<disasm>
is there a way I can force it to reinstall in a more verbose mode so I can see what's going on?
rcshm has quit [Read error: Connection reset by peer]
o1lo01ol1o has quit [Ping timeout: 245 seconds]
rcshm has joined #nixos
halfbit has quit [Ping timeout: 252 seconds]
disasm has quit [Quit: WeeChat 2.0]
__monty__ has quit [Quit: leaving]
<pie____>
bennofs[m]: well, i dont know, but nevermind for now
illegalprime has joined #nixos
disasm has joined #nixos
<elvishjerricco>
disasm: In case it's relevant, Virtualbox doesn't support persisting efi variables between boots. The efiInstallAsRemovable problem seems very weird though
<disasm>
samueldr: ah, I think that might have done it
rcshm has quit [Read error: Connection reset by peer]
<disasm>
samueldr: so it will only install the bootloader if you set --install-bootloader flag?
rcshm has joined #nixos
<disasm>
going to reboot and see if it worked, but efibootmgr -v looks better!
disasm has quit [Quit: WeeChat 2.0]
xkapastel has quit [Quit: Connection closed for inactivity]
rcshm has quit [Read error: Connection reset by peer]
camsbury has quit [Ping timeout: 245 seconds]
erasmas has quit [Quit: Lost terminal]
rcshm has joined #nixos
<samueldr>
>> Causes the boot loader to be (re)installed on the device specified by the relevant configuration options.
<samueldr>
and search for NIXOS_INSTALL_BOOTLOADER in nixpkgs to see how it ends up being used
<samueldr>
grub has a couple conditions it checks for to reinstall itself, among others that flag
<samueldr>
not sure if it should have installed in your case without the flag :/
<pie____>
im getting really strange buildPythonPath behaviour, i get one error if i use fetchPypi and I get a different error if i use src = ./the-extracted-pypi-archive
maximiliantagher has quit [Ping timeout: 240 seconds]
eadwu has joined #nixos
ym555 has joined #nixos
rcshm has quit [Read error: Connection reset by peer]
rcshm has joined #nixos
disasm has joined #nixos
disasm has quit [Client Quit]
mbrgm_ has joined #nixos
rcshm has quit [Read error: Connection reset by peer]
disasm has joined #nixos
rcshm has joined #nixos
mbrgm has quit [Ping timeout: 252 seconds]
mbrgm_ is now known as mbrgm
orivej has quit [Ping timeout: 252 seconds]
rprije has quit [Read error: Connection reset by peer]
disasm has quit [Client Quit]
rprije has joined #nixos
Cale has joined #nixos
erictape1 has joined #nixos
disasm has joined #nixos
<{^_^}>
[nixpkgs] @samueldr opened pull request #57100 → linux: Makes the -dev output reproducible → https://git.io/fhhln
jluttine has quit [Ping timeout: 245 seconds]
erictapen has quit [Ping timeout: 240 seconds]
<disasm>
samueldr: so my motherboard did something funky when I installed using efibootmgr where it lost the path after reboot, but installing as removable worked with that flag, thanks!
maximiliantagher has joined #nixos
<disasm>
but sadly memtest86+ just reboots :( so switching to grub to get that as a boot entry was a fail.
<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/fhh8q
jluttine has quit [Ping timeout: 245 seconds]
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/fhh8s
<{^_^}>
[nixpkgs] @worldofpeace pushed commit from @Infinisil to release-19.03 « turses: Fix »: https://git.io/fhh8Z
disasm has quit [Quit: WeeChat 2.0]
disasm has joined #nixos
sondr3 has quit [Quit: WeeChat 2.4]
rcshm_ has joined #nixos
rcshm has quit [Read error: Connection reset by peer]
simukis has quit [Quit: simukis]
schjetne has joined #nixos
aanderse has joined #nixos
schjetne has quit [Ping timeout: 245 seconds]
rcshm_ has quit [Read error: Connection reset by peer]
rcshm has joined #nixos
init_6 has quit []
eadwu has quit [Quit: WeeChat 2.4]
ym555_ has joined #nixos
rcshm_ has joined #nixos
rcshm has quit [Read error: Connection reset by peer]
ym555 has quit [Ping timeout: 255 seconds]
<elvishjerricco>
Hm. systemd-modules-load.service is still running when systemd-fstab-generator causes a 9p filesystem to mount, so the 9pnet_virtio module isn't loaded in time. No idea how to fix that
rcshm_ has quit [Read error: Connection reset by peer]
rcshm has joined #nixos
stepcut has joined #nixos
Ariakenom has quit [Read error: Connection reset by peer]
rcshm has quit [Read error: Connection reset by peer]
eadwu has joined #nixos
rcshm has joined #nixos
eadwu has quit [Client Quit]
o1lo01ol1o has joined #nixos
camsbury has joined #nixos
freeman42x has quit [Ping timeout: 258 seconds]
o1lo01ol1o has quit [Ping timeout: 252 seconds]
init_6 has joined #nixos
camsbury has quit [Ping timeout: 255 seconds]
rcshm has quit [Read error: Connection reset by peer]
<yl[m]>
I saw that, but I thought there's a nix-build version as well
<yl[m]>
thx!
<simpson>
Hm. I have a NixOS installation that I've had for a while. I'd like to keep having it for a while longer. I have noticed a pattern wherein I run GC and then some stuff disappears from my user profile. How can I improve this?
<infinisil>
therealwaphire[m: (i was hoping they would be here and see the ping)
<{^_^}>
[nixpkgs] @kalbasit merged pull request #57102 → doc: format the documentation → https://git.io/fhh8y
<{^_^}>
[nixpkgs] @kalbasit pushed to master « doc: format the documentation (#57102) »: https://git.io/fhhBq
ErAsMuS has quit [Ping timeout: 252 seconds]
<pie___>
andi-, so...nix aborts if theres multiple versions of the same python package but i cant use why-depends to figure out why because the package wont finish building...
vonfry has quit [Quit: WeeChat 2.4]
wykurz has joined #nixos
<yl[m]>
I'm still stuck waiting on staging to advance to master w.r.t #56927
<pie___>
is there an idiomatic way to remove something from the buildinputs when overriding a package?
<smolboye[m]>
helo
selfsymmetric-mu has quit [Ping timeout: 252 seconds]
Guest72088 has quit [Read error: Connection reset by peer]
sinner has joined #nixos
sinner is now known as Guest75742
Rusty1 has quit [Quit: Konversation terminated!]
wykurz has quit [Quit: Leaving]
ym555_ has quit [Quit: leaving...]
<wedens>
pie___: `buildInputs = remove x old.buildInputs;`
<pie___>
wedens, ok but how do you get x
<pie___>
im trying to replace a dependency with a different version
<wedens>
pie___: you don't need to modify buildInputs then. just provide different version to a derivation
pie___ has left #nixos ["Leaving"]
pie___ has joined #nixos
<pie___>
ugh, oops
<wedens>
`callPackage ./foo.nix { dep = dep_1_0; };` or something like that
<pie___>
wedens, i might have been rather opaque, there are other things as well in the buildinputs and I dont want to poke them, if I just appaend to the buildInputs then there will be two different versions of the thing?
<pie___>
wedens, im not talking about function parameters?
<wedens>
pie___: why not? where else do you get dependencies? it's a natural way to override dependencies in a derivation
<pie___>
wedens, im given a derivation and i want to replace one of its several/many dependencies
<pie___>
maybe im just not getting you
<pie___>
.overrideAttrs (a: {buildInputs = (remove x a.buildInputs) ++ [ new_x ] }) would work, if i had a decent way to obtain x
hamishmack has joined #nixos
<wedens>
`.override { x = new_x; }`
slack1256 has quit [Ping timeout: 245 seconds]
slack1256 has joined #nixos
<wedens>
that'll work if `x` from `buildInputs` is not defined in let binding
<pie___>
doh, had to stare at it for a minute
<pie___>
yeah that should work actually in this case
<pie___>
for the sake of argument, what if it was a let bindig
<pie___>
also that makes sense i guess, override lets you change the inputs, overrideattrs lets you change the output
<wedens>
if it was in a let binding, you'll have to modify `buildInputs` directly (if it's the only place where this dependency is used). filter it by name or something
rauno has quit [Remote host closed the connection]
<pie___>
wedens, it worked out in the end \o/ i needed to use your suggestion in a different place
<contrun[m]>
Anyone else encountered this problem? The ghc version required by nix is missing.
<wedens>
linarcx: what happens if you run i3status directly
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<mog>
im setting up a nix-shell environment for my team. i have the buildInputs listed out but i want to add the option to import a list of custom buildInputs from a personal.nix
<mog>
how do i get the default.nix to add and import something like that
stepcut has quit [Read error: Connection reset by peer]
stepcut has joined #nixos
<wedens>
mog: one of options: you can add a parameter for list of extra dependencies
<mog>
that sounds like what i want how do i do that?
<wedens>
mog: `{ a, b, c, extra ? [] }: mkDerivation { buildInputs = [ a b c ] ++ extra; ... }`
<wedens>
another option is to use `overrideAttrs`
gagbo has joined #nixos
<wedens>
`drv.overrideAttrs(old: { buildInputs = old.buildInputs ++ [ x y z ]; })` where `[x y z]` are extra dependencies
<mog>
i would like to do it the ++ extra way how do i structure and import the personal.nix file correctly
alex`` has joined #nixos
<immae>
Hi! I’m trying to understand why a derivation built with `previousNixpkgs.python2Packages.buildPythonApplication` works while the same one with `currentNixpkgs.python2Packages.buildPythonApplication` doesn’t (everything else unchanged, all dependencies in the derivation are using the currentNixpkgs), where previousNixpkgs is nixpkgs-18.09.1834 and currentNixpkgs is 19.03beta171477. Were there any breaking
<immae>
changes in python2 packages in 19.03 branch? (note that it compiles fine, it’s only at runtime that it complains about `UnicodeEncodeError: 'ascii' codec can't encode character u'\xe9' in position 52: ordinal not in range(128)`)
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<wedens>
mog: somnething like `drv = callPackage ./foo { extra = (import personal.nix).extraDependencies; }` or `drvWithExtra = drv.override { extra = (import personal.nix).extraDependencies; }` if it's already assigned.
kvda has joined #nixos
<mog>
wedens, thanks i think i got what i need working
Ariakenom has quit [Read error: Connection reset by peer]
<clever>
mog: what error does it give?
<mog>
lerror: cannot coerce a built-in function to a string, at /nix/store/fv7y59lyvbmqvqv7qr5glw1ndzh60rfy-nixos-18.09.2318.4c089d328fe/nixos/pkgs/stdenv/generic/make-derivation.nix:177:11
Dr8128 has quit [Quit: Leaving.]
<mog>
i tried passing as strings and it complained they werent complete paths
<clever>
mog: you need more ()
Dr8128 has joined #nixos
<clever>
] ++ (import personal pkgs);
<mog>
same error still
<clever>
its parsing as: ([ foo ] ++ import) personal pkgs;
<mog>
so before this was working with import ./personal.nix
<mog>
but i wanted it to work if the user doesnt provide one
<mog>
so i moved my personal.nix to the default_personal one
<clever>
mog: can you add all the files to a gist?
<mog>
error: undefined variable 'personal' at /home/mog/code/use_nix_please/default.nix:26:16
<clever>
mog: line 6-13 is a single attribute set
<clever>
mog: you put personal= into elixir's arguments, not the let block
<mog>
oh i see
<mog>
whoops
<mog>
fixed
<mog>
thanks
<clever>
mog: personalFile = ./personal.nix; could also be in the main let block, no real need to nest them
<mog>
ya. i stole the if block from something similar to what im trying to do
gagbo has joined #nixos
<mog>
in my nix configs i do an imports = []; to combine files. if i wanted to do that for this stdenv.mkDerivation what is the proper way to do that clever?
<clever>
mog: imports is only for modules, but stdenv.mkDerivation doesnt involve modules
iXialumy has joined #nixos
<mog>
so there isnt an easy way i could have another file where people could define their own env variables?
kvda has joined #nixos
<manveru>
mog: depends on how your personal.nix looks like
init_6 has quit [Read error: Connection reset by peer]
<mog>
so right now its just that one var pkgs
<mog>
but this is just the start of project so i can setup any way that works
<manveru>
what you could do is something like `stdenv.mkDerivation ({ your = default; stuff = is; here = and; } // (import personal))`
<manveru>
if personal is in the form of an attrset
johanot has joined #nixos
jpaus has joined #nixos
iXialumy has quit [Quit: Page closed]
<mog>
whats the format of an attrset?
<manveru>
> { a = 1; b = 2; } // { b = 3; }
<manveru>
like this
<manveru>
you can merge them this way
<clever>
thats 2 attrsets, being merged with the // operator
<jpaus>
hey there, i am totally new to nixos and wanted to get nginx to run with certbot as discribed in the manual. But on the nginx wiki page the configuration is done in configuration.nix and on the linked manual for lets encrypt its done in the nginx conf file directly
<clever>
jpaus: nixos automates the nginx and certbox config with a single flag
<laas>
I'm getting a lot of warnings like `substituteStream(): WARNING: pattern '/usr/bin/readlink' doesn't match anything in file '/nix/store/clw7q0vplzc3nb13354vmhrl9bmx7q8j-udev-rules/64-md-raid-assembly.rules' when rebuilding
<laas>
`
<laas>
is this normal?
<srhb>
laas: It's new since 5ff872aa24983cf3e1cf28bb990042846c1a97ee I think
<srhb>
I'm not sure why we need that..
<{^_^}>
[nixpkgs] @Ma27 pushed commit from @Infinisil to release-19.03 « sad: Fix build »: https://git.io/fhhEp
<srhb>
oh hm, that's old..
<srhb>
So I guess there really is a substitute there that needs tweaking..
<{^_^}>
[nixpkgs] @danbst merged pull request #56682 → network-scripting: do not run resolvconf if /etc/resolv.conf is managed manually → https://git.io/fhAPs
<{^_^}>
[nixpkgs] @danbst pushed commit from @volth to master « network-scripting: do not run resolvconf if /etc/resolv.conf is managed manually (#56682) »: https://git.io/fhhuf
<delan>
how can I point a new NixOS install to a slightly old nixos-unstable, given an existing install with a generation that was on the channel state I’m after?
<pstft>
Hi. I'm trying to open a LUKS-encrypted partition from NixOS on Raspberry Pi 3, but cryptsetup open dies from a SIGKILL. Here is the output of --debug: https://bpaste.net/show/0375bc7c4f3f Any ideas how to fix this?
<dacuna>
hey all, hope it's ok to ask questions here
<dacuna>
i installed nixos yesterday but i'm having issues doing any additional rebuilds
<dacuna>
the rebuild seems to succeed but as soon as I reboot the computer, I'm back to the old state
<lassulus>
dacuna: do you have a seperate /boot partition? is it mounted when you rebuild?
<dacuna>
yes, it's separated
<dacuna>
should I mount it before doing that?
baum has quit [Quit: Lost terminal]
<lassulus>
yes, the boot entries need to be updated, so you can boot into the new system
<dacuna>
ahh, I see
<dacuna>
i'll give that a try, thanks!
dacuna has quit [Client Quit]
asymmetric has quit [Ping timeout: 245 seconds]
eadwu has joined #nixos
juhe has joined #nixos
johanot has joined #nixos
dacuna has joined #nixos
<dacuna>
no dice, am I supposed to mount boot into /mnt/boot? somewhere else? is there a possibility that since there weren't any changes it didn't try to change the boot entry?
<clever>
dacuna: when installing, /boot must be at /mnt/boot/, after putting the rootfs at /mnt/
<{^_^}>
[nixpkgs] @markuskowa merged pull request #56144 → digikam: fix media playback by adding ffmpeg to buildInputs → https://git.io/fhFGE
<clever>
dacuna: if the install is already done, then boot has to be at /boot/ and you may need to re-run nixos-generate-config to fix hardware-configuration.ni
<clever>
x
dacuna has quit [Client Quit]
<{^_^}>
[nixpkgs] @grahamc merged pull request #55436 → Nixos manual: error out on missing IDs → https://git.io/fhHGr
<gleber_>
I remember seeing a new (possibly experimental) command to rewrite a single Nix store object into CAS form. I can no longer find it anywhere... Anyone has ideas what I might be looking for?
<gchristensen>
gleber_: maybe poke around the edolstra/nix fork's branches
<gchristensen>
or nixos/nix's branches
<gleber_>
gchristensen: I've spent 15 minutes doing it, no luck so far
<dacunap>
Also, now I see many generation entries on the grub bootloader
rprije has quit [Remote host closed the connection]
<dacunap>
How do I clean those up?
rprije has joined #nixos
<clever>
dacuna: those are supposed to be there, a new one is added every time you nixos-rebuild
dacuna has quit [Ping timeout: 256 seconds]
troydm has quit [Quit: What is Hope? That all of your wishes and all of your dreams come true? To turn back time because things were not supposed to happen like that (C) Rau Le Creuset]
<dacunap>
I understand, but over time they will just build up or does it keep a maximum amount of entries?
<clever>
dacunap: grub or systemd?
troydm has joined #nixos
fusion809 has joined #nixos
disasm has quit [Quit: WeeChat 2.0]
dacunap has quit [Ping timeout: 256 seconds]
Shouou has joined #nixos
linarcx has quit [Ping timeout: 245 seconds]
<Shouou>
I'm getting "default.nix:1:1 called without required argument 'fetchFromGithub'" when trying "nix-build -A app", but fetchurl works fine. Does anyone know a workaround/fix?
knupfer has quit [Remote host closed the connection]
xkapastel has joined #nixos
knupfer has joined #nixos
simukis has quit [Quit: simukis]
jpaus has joined #nixos
<tokudan[m]>
probably a simple question: I've got a string "abc${myport}def". during evaluation, nix complains gives me an error, stating that myport is an integer that needs to be converted first. how can I do that?
<immae>
tokudan[m]: toString
<tokudan[m]>
immae: thanks :)
disasm has quit [Quit: WeeChat 2.0]
<tokudan[m]>
wasn't able to find that through google
<Orbstheorem>
Hello, now that we have nix-locate and `nix search` is nox still useful?
<gchristensen>
more specifically, I'd like to use the standard x86_64-linux system, but set `kernelTarget = "vmlinux";` -- anyone familiar with this? (cc matthewbauer[m], fingers crossed)
Dr8128 has quit [Ping timeout: 268 seconds]
_kwstas has quit [Quit: _kwstas]
knupfer has quit [Quit: knupfer]
<Izorkin>
infinisil: checked this PR #56255 #56286 #56423 ? :)
<{^_^}>
[nixpkgs] @pSub pushed to master « mhddfs: fix build (see #53716) »: https://git.io/fhh2N
<{^_^}>
[nixpkgs] @pSub pushed to release-19.03 « mhddfs: fix build (see #53716) »: https://git.io/fhh2x
<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to release-19.03: https://git.io/fhhaT
<pie____>
lewo: hi o/ profpatsch told me you might have some good input on this: I'm trying to pakage cuckoo sandbox, and it needs a postgres / mongo / ?mysql? sever in its checkPhase for some tests, what can/should I do? (sidenote: ive never used any containers stuff before)
o1lo01ol1o has joined #nixos
<{^_^}>
[nixpkgs] @amazari opened pull request #57125 → zoneminder: Fix database name and username → https://git.io/fhhaq
jtojnar_ has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @timokau opened pull request #57126 → Sage maxima downgrade → https://git.io/fhhaY
<{^_^}>
[nixpkgs] @pSub pushed to fix-haskellPackages-xattr « haskellPackages.xattr: fix build with attr-2.4.48 (see ##53716) »: https://git.io/fhha8
<noonien>
i'm guessing it's a hash of all the fixed-output dependencies
<clever>
noonien: its a hash of the serialized .drv file, which includes the paths to its inputs, and the build instructions
<noonien>
i see
<noonien>
what does an "input" mean, in this case?
<noonien>
just nix store paths?
<clever>
every key=value in the derivation arguments
<clever>
including the buildInputs=value;
stepcut has quit [Remote host closed the connection]
<noonien>
i'm guessing it's the "resolved" derivation
stepcut has joined #nixos
<clefru>
For C include files, is there a convention to include directory prefixes? For instance #include <freetype/ft2build.h> vs #include <ft2build.h>. I have the problem that a nodejs package in its native part uses #include <node_version.h> while our derivation would require the package to do #include <node/node_version.h>. I wonder what's normal and how we should structure our derivation's include directories
<noonien>
so, no more variables
<clever>
noonien: yep
<clever>
noonien: its basically all strings at that point
<noonien>
hmm, i see. in that case, if derivation A requires B during build, and i don't have B built, nor is it cached for some reason (this should not happen). would i need to build B before being able to download A from the binary cache? provided B is not a fixed-output derivation
<ekleog>
do we have any way of knowing which path nix is speaking about when saying “warning: dumping very large path (> 256 MiB); this may run out of memory”?
<clever>
noonien: it will try to fetch B from the cache first
stepcut has quit [Read error: Connection reset by peer]
<clever>
noonien: and its not legal for a cache to have A but not B
<clever>
ekleog: add more -v's
stepcut has joined #nixos
drakonis1 has quit [Read error: Connection reset by peer]
<ekleog>
clever: oh wow I had never noticed the -v option
* ekleog
's life is about to change
<noonien>
ok, i'll rephrase, B is only a build-type dependency, would I need to download it to then install A? given that B will be of no use
<clever>
noonien: if A is already in the binary cache, it can skip downloading B
drakonis1 has joined #nixos
<noonien>
hmm, i think i'm nissing something, is there a way i can get a grimpse into a .drv?
drakonis1 has quit [Read error: Connection reset by peer]
<noonien>
hmm, i see it basically contains the paths to its sources and dependency derivations
<clever>
yep
drakonis1 has quit [Read error: Connection reset by peer]
<ekleog>
huh looks like it's a `lib.cleanSource ./.` in a directory that's <500KB large o.O
<clever>
and the hash of some of that data, is what $out is computed from
<noonien>
so, given that my system doesn't have A installed, and i tell it to install A, how does it compute the hash for A so it can check if it's available from the binary cache?
<clever>
noonien: the hash of A is based on the .drv files along
<clever>
alone*
<noonien>
oh, so a .drv is generated first
<clever>
yeah
<ekleog>
oh no right was on the wrong machine, it's 1.5G large, so that's it
<clever>
stdenv.mkDerivation is a function that generates a .drv
drakonis1 has joined #nixos
<noonien>
i see, but, to generate that derivation, does it not need B? since it's a dependency?
<clever>
it has to generate B.drv
<clever>
which is just a small text file
<noonien>
OH
<noonien>
that makes sense
kreisys has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<pie___>
wait you can add MORE -v ?? :O
<clever>
pie___: i usually start with -vvvv
<pie___>
* ekleog 's life is about to change
<noonien>
so baiscally, at the end, the hash of A is the hash of all fixed-output derivation + all commands that are executed to realise A's dependencies
<clever>
noonien: yeah
<tilpner>
gchristensen: Does ${linux.dev}/vmlinux work for you?
<noonien>
clever: cool, thanks!
<{^_^}>
[nixpkgs] @pSub opened pull request #57129 → haskellPackages.xattr: fix build with attr-2.4.48 (see ##53716) → https://git.io/fhhaE
Thra11 has joined #nixos
jpaus has joined #nixos
shibboleth has joined #nixos
astrofog has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @pSub pushed 0 commits to update-calibre: https://git.io/fhhao
ddellacosta has quit [Ping timeout: 246 seconds]
jpaus has quit [Ping timeout: 245 seconds]
drakonis1 has quit [Read error: Connection reset by peer]
drakonis1 has joined #nixos
ddellacosta has joined #nixos
drakonis1 has quit [Read error: Connection reset by peer]
drakonis has joined #nixos
vidbina has joined #nixos
drakonis_ has joined #nixos
jpaus has joined #nixos
drakonis has quit [Ping timeout: 240 seconds]
Makaveli7 has quit [Quit: Leaving]
jpaus has quit [Ping timeout: 240 seconds]
<noonien>
if i have a derivation A, that depends on D0, D1, D2... D20, and D10 fails to build, interrupting the build, will the other dependencies need to be rebuilt as well?
<clever>
noonien: any that have finished building will be in /nix/store and wont be rebuilt
vidbina has quit [Read error: Connection reset by peer]
maximiliantagher has joined #nixos
<{^_^}>
[nixpkgs] @clefru opened pull request #57131 → nodejs: Remove the requirement for dir-prefix from includes → https://git.io/fhhaj
shabius has quit [Quit: Leaving]
Glider_IRC_ has quit [Ping timeout: 245 seconds]
alex_giusi_tiri has joined #nixos
fendor has joined #nixos
fendor has quit [Client Quit]
fendor has joined #nixos
drakonis has joined #nixos
knupfer has quit [Ping timeout: 240 seconds]
endformationage has quit [Ping timeout: 255 seconds]
drakonis_ has quit [Ping timeout: 268 seconds]
johanot has quit [Ping timeout: 268 seconds]
jlslnx has joined #nixos
rauno has quit [Remote host closed the connection]
<lucus16>
Why do I need to be a trusted user on remote builders while copy closure; nix build; copy closure back does not require that and achieves the same?
<lucus16>
Either I misunderstand something, or at least one of those flows
<lucus16>
* has a mismatched security level
ErAsMuS has quit [Quit: END-AUMF]
zupo has joined #nixos
<ivan>
<ivan>I'm still curious about this from two months ago, anyone know? <vaibhavsagar> what's the reasoning behind only allowing users in `trusted-users` to use the machine as a remote builder?
<ivan>
<symphorien>a machine A using B as a remote builder can add arbitrary already built derivations to the store of B
<ivan>
<ivan>afaik non-trusted-users can still build whatever derivations they want from any nixpkgs (local builds)
<Orbstheorem>
Hello ^^/. I have a doubt: If I build somethind using nix-build and two days later I run `sudo nix-collect-garbage --delete-older-than 1d` the whatever I build will be deleted, right?
<Orbstheorem>
(and the symlink ./result will be broken)
<infinisil>
Orbstheorem: The GC only deletes things that are unreachable from the GC roots
<vaibhavsagar>
ivan: the current model expects all remote build users to have full access to change any paths in /nix/store
<infinisil>
Orbstheorem: But these result symlinks are reachable, so their contents won't be deleted
<ivan>
yeah I was answering lucus16
<infinisil>
Orbstheorem: What can get deleted however are the derivations that built these ./result's, I think the nix.conf setting keep-derivations will prevent this though
<Orbstheorem>
infinisil: But the result symlink is outside the gcroots directory, how can it know there's still a link?
<lucus16>
Hmm, so sounds like there should be an option to upload only derivations and have the remote builder build everything
<infinisil>
Orbstheorem: Go check /nix/var/nix/gcroots/auto
<Orbstheorem>
Oh
<Orbstheorem>
what's the eviction policy on those roots then?
<infinisil>
Orbstheorem: Deleting the symlink
<infinisil>
The ./result symlink that is
<Orbstheorem>
Some of those links are broken
<clever>
nix's GC will clean up the broken links
<Orbstheorem>
Okay, cool
<clever>
and deleting the result link nix-builds, makes them broken
<Orbstheorem>
Thanks !
<Orbstheorem>
15483 store paths deleted, 22317.57 MiB freed \^^/
o1lo01ol1o has quit [Remote host closed the connection]
shibboleth has quit [Quit: shibboleth]
freeman42x has joined #nixos
<Orbstheorem>
:o I still had the derivation of a rootfs I bootstrapped months ago: 24957 store paths deleted, 46223.49 MiB freed
<Orbstheorem>
:D
<lucus16>
ivan: Wait, but it only fails for me after uploading like 80 paths
ixxie has joined #nixos
<Orbstheorem>
I just recovered 70GB of space xD
<lucus16>
It does the uploading paths just fine and only after that does it complain that I "am not privileged to build derivations"
o1lo01ol1o has joined #nixos
<infinisil>
Orbstheorem: I wish my GC was that fast! ZFS is notoriously slow at deleting things, and this doesn't combine well with my seemingly extra slow mirrored HDD
<Orbstheorem>
I guess that one of the reasons why I don't want to switch the fs on my nas, is because I really enjoy btrfs-send and I already have configured my machines to push the snapper backup when devices are reachable
<clever>
Orbstheorem: same reason i use zfs, so i can `zfs send` my non-mirrored machines (pure nvme) to the nas
<mog>
how can i specify a channel in a nix-shell script , i want to import nixpkgs unstable rather than the stable
<mog>
i know i can pass -I at cli
<mog>
but i want it in the script
maximiliantagher has quit []
<clever>
mog: instead of importing <nixpkgs> import (builtins.fetchTarball URL)
<mog>
there is no way to run your own cachix backend is there?
rcshm has quit [Read error: Connection reset by peer]
schjetne has joined #nixos
rcshm has joined #nixos
<gchristensen>
you can write to S3 or DO spaces directly with nix
eadwu has quit [Ping timeout: 240 seconds]
<clever>
and nix-push/nix copy, can create a dumb directory, that you can then host on anything you want
stepcut has quit [Remote host closed the connection]
<clever>
and nix-serve can just serve /nix/store/
<gchristensen>
but no, you can't use cachix's CLI to write somewhere else afaik
<mog>
i think those nix-push and nix-copy are what i want
<mog>
thank you
stepcut has joined #nixos
ilmu has quit [Ping timeout: 268 seconds]
<gchristensen>
mog: where do you want to store your cache?
stepcut has quit [Read error: Connection reset by peer]
stepcut has joined #nixos
<mog>
i built a dev environment for my office. and everything was already in binary cache except for one package had to be built from source, so i wanted to prebuild it for macos and x64 linux
matix has joined #nixos
<mog>
i can easily store it on site and serve over nginx
rcshm has quit [Read error: Connection reset by peer]
stepcut has quit [Remote host closed the connection]
talqu has quit [Remote host closed the connection]
<ddellacosta>
do folks developing haskell code on nix generally install cabal-install and cabal2nix in their default environment, and otherwise use repo/project-specific configuration for all other necessary haskell dependencies (including ghc)? Is there any reason not to also simply include cabal* in buildInputs? It seems like based on the nixpkgs haskell docs
arjen-jonathan has quit [Ping timeout: 252 seconds]
<ddellacosta>
(I'm ignoring the examples in the nixpkgs docs showing a "global" setup in ~/.config/nixpkgs/config.nix and the like)
aexoxea has joined #nixos
vidbina has quit [Read error: Connection reset by peer]
Fare has quit [Ping timeout: 245 seconds]
stepcut has quit [Remote host closed the connection]
knupfer has quit [Remote host closed the connection]
vidbina has joined #nixos
stepcut has joined #nixos
<selfsymmetric-mu>
Yes, I install globally unless there's a reason to locally disambiguate things, but that's my personal preference.
johanot has joined #nixos
troydm has joined #nixos
camsbury has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
arjen-jonathan has joined #nixos
stepcut has quit [Remote host closed the connection]
stepcut has joined #nixos
black-brick has joined #nixos
<infinisil>
ddellacosta: I always add cabal-install to nativeBuildInputs, I don't want to install it globally
<infinisil>
Also if clever sees this he'll tell you that cabal-install won't always work due to version incompatibilities and that you should instead run `runhaskell Setup.hs build` (or so) instead of `cabal build`, which is guaranteed to work
<infinisil>
clever: (ping)
o1lo01ol1o has quit [Ping timeout: 240 seconds]
Xal has quit [Ping timeout: 250 seconds]
stepcut has quit [Ping timeout: 245 seconds]
<clever>
infinisil: ive been wanting to put a bash alias, to map cabal to `runhaskell Setup.hs`
<clever>
infinisil: so it becomes invisible to the user
<infinisil>
Yeah i thought about this too
<infinisil>
shellHook would make it pretty simple actually
<clever>
but an alias wont cover `ghcid -c "cabal repl"`
<clever>
it would have to be a bash script
<infinisil>
Ah yeah
Xal has joined #nixos
rcshm has quit [Remote host closed the connection]
<clever>
infinisil: ive also been considering weather i should just compile Setup.hs, or always runhaskell it
<infinisil>
whether* xD
rcshm has joined #nixos
* infinisil
goes to check how slow runhaskell Setup.hs is
<infinisil>
Oh, but I need runhaskell installed then
<clever>
runhaskell is in the same output as ghc
<infinisil>
Yeah
rcshm has quit [Read error: Connection reset by peer]
<infinisil>
Ah and I have ghc from the nix-shell anyways