<infinisil>
Are nix-env -i's atomic? Or can one override the others change if they happen at the same time?
<gchristensen>
in the ACID model they're Atomic but don't implement Isolation in the
<gchristensen>
s/ in the//
<infinisil>
Hmm darn
<gchristensen>
I'm fairly sure of that, anyway
<infinisil>
Won't matter that much I think for what I want to do though
<infinisil>
I wanna create an `environment.autoUpdated.packages` option :)
<infinisil>
E.g. `environment.autoUpdated.packages.youtube-dl = { url = "channel:nixpkgs-unstable"; attrPath = [ "youtube-dl" ]; period = "hourly"; }`
Jd007 has quit [Quit: Jd007]
<gchristensen>
sounds good
<infinisil>
Gets a bit tricky though, because I can't just nix-env -r, because then you can't control periodicity of individual packages (if you install all of them into a single profile)
<infinisil>
Eh, I'll just use multiple profiles for a start
MasseR has quit [Quit: Ping timeout (120 seconds)]
cartwright has quit [Remote host closed the connection]
oida has joined #nixos
<andoriyu>
I don't see it's being used in any of the included Makefiles weird
cartwright has joined #nixos
mvnetbiz_28 has joined #nixos
<clever>
andoriyu: double-check if you can cross compile the v4l loopback i linked earlier?
<clever>
maybe all out-of-tree cross compile is busted?
<andoriyu>
clever yup, that's the case
ahmed_elgabri has joined #nixos
<CMCDragonkai>
should something like electron-forge/cli be packaged up in nix?
mvnetbiz_2 has quit [Ping timeout: 245 seconds]
mvnetbiz_28 is now known as mvnetbiz_2
<CMCDragonkai>
if it is, does that mean the packages that it builds will only work on NixOS and not work on other linux distros?
<CMCDragonkai>
and if so, its debian and rpm builds become useless?
<clever>
andoriyu: then the problem is much deeper, you would have to see how the kernel cross-compile works, and then apply that fix to every module package
<CMCDragonkai>
electron-forge/cli uses electron-packager, and I'm not sure if Nix's electron contains that
<andoriyu>
clever aside from it taking forever, can I use armv6 device to build that module and push it to my binary cache?
redmp has joined #nixos
<andoriyu>
clever that's way above my understanding of nix and kernel module building.
<clever>
andoriyu: nix doesnt allow mixing native and cross compiled things
bitmapper has joined #nixos
<andoriyu>
so everything would have to be natively built on rpi?
jkachmar has quit [Quit: Connection closed for inactivity]
<andoriyu>
I mean right now my nix store is full of amd64 and armv6l things
<clever>
andoriyu: you can get ugly, and do boot.kernelPackages = (import <nixpkgs> { system = "armv6l-linux"; }).linuxPackages; for example
<clever>
then only the kernel&modules are native, but the rest remains cross
<andoriyu>
i see
<clever>
but going any finer then that gets real messy
<clever>
angerman: and this will have logs from its attempts to ssh into each one
<angerman>
clever: wait what, isn't that set by nix.machines?
<clever>
angerman: nix.machines is what generates the contents of /etc/nix/machines
<angerman>
ahgh that's annoying :L
<clever>
angerman: but hydra is more flexible then nix itself, and can take a : seperated list of those files, allowing auto-scaling scripts to create&destroy build machines in aws
<clever>
so you can have a 2nd file, that isnt managed by nix, in that env var
<angerman>
services.hydra.buildMachinesFiles defaults to []. great ;D
<clever>
it doesnt default to [], there is a trick going on
<angerman>
clever: it also means we can probably try to hack hydra and improve IFDs or something, but that will largely depend on how much time we have for it.
<angerman>
For no we'll mostly be using to do build-the-world kinda ghcjs builds.
ahmed_elgabri has joined #nixos
<clever>
angerman: i do have some ideas on how to make hydra profile the eval for you...
<angerman>
I still think hercules is pretty good in that it splits the eval and hands you the non-ifd parts, and the ifd parts lazily ... kinda.
<clever>
angerman: then you can generate graphs like this, that show how long a given attr took to eval (not counting the shared state with other attrs)
<karantan>
what exactly does it get passed to boot.nix when I do "impots = [ ./boot.nix ]"
<DigitalKiwi>
always love finding new keyboard shortcuts i didn't know existed
<karantan>
I know if I remove those three dots in the boot.nix it won't evaluate
<DigitalKiwi>
especially when it's a shortcut for "halt without confirmation"
Aleksejs_Home is now known as Aleksejs
Hurttila has joined #nixos
maier has joined #nixos
maier has quit [Client Quit]
<DigitalKiwi>
:(
<clever>
karantan: all nixos modules get passed the same args, the usual ones are config, pkgs, lib, and i think options, but others can be dynamically added
maier has joined #nixos
<karantan>
aha. so this is by default and those args are needed by nixos modules?
<clever>
yeah
<karantan>
ql thanks
waleee-cl has quit [Quit: Connection closed for inactivity]
bitmapper has quit [Quit: Connection closed for inactivity]
attila_lendvai has joined #nixos
<i1l>
DigitalKiwi: alt + up is the best (tty, rescue mode)
turlando has quit [Ping timeout: 246 seconds]
sangoma has quit [Quit: WeeChat 3.0]
bahamas has joined #nixos
<DigitalKiwi>
i1l: :)
ahmed_elgabri has joined #nixos
<karantan>
how do I know if I'm importing a module or a function? or are they basically the same? Because it looks to me that module is a function with a specific structure (usually)
<maier>
Are there utility functions like "Find all files of a specific type recursively in a directory" built-in anywhere?
dbmikus has joined #nixos
<karantan>
aha so if we import a .nix file with "foo = import ./foo.nix {a=2;b=3;};" this is a function call but if we import foo with imports "imports = [./foo.nix]" we pass those pkgs, lib, options ... args because we import it as a module
<maier>
karantan: Looks correct to me, though I'm still trying to make sense of the exact same stuff too :) The first one imports the nix expression in ./foo.nix, which seems to be a function taking a set as (first) argument, and applies it to the given set AFAIU
jess has joined #nixos
dbmikus has quit [Ping timeout: 264 seconds]
ahmed_elgabri has quit [Ping timeout: 260 seconds]
<adisbladis>
karantan: `import ./foo.nix {}` is a direct function call while `{ imports = [ ./foo.nix ]; }` is managed by the NixOS module system, and the module system knows to apply the correct arguments
o1lo01ol_ has joined #nixos
<karantan>
thanks
m0rphism has joined #nixos
avaq has quit [Ping timeout: 246 seconds]
o1lo01ol1o has quit [Ping timeout: 260 seconds]
hugolgst has quit [Quit: Lost terminal]
domogled1 has joined #nixos
ahmed_elgabri has joined #nixos
simba1 has quit [Ping timeout: 258 seconds]
domogled has quit [Ping timeout: 256 seconds]
domogled1 is now known as domogled
turlando has quit [Ping timeout: 260 seconds]
Crocodil1 is now known as Crocodillian
ericsagnes has quit [Ping timeout: 258 seconds]
simba1 has joined #nixos
<maier>
adisbladis: Are these the ones in builtins, or are there more I didn't find yet? :)
<ashkitten>
is there any way to *remove* attrs from an attrset via lib.recursiveUpdate?
jimkooch has quit [Client Quit]
simba1 has quit [Ping timeout: 264 seconds]
simba1 has joined #nixos
<clever>
ashkitten: either set it to some context-specific no-op value ({} or null or []), or use lib.filterAttrs to remove it afterwards, or use lib.filterAttrs/filterAttrsRecursive to remove every attr with a magic value (such as null)
Rainy has joined #nixos
<Rainy>
hey, I am trying to overide a setting in a nix package with the following line "services.udisks2.enable = pkgs.lib.mkOveride 0 false;" but its causing the rebuild to throw this up "error: attribute 'mkOveride' missing, at <file>" How do I declare the mkOveride option. I have tried import lib and referencing it there and got the pkg.lib.mkOveride thing from a thread on the forum (dangerous)
<Rainy>
btw trying to disable it to remove deadlock with the amazon EC2 settings and I have no need for automount usb drives
<Rainy>
ok I think I fixed it by using mkForce instead
aforemny has joined #nixos
domogled1 has joined #nixos
domogled has quit [Ping timeout: 265 seconds]
domogled1 is now known as domogled
<ixxie>
I've got a SD card reader on my Dell XPS 13 (9350), but NixOS doesn't seem to mount any of the two SD cards I have available to try out in it
<ixxie>
lspci seems to show the device but I don't know much else in the way of debugging strategies here
meh` has joined #nixos
<Rainy>
I have issues with other laptops where the SD reader presented itself in a way that made it unusable in Linux, have you used it successfully in any other distros?
<ixxie>
nope, I only ever used NixOS on this laptop
<ixxie>
but I do believe I have previously gotten the SD card to work
<cinimod>
`nix-shell` tells me `error: file 'nixpkgs' was not found` but if I do `nix-channel --list` I can see `nixpkgs https://nixos.org/channels/nixos-20.09`
<lush>
so I'm trying to install VSCode with the latex-workshop extension installed. so I added a "vscode.nix" file with this content: https://susepaste.org/95943154 and tried importing it into my configuration.nix. nixos-rebuild switch fails stating that "LET" was unexpected. what am I doing wrong?
domogled has quit [Remote host closed the connection]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
_urd[m] has joined #nixos
<DavHau[m]>
When building remote, can I tell nix somehow to just build and not download the results? I'm not interested in the results and have low bandwidth. The results should just remain on the builder which also acts as binary cache for other hosts
<aforemny>
How can I query the value of the argument `system` in my NixOS configuration? Ie. if I pass `--argstr system aarch64-linux`, how can I access that value? `config.system` appears to be something different.
novafacing has quit [Quit: novafacing]
<clever>
aforemny: pkgs.system
novafacing has joined #nixos
<aforemny>
clever++ thanks!
<{^_^}>
clever's karma got increased to like 562, I think
davidv7_ has quit [Quit: Konversation terminated!]
<cinimod>
With unstable I get ./devX11.h:64:12: fatal error: 'cairo-xlib.h' file not found
saschagrunert has quit [Remote host closed the connection]
andreas303 has quit [Remote host closed the connection]
ronthecookie has quit [Excess Flood]
ronthecookie has joined #nixos
simba1 has quit [Ping timeout: 258 seconds]
cransom has quit [Quit: WeeChat 2.9]
cransom has joined #nixos
Hurttila has joined #nixos
mkaito has joined #nixos
mkaito has quit [Changing host]
mkaito has joined #nixos
Siyo has quit [Ping timeout: 256 seconds]
<asymmetric>
how do i get a sha256 hash in SRI format when contributing to nixpkgs?
ahmed_elgabri has quit [Ping timeout: 246 seconds]
Siyo has joined #nixos
<asymmetric>
lib.fakeSha256 and run a build?
<asymmetric>
it seems like nix-prefetch-git returns the other format
simba1 has joined #nixos
<sterni>
asymmetric: sri hashes are only implemented in the experimental cli interface planned for nix 3 (I think)
<clever>
$ nix hash --help
<clever>
to-sri convert a hash to SRI representation
<sterni>
depending on which nix version you are using you can use nix to-sri or nix hash to-sri
<sterni>
also nix hash-path --sri and nix hash-file --sri work, but in nixUnstable they are called nix hash path and nix hash file and you might have to enable experimental commands
<sterni>
as said, experimental interface
<clever>
there is also the old `nix-hash`, but it lacks the sri step
<xwvvvvwx>
How can I write a file that lives under /usr with nix? I would like it to work something like the environment.etc module where a symlink into the store is added to the filesystem under /usr/lib/systemd/sleep.
<clever>
xwvvvvwx: nothing in linux will look in /usr/lib, so that file shouldnt have any effect
<clever>
nixos*
<xwvvvvwx>
hmmmm ok, so where should I put scripts that will be executed by systemd-suspend?
<clever>
xwvvvvwx: i'm guessing that systemd may have been patched by nixpkgs, to look in a different place, the nix expr would be the next place to look
<LilleCarl>
hexa-: I've just looked at status.nixos.org. Haven't the sligtest idea about hydra and the likes yet. Only reason I'm noticing is because LXD is broken
<dutchie>
i see that it did build successfully 16 hours ago though
malook has joined #nixos
<hexa->
fwiw: this builds for me locally :D
<dutchie>
the channel hasn't updated to that yet(?) though
<dutchie>
not sure if the more recent failed build will prevent that
karantan has joined #nixos
<hexa->
yeah, not sure what is up with that
<LilleCarl>
Unclear indeed, i haven't the slightest about how the bump actually works. It seems like everything NixOS except releases is 100% automated
tomberek has joined #nixos
Hurttila has quit [Ping timeout: 240 seconds]
<LilleCarl>
hexa-: Still haven't packaged FRR, I'm such a lazy POS :d
<hexa->
LilleCarl: smh
wolfshappen has quit [Ping timeout: 240 seconds]
<LilleCarl>
We've actually been thinking about it at work. Trialing NixOS as BGP edge. We only have 10Gb transit, which a potato can route safely these days. Specially with the pace the routing table is growing i wouldn't wanna go with the spectrum3 that can have 1M installed routes. That'll be toast sooner than later. And we're cheap :P
<rdk31>
I'm trying to install nixops using the unstable.nixpkgs channel on archlinux, failing on this: builder for '/nix/store/fcyrn4rwagvpmigg6yn1rjmlpcw518v5-python2.7-requests-2.25.1.drv' failed with exit code 1. I know that nixops 2.0 should be coming soon, that's why all I'm asking for is a workaround. Has anyone figured it out?
<exarkun>
is the terraform-provider-nixos packages w/ terraform 0.12 broken?
respawn_ has quit [Quit: Leaving]
<exarkun>
Error: Failed to instantiate provider "nixos" to obtain schema: Incompatible API version with plugin. Plugin version: 4, Client versions: [5]
rajivr has quit [Quit: Connection closed for inactivity]
<asymmetric>
how do i make `nix build' output the `echo' lines in setup hooks, for debug purposes?
<tpw_rules>
if you pass the -L flag it will output everything
waleee-cl has joined #nixos
<asymmetric>
thx!
sangoma has joined #nixos
<tpw_rules>
veleiro: i guess you are doing the equivalent of let blah = ./myconfig in mkDerivation { ... postInstall = "cp ${blah} $out/config.config" } ;
<winsome[m]>
I'm trying to flash the qmk firmware on a ZSA keyboard with wally-cli, but it's not working. I found this wiki article https://nixos.wiki/wiki/ZSA_Keyboards, but hardware.keyboards.zsa.enable doesn't seem to actually exist.
<winsome[m]>
I've tried udev.extraRules, but that doesn't seem to work either
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<winsome[m]>
does anybody else have this working on 20.09?
sheepfleece has quit [Quit: Lost terminal]
saschagrunert has joined #nixos
ahmed_elgabri has quit [Ping timeout: 265 seconds]
<clever>
winsome[m]: i see hardware.keyboards.zsa.enable in nixpkgs master
<clever>
winsome[m]: git logs say it was added on june of 2020, so i would expect to find it in 20.09...
<winsome[m]>
hmm, I'm getting `The option `hardware.keyboards' defined in `/etc/nixos/configuration.nix' does not exist.`
<clever>
+ options.hardware.keyboard.zsa = {
<clever>
without the s
<winsome[m]>
That's what I had first, but I saw your message and thought to try it with the 's'.
<winsome[m]>
`error: The option `hardware.keyboard' defined in `/etc/nixos/configuration.nix' does not exist.`
<clever>
i just copy/pasted your own msg, and didnt notice the typo
<clever>
what does `nix eval -f '<nixpkgs>' lib.version` report?
<clever>
cat: nixos/modules/hardware/keyboard/zsa.nix: No such file or directory
<clever>
[clever@amd-nixos:~/apps/rpi/nixpkgs-test]$ git show 88f00e7e12d
<clever>
Date: Mon Jan 18 09:16:22 2021 +0000
saschagrunert has quit [Remote host closed the connection]
<clever>
winsome[m]: stangely, that commit is newer then where zsa got added, but is lacking the file, try switching to nixos-unstable to est that out?
<mayhew>
Does nix-channel --update download an entire new copy of `nixpkgs`?
<clever>
mayhew: yes
<clever>
mayhew: based on the URL's in `nix-channel --list`
<mayhew>
Interesting! So if a server was running that every day, the harddrive would fill up fairly quickly. Is there some documented strategies on how to reduce the space of `nixpkgs`?
<clever>
mayhew: `nix-store --optimize` will hardlink all of the duplicate files together, so they dont consume as much space, and the normal flags in `man nix-collect-garbage` can delete old generations, allowing gc to clear it up fully
rj has quit [Client Quit]
<mayhew>
clever: thanks, I will read the docs on both of those. Would copy-on-write filesystems like btrfs/zfs dedup these anyways (I've never used them)?
domogled has quit [Read error: Connection reset by peer]
domogled has joined #nixos
<clever>
mayhew: zfs would only do it if you turned the dedup flag on before you did all of those --update's
<Ke>
dedup is expensive, no filesystem I know of does it by default
<Ke>
normal use cases do not make it super useful anyway
<clever>
yeah
<clever>
`nix-store --optimize` is a far cheaper file-level dedup
<clever>
zfs dedup is block level, and complicated by compression and encryption
<Ke>
if I have package.nix what is the way to get interactive nix-shell or whatever based on that
<clever>
Ke: `nix-shell package.nix`
<clever>
or if it needs callPackage, `nix-shell -E 'with import <nixpkgs> {}; callPackage ./package.nix {}`
<Ke>
thanks, latter
<Ke>
first one was what I tried
<clever>
both, will give you an env suitable for building the package
<clever>
if you instead want nix to build it for you, and to just run it directly in the shell, you want:
<clever>
ixxie: its currently chewing up 39gig on my nas, and lets me re-fetch things at full gigabit speeds
<Ke>
pinning nixpkgs version is horrible
<Ke>
if something is not maintained, it's should not be used
<clever>
Ke: id say pin, but also test the latest regularly, and keep the pin moving
<clever>
ixxie: cachecache also includes prometheus exporters, which report that i have had a total of 5.3gig worth of misses, and 4.5gig worth of hits in the last week
nicoo has quit [Remote host closed the connection]
<Ke>
anyway, my usecase is not to distribute an environment but to check, what sources I get for a build, so I can create a patch easily
<clever>
ixxie: and on a 6 month scale, its almost 50/50, so its halving my download requirements
berberman_ has joined #nixos
<mayhew>
it seems like pinning is good for dev environments, but terrible for set-and-forget servers where you want to get glibc/openssl/etc updates
<ixxie>
clever: perfect... this is exactly what I need. I moved to a remove location and my download speed has plumetted...
<clever>
ixxie: the only real problems it has right now, is no config (upstream is always iohk + nixos), and no gc policy (i just randomly rm -rf the whole cache dir, whenever i want more space)
nicoo has joined #nixos
<ixxie>
what is iohk?
<cole-h>
a company that uses Nix -- input-output-hk
<clever>
ixxie: hydra.iohk.io's cache, but if you dont have the public key in your end machine, then that wont really impact things
<asymmetric>
and why it mentions the wrong package version
vidbina_ has joined #nixos
<asymmetric>
it's as if overrideAttrs isn't applying
zupo has joined #nixos
<gchristensen>
manveru[m]: not that specific revision, I'll help cole-h run it through a test
<manveru[m]>
ok, i'm not gonna test it until next week i fear
zupo_ has joined #nixos
<manveru[m]>
but changes look good, so take my blessing for now :)
<manveru[m]>
i really like that new review bot
<cole-h>
Great, thank you :) (Maybe add a comment to the effect of "Diff LGTM", if you wouldn't mind?)
simba1 has joined #nixos
<Ke>
clever++ anyway for the line
<{^_^}>
clever's karma got increased to 565
<Ke>
eacameron, clever: if you want to view a nixpkgs package that does not build that gets built with callpackage normally, what do you do?
<asymmetric>
nix-build -L has this output line
<asymmetric>
polkadot> Validating consistency between /tmp/nix-build-polkadot-0.8.29.drv-0/source//Cargo.lock and /tmp/nix-build-polkadot-0.8.29.drv-0/polkadot-0.8.28-1-vendor.tar.gz/Cargo.lock
<Ke>
view as in interactively inspect filesystem state
<asymmetric>
which shows that the vendor tarball is using a different version than the package itself
zupo has quit [Ping timeout: 276 seconds]
dominikh has quit [Remote host closed the connection]
dominikh has joined #nixos
sangoma has quit [Ping timeout: 245 seconds]
orivej has quit [Ping timeout: 246 seconds]
meh` has quit [Ping timeout: 256 seconds]
<cole-h>
manveru[m]++ Thank you
<{^_^}>
manveru[m]'s karma got increased to 1
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/JqU2m
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage2nix: update list of broken builds »: https://git.io/JqU2l
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/JqU2B
<{^_^}>
[nixpkgs] @peti pushed 18 commits to haskell-updates: https://git.io/JqUa2
<drozdziak1>
For remote builders, is there an alternative to adding `command=<some-nix-wrapper <key type> <pubkey> <name@host>` in front of my key in ~/.ssh/authorized-keys?
domogled has quit [Ping timeout: 276 seconds]
domogled1 is now known as domogled
<{^_^}>
[nixos-homepage] @garbas pushed 74 commits to blog: https://git.io/JqUai
<drozdziak1>
I'm on a machine that has systemwide ssh conn reuse
<drozdziak1>
meaning it doesn't let me override the file used for connecting
<{^_^}>
[nixpkgs] @peti pushed 9 commits to haskell-updates: https://git.io/JqUaX
<drozdziak1>
and I'd prefer not to kill all sessions just to unstick the persisted connection
<drozdziak1>
(the file used for connecting = the key used for connecting)
selfsymmetric-mu has left #nixos ["gone to the land of dead hiccups and extinguished light bulbs"]
simba1 has quit [Ping timeout: 260 seconds]
rj has joined #nixos
vidbina_ has joined #nixos
simba1 has joined #nixos
zupo has joined #nixos
zupo_ has quit [Ping timeout: 264 seconds]
EmoSpice has joined #nixos
EmoSpice has left #nixos [#nixos]
<{^_^}>
[nixos-homepage] @garbas pushed to blog « Rename the svg's in blog section »: https://git.io/JqUVU
Hurttila has joined #nixos
<{^_^}>
[nixos-homepage] @garbas pushed 77 commits to commercial-page: https://git.io/JqUVt
<LilleCarl>
hexa-: unstable updated now, yay
maier has quit [Ping timeout: 276 seconds]
nf has quit [Quit: Fairfarren.]
nf has joined #nixos
dsg has quit [Ping timeout: 245 seconds]
dsg has joined #nixos
zakame has quit [Ping timeout: 245 seconds]
zakame has joined #nixos
sangoma has quit [Read error: Connection reset by peer]
<kakakakak>
Hi folks, I am on nixos stable 20 09 and I see that update builds coreutils,openssl,curl, etc, but nothing else. Is that a graft or something?
ixxie has quit [Ping timeout: 260 seconds]
ixxie has joined #nixos
vidbina_ has joined #nixos
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « haskell-streamly: drop obsolete patches »: https://git.io/JqUro
apache8080 has quit [Ping timeout: 245 seconds]
<infinisil>
kakakakak: What do you mean by it building these packages but nothing else?
<kakakakak>
Maybe Stable dose not provide binaries for each update anymore? Sorry I am not up to date with Nix anymore. Did I hit a recent Stable bump and my issue is just by accident?
stree has joined #nixos
<kakakakak>
infinisil I did usual stuff. sudo nix-channel --update, sudo nixos-rebuild boot
<kakakakak>
I see my system builds curl and its deps, but nothing much besides that.
<kakakakak>
So I am surprised. That seems unusual.
<kakakakak>
I assume you adapted some grafts like Guix and that is a openssl/curl graft. That would make sense.
apache8080 has joined #nixos
lsix has quit [Ping timeout: 240 seconds]
<supersandro2000>
kakakakak: we don't backport that much
<supersandro2000>
if you are on unstable you will get regular full rebuilds
<kakakakak>
Not sure how is that related. I have machine with unstable and even master. I am curious about behaviour on Stable. It does not work like in the past. In the past Stable had all binaries ready. Did you change that policy?
<kakakakak>
Curl/ssl changes is a mass rebuild, it does not make sense those binaries are not ready on Stable.
<kakakakak>
Any Stable users here?
<redmp>
are flakes official yet?
henri has joined #nixos
<henri>
kakakakak: No, there was no policy change. It's your overlays that are causing rebuilds.
<kakakakak>
It's a plain box with no overrides. Could you confirm hash for curl on latest stable?
<henri>
redmp: No, there is not even an RFC. (There was one but it was closed)
<kakakakak>
And why building only curl and nothing that depends on it?
<henri>
redmp: It was closed because it was decided that the feature needs more work before being proposed.
rj has quit [Ping timeout: 268 seconds]
<henri>
redmp: Flakes are alive and well but it's going to take a lot more time to land in stable Nix.
<henri>
redmp: However, eelco has started writing documentation recently and I doubt that anyone wants to rewrite that because the features change.
cmk_zzz has joined #nixos
<redmp>
henri: as somebody with a whole bunch of interrelated projects outside of nixpkgs, but building with nix, is there a recommended model for composing these projects? my assumptions was flakes+overlays..
<kakakakak>
Thak you henri, I have only one override in config with mpv to use vaapi. But I see also services.accounts.daemon.enable = pkgs.lib.mkForce false; and maybe that is the past issue triggering this behaviour, because libkrb5 also builds.
<henri>
redmp: This doesn't give you the real Flakes features like hermetic evaluation but you can use Flake references.
sigmundv_ has quit [Read error: Connection reset by peer]
sigmundv_ has joined #nixos
wnklmnn has joined #nixos
tilt[m] has joined #nixos
fuiltilt has joined #nixos
<redmp>
henri: that repo only has the one default.nix with a short docstring at the top, as somebody who's not yet learned flakes, i'm not really sure what to do with this.. do you have any recommendation for documentation i can read about flake-compat?
dev_mohe has joined #nixos
Ollie[m]1 has joined #nixos
yangm4 has joined #nixos
<yangm4>
hello
<winsome[m]>
upgrading to unstable fixed the `hardware.keyboard.zsa.enable` being missing. Thanks clever
henri has quit [Quit: ERC (IRC client for Emacs 27.1)]
<redmp>
henri: thank you! one last question before i dive into this: do i write flake.lock manually, or is that generated by some tool.. and if generated, which tool?
<redmp>
tilt[m]: ok, there isn't an example of exactly how to do it here, but .. https://releases.nixos.org/nixops/latest/manual/manual.html in the table of contents, click 3.1 to see the VM example.. in that figure, you can see that `webserver` is a function taking `{ config, pkgs, ...}` arguments and returning a _machine configuration_ .. so in the _machine configuration_ you write nix just like in your
atcasdadasdas has joined #nixos
<redmp>
configuration.nix file!
<tilt[m]>
thanks @redmp
<tilt[m]>
I'll give that a try!
<redmp>
tilt[m]: so you'd take that example and add this into the mix: `environment.systemPackages = with pkgs; [java vim git];`
atcasdadasdas has quit [Client Quit]
philr has joined #nixos
<{^_^}>
[nixos-homepage] @garbas pushed to commercial-page « Add 2 more locations of Tweag »: https://git.io/JqUiw
<tilt[m]>
@redmp
dev_mohe has quit [Remote host closed the connection]
<iyefrat>
is switching between nix 2.3 and 2.4pre achived by installing and uninstalling nixUnstable?
<ambro718>
Hi. How can I add some shell code to buildFHSUserEnv to be run automatically (e.g. export)?
vidbina_ has quit [Ping timeout: 246 seconds]
garionion has joined #nixos
<garionion>
hi, has anyone some ideas why my nginx keeps failing with »[emerg] 1403#1403: open() "/run/nginx/nginx.pid" failed (2: No such file or directory)« ?
<immae>
garionion: Did you modify the systemd service? (in particular the RuntimeDirectory entry)
<{^_^}>
[nixos-homepage] @garbas pushed to commercial-page « Remove not needed files and create redirects »: https://git.io/JqUPE
ambro718 has quit [Remote host closed the connection]
<garionion>
immae: no, the problem was, that nginx could not open log files (because they where already owned by root)
<exarkun>
Hm. This weird thing is happening where services.postfix.enable defaults to true unless I explicitly set it to false. I don't have to mkForce it, either.
<immae>
oh ok
<exarkun>
If it were being turned on by some module, there'd be a conflict, right?
<dev_mohe>
Hi, is it really a good idea that environment.noXlibs adds some overlays that you really don't detect if you just use the minimal profile. It took me ages to find out that's the reason jitsi-meet didn't work especially as building the mentioned packages imperatively did work. Is there any idea except RTFM?
<immae>
exarkun: as you suggested, a module is able to set it to true with some priority
<exarkun>
any tricks to discovering which module is doing it?
starlord has joined #nixos
<immae>
(mkDefault and mkForce are just priorities set to 1000 and 50 respectively)
<immae>
git grep postfix.enable ?
<exarkun>
That's a big haystack :)
simba1 has quit [Ping timeout: 240 seconds]
<immae>
yeah
<immae>
I know no better way
<starlord>
I'm playing around with flakes and trying to cross compile the NXP kernel fork for aarch64, however it seems to require me to enable binfmt wrapper for aarch64, can't i cross compile this stuff? If not, why and what do I need to fix?
<immae>
exarkun: another way is to find with which priority the nasty module is setting postfix.enable to, then nix will complain that foo.nix and bar.nix have the same priority but incompatible values and will leak the module file
<immae>
(it might not work depending on how verbose nix decides to be)
<starlord>
redmp: you need to specify the outputs.packages.<system>.<package> for every package you want to support, you can use flake-utils https://github.com/numtide/flake-utils to automatically set that up for every supported arch in nixpkgs
civodul has quit [Quit: ERC (IRC client for Emacs 27.1)]
IRCsum has joined #nixos
starlord has quit [Remote host closed the connection]
cole-h has quit [Ping timeout: 265 seconds]
dev_mohe has quit [Quit: dev_mohe]
KarlJoad has joined #nixos
<KarlJoad>
I am using mkRenamedOptionModule on a set and I believe it is erroring out on a list somewhere. The error message is very vague, and doesn't actually give me any help. Does someone have resources on this area?
andymandias_ has quit [Quit: ZNC 1.8.1 - https://znc.in]
<colemickens>
I bet it's some nix process trying to do the download, gets stuck and then jams everything up. maybe there's a missing http timeout or something
lovesegfault has quit [Quit: WeeChat 3.0.1]
tom39291_ has quit [Quit: WeeChat 2.9]
iyefrat has quit [Remote host closed the connection]
<supersandro2000>
colemickens: happens to me like once a week for some reason