined #nixos
elorm has quit [Changing host]
peelz has joined #nixos
zupo has joined #nixos
elorm has quit [Quit: leaving]
ok2` has quit [Ping timeout: 256 seconds]
<peelz>
What is the recommended way for creating an overlay for a gnome3 package?
<rycee>
The most straight forward would probably be to add support for some kind of filter w.r.t. `includeAllModules` that could be used to exclude the HM module.
<qy[m]>
mm, in my case i was enabling it specifically to add HM modules
zupo_ has joined #nixos
<qy[m]>
but i'll settle for having two separate manpages
<qy[m]>
no other flake import i have is large enough to be significant
zupo has quit [Ping timeout: 260 seconds]
<simukis__>
grahamc[m]: does buildLayeredImage have a way to prioritize layerization and/or is there a way to layerize some higher level closure as a single layer? Say, I want to put a python runtime and all of its dependencies into a single layer and then create multiple layers for the rest of the stuff?
simukis__ is now known as simukis_
<rycee>
I think HM in particular is a bit tricky since it is used both standalone and as a NixOS module but the module documentation is limited to describe one of the use cases.
<jared-w>
simbergm: I don't believe so. I think the closest you can get is building a manual layer and then using buildLayeredImage with that. I've had difficulties mixing buildLayeredImage with other layers and with public docker images in general
<nix-build>
[nixpkgs] @veprbl pushed 2 commits to release-19.09: https://git.io/Jv7N0
<evanjs>
or just getting cstddef into the nix-shell, I guess :P
<evanjs>
With clang in this case, I think?
cosimone has joined #nixos
growpotkin has joined #nixos
xkapastel has quit [Quit: Connection closed for inactivity]
ixxie has quit [Ping timeout: 256 seconds]
<tilpner>
evanjs: I have LIBCLANG_PATH="${llvmPackages.libclang.lib}/lib/libclang.so";
<tilpner>
evanjs: And BINDGEN_EXTRA_CLANG_ARGS = "-I ${linux.dev}/lib/modules/${linux.modDirVersion}/build/include -I ${linux.dev}/lib/modules/${linux.modDirVersion}/build/arch/x86/include/generated"; for something else
<nix-build>
[nixpkgs] @NickHu opened pull request #83806 → zotero: make wrapGAppsHook work → https://git.io/Jv7Nx
lsix has quit [Ping timeout: 272 seconds]
rauno has joined #nixos
endformationage has joined #nixos
<benny>
tilpner: several builds later and yeah, 16GB memory isn't enough and the reason the system crashes is due to this bug: https://github.com/openzfs/zfs/issues/7734
<nix-build>
openzfs/zfs#7734 (by runderwo, 1 year ago, open): Swap deadlock in 0.7.9
<glittershark>
it looks like this is an issue with cairo's ./configure
linarcx has joined #nixos
<aranea>
Is there a builtins.getAttr value that does something more useful than erroring out when the attribute isn't found? Or do I need to write that myself based on getAttr and hasAttr?
<FRidh>
and that would be NixOS onl
<aranea>
s/value/alternative/
<glittershark>
aranea: the ? operator iirc
<simpson>
glittershark: Perhaps a flag of some sort is required in the builder; around l70, maybe, something like `++ optional (X11Support && xcbSupport) "--enable-x11"`. Don't forget that XCB support implies X11 support, I think?
<evanjs>
tilpner: do you still wrap the NIX_CFLAGS in build.rs?
pjt_014 has joined #nixos
linarcx has quit [Client Quit]
<glittershark>
simpson: yep
<glittershark>
will try that and open a PR if it fixes it
worldofpeace_ has joined #nixos
<evanjs>
e.g. `.clang_args(nix_cflags.split(" "))`
<tilpner>
evanjs: No
<aranea>
That's just a static variant of hasAttr, not what I need. Thanks though. :)
<tilpner>
evanjs: But I also have very little context about your problem
linarcx has joined #nixos
reallymemorable has joined #nixos
lsix has joined #nixos
<evanjs>
tilpner: basically, I'm trying to use rust-bindgen (nix-shell w/lorri), and I am able to reference some headers (from a local nix package) just fine, but those files reference cstddef, which cannot be found with my current setup
coco has joined #nixos
lsix has quit [Client Quit]
<tilpner>
evanjs:
lsix1 has quit [Ping timeout: 252 seconds]
<tilpner>
... and did you try BINDGEN_EXTRA_CLANG_ARGS = "-I ${libcxx}/include/c++/v1";?
<evanjs>
tilpner: ahh I did not try putting it into BINDGEN_EXTRA_CLANG_ARGS
<glittershark>
oh dang I think this a sandboxing issue :cry:
<tilpner>
Huh, where else did you try putting it?
<glittershark>
the cairo configure script is just finding my system-wide xlib
<glittershark>
and enabling it since it's there
<evanjs>
I did see it there and tried adding it to buildInputs earlier. I wasn't sure how to like -I it for bindgen
<glittershark>
goddammit, darwin
<evanjs>
that makes sense, though
<glittershark>
dang, configure with --disable-x11 still fails
<nix-build>
[nixpkgs] @Ericson2314 pushed 8 commits to release-20.03: https://git.io/Jv7xT
<evanjs>
tilpner: I get a whole slew of errors, now :D e.g. /nix/store/rpinhskd2g4a9cz620w2b2frm7hrxw7j-libc++-7.1.0/include/c++/v1/__nullptr:55:1: error: unknown type name 'namespace'
<glittershark>
wheeee that worked
<glittershark>
thanks for your help simpson
chagra has quit [Ping timeout: 265 seconds]
<simpson>
glittershark++ you did all the work.
<nix-build>
glittershark's karma got increased to 1
bdju has quit [Read error: Connection reset by peer]
<betaboon>
i just noticed that there are packages that create entries in $out/share/dbus-1/services which allows them to be automatically started by dbus. i dont like that ...
<nix-build>
[nixpkgs] @AndersonTorres pushed 2 commits to master: https://git.io/Jv7x2
fragamus has joined #nixos
andreas303 has quit [Remote host closed the connection]
webdavis has quit [Quit: Idle for 30+ days]
<nix-build>
[nixpkgs] @glittershark opened pull request #83808 → cairo: pass --disable-xlib to config if no x11 → https://git.io/Jv7x6
ngerstle[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
OmnipotentEntity has joined #nixos
linarcx has quit [Quit: WeeChat 2.8]
siiky has quit [Quit: leaving]
andreas303 has joined #nixos
linarcx has joined #nixos
wakest[m] has quit [Quit: Idle for 30+ days]
Darkmatter66_ has joined #nixos
<pie_[bnc]>
tbenst[m]: I remembered the packaging prolem with hanging tests :( It doesnt tell you which test is currently running...
<pie_[bnc]>
and i dont see it in the ^C result yet
erictapen has quit [Ping timeout: 258 seconds]
aveltras has joined #nixos
Darkmatter66 has quit [Ping timeout: 246 seconds]
gsals[m] has quit [Quit: Idle for 30+ days]
killercup[m] has quit [Quit: Idle for 30+ days]
erictapen has joined #nixos
<pie_[bnc]>
but pytest-timeout is crap and doesnt tell you where it killed itself
<aveltras>
i have a systemd service like the following. The server doesn't restart and isn't stopped when running "systemctl stop server", the journalctl entry indicates that it tries to restart a second instance before stopping the running one (i think cause it complains about an already in use port). Is there anything incorrect ?
<nix-build>
[nixpkgs] @teto opened pull request #83809 → vimPlugins: make it more extensible → https://git.io/Jv7xj
<nix-build>
swaywm/sway#4091 (by grahamc, 48 weeks ago, closed): `exec sway` is security-critical, but underdocumented
<syd>
sudo nixos-rebuild switch 22.07s user 6.86s system 105% cpu 27.353 total
<syd>
for a no-op nixos-rebuild switch.
<bdju>
ah thanks for the tip
<glittershark>
bdju: I have `VGA compatible controller: Intel Corporation UHD Graphics 630 (Mobile)` on this computer running 5.4.28 and it seems to work fine
<glittershark>
<glittershark>
bumblebee/optimus is completely broken but that's another story
<cransom>
syd: thats basically how it will be. it still has to go through and evaulate everything to be sure nothing changed.
<syd>
cransom Right, but that's a _lot_ to evaluate.
<bdju>
glittershark: to clarify I never had problems on sandy bridge igpu, but have on haswell, it's not all intel stuff.
<cole-h>
Means you have a lot in your configs that needs to be evaluated.
<syd>
It doesn't help that I'm using like 4 different versions of nixpkgs pinnings because of how broken the haskell ecosystem is
worldofpeace_ has quit [Quit: worldofpeace_]
<gchristensen>
syd: yeah, each one of those makes everything extremely worse, that thorouhly explains it :P
<syd>
Would it help if they all shared the same nixos pinning, or does it still need to evaluate them all seperately?
<glittershark>
bdju: this is an 8th generation (coffee lake)
<bdju>
getting build failure on wofi... was hoping to update before rebooting, still in the tty after the crash.
<bdju>
glittershark: ah okay if it's new then that makes me feel better
<glittershark>
yep
reallymemorable has quit [Quit: reallymemorable]
pantomath[m] has quit [Quit: Idle for 30+ days]
<OmnipotentEntity>
If I have an application that requires SSE, what's the best way to annotate that in a nixpkgs contribution? I assuming that it's best to restrict platforms, but which platforms ought I restrict to (assuming it's a linux only style contribution), is x86_64-linux both necessary and sufficient?
<gchristensen>
should be, and if it isn't sufficient, a bug report will happen and we can deal :)
<tbenst[m]>
pie_[bnc]: ha! I forgot about that. not in nixpkgs, but I do have that in my nix shells
<pie_[bnc]>
I dont know if its only needed for the tests or should be propagated to downstream packages though
morgib[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
nrdxp[m] has quit [Quit: Idle for 30+ days]
bobvanderlinden_ has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
xavierzwirtz[m] has quit [Quit: Idle for 30+ days]
index-0[m] has quit [Quit: Idle for 30+ days]
srid[m] has quit [Quit: Idle for 30+ days]
truby has quit [Quit: Idle for 30+ days]
Ovsyanka[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
jamiemagee has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
risebell[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
isgy[m] has quit [Quit: Idle for 30+ days]
gcoakes[m] has quit [Quit: Idle for 30+ days]
gallible[m] has quit [Quit: Idle for 30+ days]
xavierm02 has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
pwaller[m] has quit [Quit: Idle for 30+ days]
dreadward[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
iceychris[m] has quit [Quit: Idle for 30+ days]
tek[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
val2229[m] has quit [Quit: Idle for 30+ days]
fstamour[m] has quit [Quit: Idle for 30+ days]
gotofail[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
mupf[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
TimePath has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
l_inus[m] has quit [Quit: Idle for 30+ days]
Sisyphe[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
neonfuz2 has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
joelwallis1[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
HailSagan[m] has quit [Quit: Idle for 30+ days]
TroelsAgergaardJ has quit [Quit: Idle for 30+ days]
sterninator3000[ has quit [Quit: Idle for 30+ days]
dot-product[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
tobia_[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
sehqlr[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
jbbr[m] has quit [Quit: Idle for 30+ days]
sreybastien[m] has quit [Quit: Idle for 30+ days]
Smith[m] has quit [Quit: Idle for 30+ days]
akapaka[m] has quit [Quit: Idle for 30+ days]
DenisLehmann[m] has quit [Quit: Idle for 30+ days]
zml[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
nbathum[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
WillW has quit [Quit: Idle for 30+ days]
gilescope[m] has quit [Quit: Idle for 30+ days]
kraem[m] has quit [Quit: Idle for 30+ days]
organixpear has quit [Quit: leaving]
<clever>
gchristensen: and we got matrixed again, lol
<gchristensen>
heh
sMuNiX has joined #nixos
<gchristensen>
it'd be nice if matrix spread that out a bit :)
<drakonis>
wow that's nice
<drakonis>
it actually axes afk users
<drakonis>
this is the best thing, as there was an irc channel i chatted on that was pretty much 80% matrix users
<bdju>
Unusably broken after updote and reboot. Wonky monitors. Briefly fixed after redocking. They went from DP-3/4 to 5 and 6. Got a crash to TTY (not logged in this time because I used exec) attaching to remote TTY. Back to old kernel I guess.
<drakonis>
700 users and a dozen werent matrix users nor were lurkers
tokudan_ has quit [Quit: Dunno.]
<pie_[bnc]>
tbenst[m]: on a sidenote then, does wrapgappshook not work in an fhsuserenv? or does it not deal with hooks because its not a genericBuild derivative?
tokudan has joined #nixos
<pie_[bnc]>
(i dont know much about fhsuserenv
<pie_[bnc]>
clever: i heard soemthing about ovh going down earlier
<pie_[bnc]>
oh
<pie_[bnc]>
just noticed the idle thing
<iceypoi>
is setting `boot.binfmt.emulatedSystems` all that I have to do to be able to cross-compile for the raspi?
<tbenst[m]>
pie_[bnc]: no, I could never get it working with fhs. Honestly I don’t know much about what is happening behind the scenes with the builders
<bdju>
all fine on old kernel so guess it was the kernel. I did updates before changing kernels so that it'd be the only difference if I had to roll back a gen
<glittershark>
heck yeah nixos tbh
linarcx has quit [Quit: WeeChat 2.8]
<bdju>
yeah
raingloom has quit [Ping timeout: 258 seconds]
<bdju>
if I were on arch I'd probably just cry, never had to roll back a kernel on a normal distro haha
<pie_[bnc]>
tbenst[m]: its black magic till you start looking
<MichaelRaskin>
Of all things, kernel is what is reasonable to rollback wherever
<pie_[bnc]>
then its slightly less black magic and a lot of bash :P
<bdju>
yikes... firefox crashed trying to save an image... happened a couple time in the last month but hadn't had problems in a few days
rui has quit [Quit: rui]
raingloom has joined #nixos
hio has quit [Quit: Connection closed for inactivity]
<nix-build>
[nixpkgs] @jonringer opened pull request #83813 → python3Packages.subunit: fix build and add tests → https://git.io/Jv7j4
<nix-build>
[nixpkgs] @jonringer pushed commit from @r-ryantm to master « stacks: 2.41 -> 2.52 »: https://git.io/Jv5fC
reallymemorable has quit [Quit: reallymemorable]
philip_qumulo has joined #nixos
<NemesisD>
how would you all suggest installing a global npm dependency with nix? i'm trying `yarn2nix.mkYarnPackage` but it sort of seems like that's for projects. i don't have anything to provide to `src`
o1lo01ol1o has quit [Remote host closed the connection]
mexisme has quit [Ping timeout: 272 seconds]
snicket has quit [Ping timeout: 240 seconds]
reallymemorable has joined #nixos
philip_qumulo has quit [Client Quit]
thongpv871 has joined #nixos
erictapen has quit [Ping timeout: 256 seconds]
thongpv87 has quit [Ping timeout: 260 seconds]
justanotheruser has joined #nixos
vandenoever has quit [Ping timeout: 264 seconds]
<nix-build>
[nixpkgs] @orivej-nixos pushed commit from @orivej to master « qpaeq: fix build with desktop file validation after #75729 »: https://git.io/Jv5fX
<nix-build>
[nixpkgs] @jonringer pushed commit from @r-ryantm to master « leatherman: 1.10.0 -> 1.11.0 »: https://git.io/Jv5fH
bvdw has quit [Read error: Connection reset by peer]
xelxebar_ has quit [Remote host closed the connection]
bvdw has joined #nixos
<bsima>
is the nixos configuration linked somewhere in /run/current-system?
<bsima>
i can't find it
<gchristensen>
no
<MichaelRaskin>
And it is not easy to do correctly…
<gchristensen>
NixOS doesn't capture the configuration of the system which was built. there is an option which attempts to do this, but doesn't do a great job (I can't even find the name)
xelxebar has joined #nixos
<MichaelRaskin>
Maybe it was so buggy it was dropped?
philip_qumulo has joined #nixos
thongpv87 has joined #nixos
mbrgm_ has joined #nixos
mbrgm has quit [Ping timeout: 272 seconds]
mbrgm_ is now known as mbrgm
<gchristensen>
I hope so :P
glittershark has quit [Read error: Connection reset by peer]
<bsima>
so how can i get the current config?
<philip_qumulo>
Hey, I'm running 20.09pre218613.ae6bdcc5358 (Nightingale) -- a quite recent build on unstable -- and I'm trying to run this command: `machinectl pull-tar https://nspawn.org/storage/ubuntu/bionic/tar/image.tar.xz`. When I do, I eventually get this error: Failed to rename to final image name to
<philip_qumulo>
/var/lib/machines/.tar-https:\x2f\x2fnspawn\x2eorg\x2fstorage\x2fubuntu\x2fbionic\x2ftar\x2fimage\x2etar\x2exz.\x22q3ndaffkja0\x22: Read-only file system
<bsima>
(i'm using switch-to-configureation, not nixos-switch)
thongpv871 has quit [Ping timeout: 256 seconds]
<MichaelRaskin>
Well, you cannot
<philip_qumulo>
I can see that /var/lib/machines isn't read-only. How can I get more information about what's going on?
Guest99424 has joined #nixos
<Guest99424>
Hello, I don't have sound on my nixos system. What'd the recommende troubleshooting be?
thc202 has quit [Ping timeout: 252 seconds]
<gchristensen>
did you set sound.enable = true; ? this'd be a good start