<{^_^}>
[nixpkgs] @alyssais pushed commit from @burke to master « ruby-modules: parse build_flags correctly: »: https://git.io/Je7cB
ng0 has quit [Quit: Alexa, when is the end of world?]
gustavderdrache has joined #nixos
selfsymmetric-mu has joined #nixos
<selfsymmetric-mu>
I'm trying to get this issue rolling: https://github.com/NixOS/nixpkgs/issues/70261, and this nice user Dema created a derivation. Now...the derivation doesn't put anything on the PATH, it just has an important file called `emacspeak-setup.el`. But uh, how do I know that I have the right one? If I install the Nix expression it puts it in my store, but how do I know which one is the right version?
<ddima>
selfsymmetric-mu: Not sure I follow - could you please expand on "how do I know I have the right one"? Does it mean you might have multiple versions of emacspeak installed? How did you install the derivation, do you have an overlay or so?
<systo>
question, Since Nix doesn't follow the FHS, what are the basic mountpoints I should create? Coming from FreeBSD so used to a more rigid structure, but I get the premise of the FS being used as a database i.e. a mountpoint of /nix/store/
<systo>
so /, /nix/store, /tmp, /home, anything else? /var/log?
<gchristensen>
with ZFS I do a dataset at /, /nix, /home, and /boot
<lovesegfault>
gchristensen: Do you use BIOS instead of UEFI?
<systo>
TY
<gchristensen>
lovesegfault: everywhere I can, yeah :)
<systo>
I do not, trying to use UEFI where ever possible
<gchristensen>
oh yeah /boot isn't zfs, its vfat of course
<lovesegfault>
gchristensen: Why? just the better FDE support?
<lovesegfault>
gchristensen: Aaaaah, there we go :P
<lovesegfault>
I was like "wtf his /boot is ZFS?!"
<gchristensen>
hehe
<systo>
yea /boot as fat32
<infinisil>
systo: You don't need any special partitions really, you're free to choose whatever
<infinisil>
(Well as long boot works)
<gchristensen>
systo: if you like to do recursive snapshots, make sure to put /nix/store in a different subtree than / and /home :P
<systo>
I've already got a scripted FreeBSD install using legacy mountpoints, with only the linux-compatable options set, so I'm on the hunt for a distro with musl-c
<qyliss>
My /boot will be ZFS as soon as I can get heads to build with ZFS support :)
<selfsymmetric-mu>
ddima: I have multiple directories labeled "emacspeak" in my `/nix/store/`. They are all set to epoch date to preserve determinism but one of them is in fact the most recent. How do I know which one it is?
<lovesegfault>
qyliss: heads?
<infinisil>
ddima: ls /nix/store is an antipattern in nix
<infinisil>
Oh, well usually they should have a version in the name actually
<systo>
I'm also interested in the idea of alpine+nix for IoT dev
<ddima>
infinisil: thanks ;)
<systo>
or until I get OpenNHRP ported over
<infinisil>
ddima: If it's not in the name you can't really know just from looking at the paths
<ddima>
selfsymmetric-mu: you probably should resolve the outPath by evaluating the derivation, thus I was wondering how you installed it. If you have an overlay/have this available as an attibute, it could be as simple as `nix-instantiate '<nixpkgs>' -A fooBar.emacspeak' - if you only have the nix expression randomly flying around, you'd have to do a bit more an `callPackage` to evaluate. Does that make sense?
<qyliss>
I have Linux on my motherboard instead of a BIOS
<lovesegfault>
108414669 is reported finished but the status is still ?
<lovesegfault>
is this a bug?
<infinisil>
Seems like it's too late for me
[Leary] has quit [Read error: No route to host]
<selfsymmetric-mu>
ddima: I do have the Nix expression. I don't know how to do what you're describing unfortunately but I've made a note on the issue.
<selfsymmetric-mu>
Thank you for your help regardless.
selfsymmetric-mu has quit [Remote host closed the connection]
<pie_>
oh my god this tool is amazing
<pie_>
re: i got gammaray to work
<ddima>
selfsymmetric-mu: so, lets say you have it in /tmp, you could do it like `nix-instantiate --eval -E 'let pkgs = import <nixpkgs> {}; in pkgs.callPackage /tmp/emacspeak.nix { }' -A outPath`
knupfer1 has joined #nixos
vld has quit [Ping timeout: 248 seconds]
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/Je7CL
slack1256 has quit [Remote host closed the connection]
shibboleth has quit [Quit: shibboleth]
supermathieu has quit [Quit: Leaving]
supermathieu has joined #nixos
<supermathieu>
Howdy, I am trying to parse a yaml file from a nixos module. I couldn't find a builtin parser, is there a way to do it 'natively' or to call some bash code in a let binding?
rardiol has quit [Ping timeout: 276 seconds]
mexisme has quit [Ping timeout: 250 seconds]
<qyliss>
You could do IFD, but that wouldn't be allowed in Nixpkgs. I think your only other option is to convert to JSON or TOML outside Nix
mexisme has joined #nixos
noudle has quit []
<supermathieu>
qyliss: What is IFD?
<clever>
,ifd
<{^_^}>
import-from-derivation (IFD) is when you evaluate nix from a derivation result, for example `import (pkgs.writeText "n" "1 + 1")` will evaluate to 2. This is sometimes problematic because it requires evaluating some, building some, and then evaluating the build result. It has been described as "such a nice footgun."
<supermathieu>
It was too obvious ;)
h0m1 has quit [Ping timeout: 250 seconds]
<supermathieu>
Just wondering, would it be possible to call some bash code?
dirkx has quit [Ping timeout: 250 seconds]
<clever>
supermathieu: pkgs.runCommand takes a bash script, but to get the value within an expression, you must import from that derivation
selfsymmetric-mu has joined #nixos
<supermathieu>
clever: alright, thanks. In that case, it effectively sounds easier to simply convert to json
h0m1 has joined #nixos
dirkx has joined #nixos
<pie_>
hm apparently nixos has flatpak support, i should try it out
<pie_>
clever: now if only ,ifd referred to why its a footgun or something
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
<selfsymmetric-mu>
if it gets snap support too then maybe there will finally be an Aether client available
<pie_>
supermathieu: yeah i probably just need to update
lambda-11235 has quit [Quit: Bye]
<mla>
clever: oh cool looks simple, thanks i'll give that a try
<pie_>
supermathieu: seems to work now
<supermathieu>
pie_: glad to hear that!
<pie_>
now if only i actually knew how to use this tui
<supermathieu>
I am actually trying to find a way to use it properly. I keep trying to find playlist, but what I love with the youtube ui is that I can simply start at one song and it just autoplay everything. (Thats how I discover new songs)
<supermathieu>
I haven't found a way to do that with mpsyt
ddellacosta has quit [Read error: Connection reset by peer]
<supermathieu>
At least, not yet
<pie_>
cant fet i to actually play anything though. odd.
ddellacosta has joined #nixos
<pie_>
supermathieu: im curious about freetube but that probably doesnt have a clu
<pie_>
*cli
<pie_>
whelp. :I 'Problem playing last item: <urlopen error [Errno 101] Network is unreachable>'
<mla>
my learning method so far has been trial and error )
jjakob has joined #nixos
<pie_>
if you have a path object like ./something, when you cast it to a string via ${} ...makes things such... that the path gets copied to the store, and it substitutes the new path
<pie_>
you can put any nix expression in ${}
<pie_>
mla: i strongly recommend playing around in the repl a bit with some random stuff
<pie_>
> "${1 + 1}"
<{^_^}>
cannot coerce an integer to a string, at (string):271:2
<pie_>
:I
* pie_
trips on a banana
<pie_>
> "${builtins.toString (1 + 1)}"
<{^_^}>
"2"
commander has quit [Remote host closed the connection]
<lordcirth_>
Is that the client or the server config?
<clever>
lordcirth_: config for the justdoit script, if you run `justdoit` on the netboot client, it will nuke /dev/vda, and install nixos using luks
<clever>
lordcirth_: the main file, is meant to be a netboot server, on a laptop, you aim lan to an ethernet card, and wan to a wifi card, and it will NAT between them for you
<clever>
lordcirth_: lines 20-27 are then a nixos config for the client, that will boot over the network
<clever>
lordcirth_: the original use, is that you plug any machine into the laptop ethernet, netboot it, run justdoit, and boom, its now nixos!
<lordcirth_>
neat. So if it's to run a server, I can remove some parts
<clever>
lordcirth_: yeah, customize lines 20-27 and 12 to suit your needs
<clever>
lordcirth_: and if its the main router, you dont need the nat stuff
<lordcirth_>
Might need NAT at some point, for clusters
<Church->
I did on a brand new laptop and had to work in a framebuffer for a few days till I got graphics working with some help lol
<clever>
lordcirth_: this will use 2 instances of nixpkgs, to build a qemu that runs nixos
<Church->
did it*
<lordcirth_>
I want to be able to start with a basic NixOS server, use nixops to turn itself into a cluster head node, most importantly a PXE server, then be able to 0-touch deploy the rest of the cluster
<clever>
lordcirth_: one nixpkgs is forced to x86_64-linux (for the nixos guest)
<clever>
lordcirth_: but the 2nd nixpkgs (for qemu and bash) are left to default to the current os
<clever>
lordcirth_: so if you nix-build that on a mac, it winds up running nixos under the darwin build of qemu!
<{^_^}>
nixops#1189 (by cleverca22, 17 weeks ago, open): plan for supporting custom partition layouts and custom FS's on any backend
endformationage has quit [Quit: WeeChat 2.6]
<clever>
lordcirth_: for example, you could just netboot a machine, then just point nixops towards the ip, and let nixops do the entire install
<lordcirth_>
clever, from the OS in RAM? Ooh, didn't think of that
<lordcirth_>
No need for a separate preseed config, nice
<clever>
lordcirth_: yep
<clever>
lordcirth_: step 3 in the issue is how to copy-closure with a chroot on the remote end
<zeta_0>
i ran `nix-collect-garbage -d`, which deleted a bunch of stuff but when i rebooted it showed the same number of generations, i thought that command was supposed to remove all previous generations and your most recent configuration starts at generation 1 ?
<clever>
lordcirth_: and step 4 tells nixos-install that its already been built, just fix the bootloader
<zeta_0>
the older generations are still there?
<clever>
zeta_0: the grub config isnt updated until you nixos-rebuild switch
<clever>
lordcirth_: the rest of the stuff in the issue, is just how to get a nixos env running on the target, in ram
<zeta_0>
clever: oh, thanks
<zeta_0>
clever++
<{^_^}>
clever's karma got increased to 283
sigmundv has quit [Read error: Connection reset by peer]
<systo>
anyone encountered systemd-boot-builder.py complaining about machineid list index out of range when trying to install
lord| has quit [Read error: Connection reset by peer]
dansho has quit [Quit: Leaving]
dsg has quit [Ping timeout: 276 seconds]
<zeta_0>
clever: i am finally able to use ghc.nix to build ghc, i couldn't get it to build earlier because i missed a few tiny details
lord| has joined #nixos
<clever>
zeta_0: ah
sigmundv has joined #nixos
<zeta_0>
clever: it takes about an hour to build ghc, it sucks waiting, but i am glad i finally got it to work
<systo>
got all my zfs datasets mounted, but can't seem to cmplete install
<zeta_0>
clever: i ran nixos-rebuild switch and when i rebooted the previous older generations are still there?
<clever>
zeta_0: was /boot mounted correctly?
<lordcirth_>
zeta_0, did you run it as root/sudo?
dsg has quit [Ping timeout: 265 seconds]
<zeta_0>
when i installed nixos months i followed the instructions the best i could so i think i mounted it correctly, and i ran it as root, so i don't know
<clever>
zeta_0: is it booting with efi or legacy?
<zeta_0>
efi
<clever>
zeta_0: what does `df -h /boot` report?
<zeta_0>
Filesystem Size Used Avail Use% Mounted on
<zeta_0>
[zeta@nixos:~]$ df -h /boot
<clever>
zeta_0: the 3rd line is missing
<zeta_0>
clever: what is missing?
<zeta_0>
on the 3rd line?
<clever>
zeta_0: compare what you said on irc to what you copied from the terminal
<lordcirth_>
It's probably on their screen, along with a flood control warning
<zeta_0>
it has the same stuff, it just pasted all messed up
<clever>
lordcirth_: it starts with a /, which is the trigger for commands
<lordcirth_>
ah, right, lol
dsg has joined #nixos
<clever>
zeta_0: what is the last line of output from `df -h /boot` ?
<lordcirth_>
put a space in front this time
<zeta_0>
are you guys not able to see this output:
<lordcirth_>
no, not with a leading /
<zeta_0>
dev/nvme0n1p3 510M 14M 497M 3% /boot
dx_ has quit [Quit: leaving]
<lordcirth_>
or that
<zeta_0>
ok i removed /
<clever>
it must be a bug in the emacs irc client
<lordcirth_>
To not escape pastes?
<clever>
lordcirth_: to make it look like its still sending a msg, even if its a cmd
<zeta_0>
so you guys can't see the output, how strange
<clever>
zeta_0: all commands in irc start with a /
<lordcirth_>
zeta_0, lines beginning with / are IRC commands
<clever>
zeta_0: can you pastebin the full output of `fdisk -l /dev/nvme0n1 ; df -h` ?
<zeta_0>
i must of made a little mistake somewhere, i think i will redo the`nix-collect-garbage -d` one more time just to make sure
<clever>
zeta_0: `fdisk -l` only works as root
<clever>
zeta_0: `nix-collect-garbage -d` cant delete system profiles if it lacks root
<zeta_0>
clever: yes, fdisk -l works as root
bvdw has quit [Read error: Connection reset by peer]
<clever>
zeta_0: but the pastebin only shows the error, nothing useful
bvdw has joined #nixos
<zeta_0>
let me try it one more time, i'll be back in a couple minutes
zeta_0 has quit [Quit: rcirc on GNU Emacs 26.3]
dsg has joined #nixos
<pie_>
i think when something ctually fails then the whole system goes down (maybe theres only one thing) - can i do this on the journalctl logs after the fact? <lordcirth_> pie_, systemctl list-units --failed might be useful
<clever>
pie_: there is also `journalctl -b -1`
zeta_0 has joined #nixos
<pie_>
clever: yeah i know about that one \o/
<pie_>
maybe it does tell you what failed
<clever>
it should
<clever>
if it was a clean shutdown
<pie_>
in any case i got the container working a few minutes ago, i needed enableTun = true and then the tinc service didnt fail
<pie_>
clever: i think the container runner was killing it so probably not
dansho has quit [Quit: Leaving]
<zeta_0>
it worked, i am back at generation 1, everything loads so much faster now, i had 242 generations before i executed the `nix-collect-garbage -d` command
<clever>
zeta_0: who does it load faster? it shouldnt have any real impact on the speed
mexisme has joined #nixos
<zeta_0>
clever: it definitely seems to boot faster than before, as for the program load times its hard to tell
<clever>
nothing in the boot process should care about how many generations you have
viric has quit [Read error: Connection reset by peer]
medvid has quit [Ping timeout: 268 seconds]
mexisme has joined #nixos
mexisme has quit [Ping timeout: 248 seconds]
zupo_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sigmundv__ has quit [Read error: Connection reset by peer]
<pie_>
I can never remember what the right way to do it is. Is there some easy way to remember how to turn a path object into a string without copying into the nix store?
sigmundv__ has joined #nixos
<clever>
> toString ./.
<{^_^}>
"/var/lib/nixbot/state/nixpkgs"
<clever>
> ${./.}
<{^_^}>
error: syntax error, unexpected DOLLAR_CURLY, at (string):271:1
<clever>
> "${./.}"
<{^_^}>
access to path '/var/lib/nixbot/state/nixpkgs' is forbidden in restricted mode
<clever>
pie_: toString will just give the path as a string, casting it will copy
mexisme has joined #nixos
talvdav has joined #nixos
<pie_>
clever: oh ok
<pie_>
clever: casting-copying works by keeping "contexts" around right?
<clever>
pie_: when it makes a copy, the new string will depend on the copy via a context
veske has joined #nixos
<pie_>
clever: what i meant to ask was i guess toString does the cast and drops the context?
<clever>
pie_: context only works for things in /nix/store, so toString doesnt have anything it could add as context
<pie_>
i also mean i guess im failing to parse this unambiguously <clever> pie_: when it makes a copy, the new string will depend on the copy via a context
<clever>
pie_: "${./}" will create a new string, that depends on the copy at /nix/store/hash-foo
<pie_>
aha
<clever>
pie_: but toString ./. just turns the path into a string, without copying, and it cant have context on things outside of the store
<pie_>
ok
<pie_>
so does ${} specifically have some magic for path casts?
<pie_>
id check the source but it would probably take me a while to find it
<keithy[m]>
guys... suggestion, how about making the NixOS options search on the website, a regex?
<clever>
pie_: but you found the other half i didnt link!
<pie_>
ok i should get some breakfast
<pie_>
o/
<clever>
pie_: line 1600, if you try to coerceToString another string, it will return the string, and add the context to the given list
Ariakenom has joined #nixos
<clever>
pie_: line 1605, it will return either the copy or the original string (based on the bool), and only copyPathToStore is able to modify the context
<clever>
pie_: 1610, it will check a set for a .toString function or .outPath value
<clever>
both of which can add more context
CMCDragonkai has joined #nixos
<CMCDragonkai>
It used to be that mounting the `.Xauthority` into a docker container was enough to allow the container to run an X application with the same permissions as the user.
<CMCDragonkai>
However now, that doesn't work, unless I first enable `xhost +local:root`
<CMCDragonkai>
Did something change...
<CMCDragonkai>
?
<clever>
CMCDragonkai: .Xauthority contains the secret password that lets you connect, but you also need to be able to connect to the unix socket at /tmp/.X11-unix/X0
<CMCDragonkai>
clever: I am able to access it, that's what the `--net=host` is for
<CMCDragonkai>
It enables access to the unix domain socket
<clever>
CMCDragonkai: you can also use the xauth command, to export and import records from .Xauthority
<CMCDragonkai>
But even if that were not true
<CMCDragonkai>
I tried this, and it still didn't work
<CMCDragonkai>
the `--net=host` obviates the need for `--volume /tmp/.X11-unix:/tmp/.X11-unix`
<CMCDragonkai>
regardless, I had to set `xhost +local:root` otherwise, the GUI app inside the container wouldn't work
<clever>
CMCDragonkai: what is the listing from inside docker?
<CMCDragonkai>
the container doesn't have anything like xauth
<CMCDragonkai>
unless you're saying to give it that command?
<clever>
CMCDragonkai: you may need to install it within the container first, the hostname may also matter
<clever>
when `xauth list` starts with amd-nixos/unix:0, then that means it will only use that record when connecting to amd-nixos
<CMCDragonkai>
it says `file /.Xauthority does not exist`
<CMCDragonkai>
even though i have this `--volume="$HOME/.Xauthority:/root/.Xauthority:ro"`
<clever>
CMCDragonkai: then $HOME isnt set right, HOME=/root/ in your case
<clever>
CMCDragonkai: its not looking in /root/, is the problem
<CMCDragonkai>
huh
fusion809 has quit [Remote host closed the connection]
<clever>
CMCDragonkai: because $HOME isnt set, X11 apps are looking as /.Xauthority, not /root/.Xauthority
<CMCDragonkai>
that seems seems like it
<CMCDragonkai>
that must have been what is changed
<CMCDragonkai>
i didn't need to do this before with `dockerTools.buildImage`
<CMCDragonkai>
with images built by that nix function
<CMCDragonkai>
and now i do
hyper_ch2 has joined #nixos
sigmundv__ has quit [Ping timeout: 265 seconds]
knupfer has quit [Quit: knupfer]
knupfer has joined #nixos
Izorkin_ has joined #nixos
Izorkin has quit [Ping timeout: 246 seconds]
Izorkin_ is now known as Izorkin
lsix has joined #nixos
knupfer has quit [Remote host closed the connection]
philr has joined #nixos
hyper_ch2 has quit [Remote host closed the connection]
pie_ has joined #nixos
vld has joined #nixos
ng0 has joined #nixos
vld has quit [Ping timeout: 245 seconds]
<danderson>
srhb: (re using nixos test infra for testing other stuff) as long as I can build terrible non-hermetic packages by slurping up freshly built binaries, that would still work out. Not much nixpkgs/nixos code to write to drop a binary onto the machine and run it, before the test script.
<danderson>
really the things I want is the ease of bringing up a non-trivial small network of VMs, with various firewall configs (this is to test NAT traversal tools), and have helpers in test scripts for stuff like "wait for this thing to be running"
<danderson>
for these tests I need to start and configure a bunch of VMs in non-trivial ways, run my service binary on some of the machines, then poke at the binary to see what it makes of the environment. So really, it's almost entirely "make a nixos test", except for the last bit where I'm testing my code, not nixos
o1lo01ol1o has joined #nixos
vika_nezrimaya has joined #nixos
o1lo01ol1o has quit [Ping timeout: 248 seconds]
talvdav has quit [Remote host closed the connection]
<srhb>
danderson: I think that sounds like a reasonable use case for NixOS tests. I mean, inevitably they also test software.
<clever>
even chromium has a test in the nixos framework
jgt2 has joined #nixos
zupo has joined #nixos
jgt1 has quit [Ping timeout: 248 seconds]
Anton43 has joined #nixos
<Anton43>
Hey all! I just installed nix with 'curl https://nixos.org/nix/install | sh' on ubuntu and can't find my nix.conf. In particular, 'find ~ -name 'nix.conf'' returns nothing..
<srhb>
Anton43: It doesn't exist by default iirc
zupo has quit [Client Quit]
<Anton43>
Ah, so I have to create it under .config/nix/nix.conf ?
<Anton43>
And then nix just accepts it?
<srhb>
Anton43: You can see the default values with `nix show-config` and adjust as needed
<srhb>
Anton43: Yeah
<Anton43>
Ah, thanks!
<srhb>
Anton43: (I should rephrase, nix show-config shows your current config, but with no nix.conf it shows the default values)
<{^_^}>
[nixpkgs] @danbst opened pull request #75813 → [RFC] mkDerivation: support version suffix, which is part of derivation nam… → https://git.io/Je7zq
chloekek has joined #nixos
Anton43 has quit [Remote host closed the connection]
lovesegfault has quit [Ping timeout: 252 seconds]
__monty__ has joined #nixos
avn has quit [Remote host closed the connection]
lovesegfault has joined #nixos
lukash_|away is now known as lukash_
<{^_^}>
[nixos-weekly] @domenkozar merged pull request #109 → Add nix-notes, a collection of short notes about Nix, each contributing to the same virtual machine image → https://git.io/Je7Yu
<{^_^}>
[nixos-weekly] @domenkozar pushed 2 commits to master: https://git.io/Je7zl
kraem has joined #nixos
kraem has quit [Client Quit]
hyper_ch2 has joined #nixos
fenedor has joined #nixos
kraem has joined #nixos
dirkx1 has joined #nixos
Syrup has joined #nixos
chiefgoat_ has joined #nixos
m4ts_ has joined #nixos
obey has joined #nixos
cqc_ has joined #nixos
bgupta_ has joined #nixos
lordcirth has joined #nixos
drozdziak1_ has joined #nixos
tetdim_ has joined #nixos
smj has joined #nixos
nixy- has joined #nixos
obey is now known as Guest86177
jeregrine_ has joined #nixos
scoates_ has joined #nixos
etrepum_ has joined #nixos
midchildan_ has joined #nixos
rys has joined #nixos
eddyb[legacy]_ has joined #nixos
xwvvvvwx- has joined #nixos
atriq has joined #nixos
kitemikaze_ has joined #nixos
adamse_ has joined #nixos
Czen11 has joined #nixos
lpsmith_ has joined #nixos
puckipedia has joined #nixos
exarkun_ has joined #nixos
tdeo_ has joined #nixos
fiddlerwoaroof_ has joined #nixos
asheshambasta has joined #nixos
mpickering_ has joined #nixos
qyliss_ has joined #nixos
silver_hook_ has joined #nixos
infinisi1 has joined #nixos
raboof_ has joined #nixos
rys is now known as Guest28246
pasukon_ has joined #nixos
wildsebastian_ has joined #nixos
jbetz_ has joined #nixos
srhb_ has joined #nixos
jackdk_ has joined #nixos
lirzhv_ has joined #nixos
linuus_ has joined #nixos
bigvalen_ has joined #nixos
lemsip has joined #nixos
alanz_ has joined #nixos
djanatyn_ has joined #nixos
georges has joined #nixos
<rizary_>
for deployment using nixops, do I need to add binaryCaches and binaryCachePublicKeys in the server? or using nix-copy-closure instead of nixops is sufficient? my target deployment is digital ocean machine ? (I am using custom nixos image in digital ocean)
veske has quit [Quit: This computer has gone to sleep]
ehmry_ is now known as ehmry
hyper_ch2 has quit [Remote host closed the connection]
p01ar_ is now known as p01ar
tetdim_ is now known as tetdim
lemsip is now known as emilsp
<ocharles>
niksnut: are experimental features passed down through recursive Nix? From my shell, I can do `nix make-content-addressable`, but I can't seem to do that from within a Nix build (even the "outermost" Nix build)
<ocharles>
My main mkDerivation call has `requiredSystemFeatures = [ "recursive-nix" "nix-command" ];`, but calling `nix make-content-addressable` produces: `error: experimental Nix feature 'nix-command' is disabled`
<ocharles>
My /etc/nix/nix.conf has `system-features = benchmark big-parallel kvm nixos-test recursive-nix ca-references nix-command` and `experimental-features = recursive-nix nix-command ca-references`
<DigitalKiwi>
ocharles: sorry for tangent but i was wondering if you've done/are planning more twitch zero to quake? i really liked it :D
lsix has quit [Ping timeout: 246 seconds]
<ocharles>
DigitalKiwi: I'd like to at some point, but no plans yet!
<DigitalKiwi>
oh ok. well you've at least one fan!
noudle has joined #nixos
<ocharles>
Great to hear, thank you!
* DigitalKiwi
still uses doom because of you too lol
<__monty__>
Zero to quake? Implementing quake from scratch on-stream?
<keithy[m]>
My sshd is not enabling is there an easy fix?
h0m1 has joined #nixos
h0m1 has quit [Client Quit]
iyzsong has joined #nixos
h0m1 has joined #nixos
<keithy[m]>
ok, needed lib
<keithy[m]>
nope not goodenough...
<__monty__>
,dnw
<{^_^}>
Provide as much info as possible on problems. What specifically doesn't work? What command did you run? What's the error? This makes debugging easier and helps getting faster responses
o1lo01ol1o has joined #nixos
<keithy[m]>
Ok I am a brand new user, have set nothing of much interest except sshd enable, but it hasnt enabled.
<keithy[m]>
so essentially out of the box sshd enable doesnt appear to work
o1lo01ol1o has quit [Remote host closed the connection]
<simpson>
,sshd
<keithy[m]>
a manual attempt... get an error The unit files have no installation config (WantedBy=, RequiredBy=, Also=,
<clever>
keithy[m]: did you enable ssh in configuration.nix?
<simpson>
Ah, disappointing. Anyway, I'm guessing that you didn't open the port. You need to add port 22 to your firewall.
<keithy[m]>
I am sshing in, but I have to manually start the demon on reboot
<__monty__>
Not in the default installer, no.
hyper_ch2 has joined #nixos
<clever>
keithy[m]: if you remove line 66 of configuration.nix, and try another reboot, what happens?
<clever>
keithy[m]: you will also need to nixos-rebuild switch for the change to take effect
<keithy[m]>
I can see how you guys get to be such good mates, you really need to lean on each other here!
lsix has joined #nixos
Anton43 has joined #nixos
<keithy[m]>
you mean the line I put in to try and fix it!
<simpson>
Part of it is unlearning and comfort. Gotta get used to the simpler, more declarative way of configuring the entire system at once. It takes time.
<clever>
keithy[m]: wait a second, line 66 should just fail hard
<clever>
keithy[m]: thats not a valid option!
<clever>
keithy[m]: what happens if you run `nixos-rebuild switch` ?
<clever>
keithy[m]: yep, because services.sshd.wantedBy isnt a valid option
<clever>
keithy[m]: the stackoverflow said to use systemd.services.sshd.wantedBy
<__monty__>
keithy[m]: Next step is to remove that and run the switch again.
<Anton43>
clever: Okay, KDE shows a "Dummy output: *volume 0 sign*" message, no output from the command
<clever>
keithy[m]: can you pastebin the entire output from `nixos-rebuild switch` before you remove it
<clever>
Anton43: its possible something is already using the alsa device, so pulse cant use it
<niksnut>
ocharles: but it's a good point, the recursive build (at least the daemon side) is currently affected by the host's experimental features which could be seen as an impurity
<__monty__>
Anton43: Maybe you installed nix from the ubuntu repo and it pulled in a dependency that affects sound?
<keithy[m]>
remove what, I did remove that line already
<Anton43>
__monty__ I installed it from the website
<__monty__>
keithy[m]: You removed the line and got an error about it?
<clever>
keithy[m]: what does `nixos-option systemd.services.sshd.wantedBy` output?
<clever>
__monty__: the error was trying to query it with nixos-option
<clever>
keithy[m]: it should just work then, with only enable set
<__monty__>
keithy[m]: Maybe you forgot to nixos-rebuild switch before?
klntsky has quit [Ping timeout: 260 seconds]
<keithy[m]>
hmmm puzzled by happy for now.
<keithy[m]>
* hmmm puzzled but happy for now.
Anton43 has quit [Remote host closed the connection]
klntsky has joined #nixos
Anton43 has joined #nixos
<keithy[m]>
I might have a clue, found "Failed to start Networking Setup." in the journal
<keithy[m]>
so network-setup.service may not be happy
erictapen has joined #nixos
o1lo01ol1o has joined #nixos
<duairc>
I have a server with an encrypted root that I have to SSH into to provide the key via initrd networking. I use networkd, which on recent nixpkgs means I have to disable networking.useDHCP. This broke my initrd networking and I've spent the best part of the last 36 hours trying to figure out why
<duairc>
Firstly, it seemed that my ip= config parameter was never working in the first place and that it was always using DHCP (due to this bug: https://github.com/NixOS/nixpkgs/issues/75314)
<{^_^}>
#75314 (by schmittlauch, 1 week ago, open): boot.initrd.network.enable: documentation links to wrong/outdated ip argument path
<duairc>
The bit that does the DHCP, it calls udhcpc and then passes the values it gets to a script which runs ip commands to set everything up
<duairc>
So I put set -x on that script and looked at dmesg after the boot to see what commands need to be run to get a successful networking setup
<duairc>
Then I replaced the udhcpc script with just those bare commands, but that didn't work
ng0 has quit [Ping timeout: 260 seconds]
<duairc>
After much expensive trial and error (an error costs 10 minutes because I have to reboot the server into rescue mode and log into that over VNC to do a rollback), I discovered that even if I statically run the exact same commands that udhcpc runs in its script, I don't have networking
<duairc>
But if I add this command: udhcpc --quit --now -i eth0 -O staticroutes --script true, then it works.
<duairc>
(true is just the "true" command, as in, it's basically a null script that exits 0)
<duairc>
What on earth could udhcpc be doing that makes it work? Given that it doesn't actually change my ip configuration at all
<clever>
duairc: does the script that --script points to, actually exist within the initrd env?
ng0 has joined #nixos
Anton43 has quit [Remote host closed the connection]
<duairc>
I'm not certain (it's just the "true" command so I think so), but it works regardless
<ocharles>
niksnut: thanks, I'll try that. Want an issue to track the possible impurities?
rardiol has joined #nixos
<duairc>
My understanding is that udhcpc doesn't actually change anything, all it does is get values which it then passes to the script which does the actual ip configuration
<duairc>
But it seems to be doing something in my case, even when I pass it a "null" script, because without it my server never comes up
<duairc>
I've even diffed the results of `ip a` and `ip r` before and after the udhcpc command and there's no difference
<duairc>
Maybe it signals something to the router?
viric has quit [Read error: Connection reset by peer]
<__monty__>
duairc: Maybe it "wants" another service that actually does something?
viric has joined #nixos
gyroninja_ has joined #nixos
<duairc>
__monty__: It's not a systemd service, this is all in the initrd before then. Is that what you mean?
<__monty__>
Yes. Out of my depth.
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
sheenobu_ has quit [Ping timeout: 250 seconds]
sheenobu has joined #nixos
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
emily has joined #nixos
<keithy[m]>
clever: just wanted to say thx
viric has quit [Read error: Connection reset by peer]
viric_ has joined #nixos
<keithy[m]>
has anyone had a go with cowrie ssh honeypot?
emily has quit [Client Quit]
grahamc is now known as gchristensen
viric_ has quit [Read error: Connection reset by peer]
viric has joined #nixos
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
emily has joined #nixos
emily has quit [Client Quit]
emily has joined #nixos
viric_ has joined #nixos
viric has quit [Read error: Connection reset by peer]
orivej has joined #nixos
veske has quit [Quit: This computer has gone to sleep]
<Ariakenom>
is there somewhere I should go to find out how to set up a nix-shell, or similar, for dev?
viric_ has quit [Read error: Connection reset by peer]
hyper_ch2 has quit [Remote host closed the connection]
<__monty__>
Ariakenom: The language infrastructure sections of the nixpkgs manual are good, yes.
<__monty__>
And slap lorri on the todo list to pick up once you've got a nice shell.nix set up : ) (And haskell.nix if haskell's your jam.)
viric_ has joined #nixos
o1lo01ol1o has quit [Ping timeout: 265 seconds]
<Ariakenom>
__monty__: ah yes, the language section. thanks
sigmundv has quit [Ping timeout: 246 seconds]
Chiliparrot has joined #nixos
viric has quit [Ping timeout: 265 seconds]
gxt has quit [Ping timeout: 260 seconds]
gxt has joined #nixos
viric_ has quit [Read error: Connection reset by peer]
viric has joined #nixos
<{^_^}>
[nixpkgs] @filalex77 opened pull request #75817 → tuir: init at 1.28.03 → https://git.io/Je7rN
jgeerds has joined #nixos
Neo-- has joined #nixos
afics has quit [Ping timeout: 246 seconds]
viric_ has joined #nixos
gxt has quit [Ping timeout: 260 seconds]
reanimus has joined #nixos
<DigitalKiwi>
what's haskell.nix
dirkx has joined #nixos
<DigitalKiwi>
lorri++
timmw has joined #nixos
sigmundv has joined #nixos
viric_ has quit [Read error: Connection reset by peer]
viric has quit [Ping timeout: 252 seconds]
dirkx1 has quit [Ping timeout: 248 seconds]
viric has joined #nixos
ztrawhcse is now known as elibrokeit
<michaelpj>
do other people have TMPDIR as /tmp/$USER? if so, do you know how it gets set that way? systemd user services seem to have it unset, which results in them using /tmp
<__monty__>
DigitalKiwi: It's a new approach to the haskell infrastructure.
<__monty__>
michaelpj: It's not set for me.
<michaelpj>
hm
<__monty__>
It is on a mac but to a path in /var/private
<michaelpj>
this is on nixos, I'm assuming for nix-on-other-os it'll depend on the os
<gchristensen>
michaelpj: maybe bash --login -x -c exit 2>&1 > startup-log then look in startup-log for TMPDIR=...
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
<michaelpj>
gchristensen: nothing. It's set in the process environment for my shell (looking in /proc/PID/environ)
<michaelpj>
so it's coming from outside somewhere
<michaelpj>
I can't see anything in a nixos module that sets it
<gchristensen>
how about adding --interactive
veske has quit [Quit: This computer has gone to sleep]
viric has quit [Read error: Connection reset by peer]
viric_ has quit [Read error: Connection reset by peer]
viric has quit [Read error: Connection reset by peer]
civodul has joined #nixos
<Ariakenom>
iirc I was told it didnt work when I asked months ago. interesting
<rajivr___>
Where can I find the source for `nixos-rebuild` derivation?
<grw>
i think problem with this way is that the venv is not registered as gcroot so the python it points to may get garbage collected
hyper_ch2 has quit [Remote host closed the connection]
<blackriversoftwa>
If I have a derivation that has on `(haskellPackages.ghcWithPackages (ps: with ps; [ HaTeX ]))`in the `buildInputs`, and I go into a shell and run `ghc-pkg list`, shouldn't I see `HaTeX`?
viric has joined #nixos
hyper_ch2 has joined #nixos
<Ariakenom>
grw: it doesnt work if I run the command without nix-shell --run
<Ariakenom>
that seems strange
<blackriversoftwa>
(My builder is using haskintex, which calls command-line ghc(i) to evaluate haskell inlined in TeX files)
<grw>
Ariakenom: maybe you have python2 in your $PATH- venv module is not included in python 2
knupfer has joined #nixos
<grw>
can check `which python`
<LnL>
hmm yeah, try adding --pure to the shell invocation
eacameron has joined #nixos
<Ariakenom>
grw: only python 3.6 in path. for some reason, I expected others
<eacameron>
Is it possible to specify "NIX_PATH=nixpkgs=PATH" where PATH has a space in it??
<Ariakenom>
since I have others outside the shell
<LnL>
could be trying to make a python3 venv while it's a python2 shell
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
<hodapp>
raaargh, my last two reboots/shutdowns have been alongside something like "Failed to talk to init daemon" to where halt/shutdown both were failing
<hodapp>
can't remember if I tried 'reboot -f' or not, as I'm now reading online should be able to bypass the init system
<DigitalKiwi>
hodapp: if you figure it out let me know i think i've been having the same
gxt has joined #nixos
<hodapp>
DigitalKiwi: what sort of machine?
<DigitalKiwi>
macbookpro11,3
<hodapp>
ahh, Thinkpad T410 here
<hodapp>
it looks like it happened after a suspend
<hodapp>
was seeing things like: "traps: xfsm-shutdown-h[17603] general protection ip:7f8bdb778311 sp:7fff79d0dec0 error:0 in libc-2.27.so[7f8bdb718000+13d000]"
<hodapp>
the first time it looked like systemd was just completely out to lunch
cosimone has joined #nixos
<DigitalKiwi>
less recently i was having it take several minutes to do a shutdown/reboot, and then more recently i get kernel panics for restarting display-manager, and the same for reboot/poweroff, forcing me to hold the power button after yelling something that i should probably take a picture of
<hodapp>
huh
<evils>
hodapp: T410 user here, don't think i've seen that issue before
<DigitalKiwi>
also had a ton of problems with wifi ~19.09 release (i'm on nixos-unstable though)
cosimone has quit [Client Quit]
viric has quit [Read error: Connection reset by peer]
<Ariakenom>
if I use --system-site-packages with virtualenv, will those system packages point to the ones I installed in my shell.nix?
<evils>
hodapp: wouldn't happen to know how to get tp_smapi working on this thing? and how to prevent the lid from suspending it?
viric has joined #nixos
seanparsons has quit [Ping timeout: 245 seconds]
<hodapp>
didn't even know tp_smapi was a thing >_>
<hodapp>
I used this laptop almost solely at a desk with a docking station
ixxie has joined #nixos
viric has quit [Read error: Connection reset by peer]
viric_ has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
viric_ has quit [Read error: Connection reset by peer]
<evils>
hodapp: acpi seems to have the info i want, not sure if that's dependent on tp_smapi xD
stites has joined #nixos
ixxie has joined #nixos
<stites>
Hey all, I recently had a hardware failure when building a nix derivation and the sqlite.db got corrupt
<stites>
the db is fried, so I'm wondering how one goes about rebuilding it
halfbit has joined #nixos
<jared-w>
new fun project of mine. Setting up a dell precision 7740 with nixos + encryption. Surprisingly annoying so far
<stites>
So far, I've moved the db and attempted to run nix-store --init (from https://github.com/NixOS/nix/issues/3091 ) but I just get "cannot open Nix database '/nix/var/nix/db/db.sqlite'"
<{^_^}>
nix#3091 (by bhipple, 13 weeks ago, open): Rebuild sqlite db from scratch?
<jared-w>
stites: I've wondered this myself; I ended up nuking everything and restarting
<jared-w>
(well, my particular thing is I wanted to completely nuke the entire /nix/store and pretend like it was a clean install without actually formatting stuff)
<duairc>
Is there a way of doing something like nixos-rebuild test, but like, automatically reverting after a minute unless instructed otherwise? I'm messing around with networking at the moment and every time I mess up I have to reboot my server because SSH goes down
<jared-w>
duairc: there's a dry-run command of some sort; would that do it?
<stites>
duairc: I use mosh, which reconnects automatically if ssh goes down (since the mosh-server is still in the background)
<duairc>
stites: No, I mean, if I mess up the network configuration neither mosh nor ssh nor anything work
<duairc>
jared-w: That's a bit better, but I do want to actually run it
fusion809 has joined #nixos
<stites>
ah, sorry i meant "if networking services are rebooted"
<stites>
but I guess it doesn't fit the bill
<duairc>
I guess I mean, what's the best way to revert to a specifc configuration? As far as I know nixos-rebuild switch --rollback will switch to the previous generation, not the current one if being run inside a nixos-switch test
<duairc>
Like ideally I could just do nixos-rebuild test && sleep 60 && nixos-rebuild untest
<stites>
oh, also, isn't there a vm-build command for testing out nixos builds without interfering with your setup?
ixxie has quit [Ping timeout: 268 seconds]
<duairc>
There is, but I don't think will be much good for testing networking because I want to test if the configuration does what I need it to do on the real network with the real network interfaces, not the virtual network
erictapen has quit [Ping timeout: 268 seconds]
m4ts_ has quit [Quit: bye]
m4ts has joined #nixos
gchristensen has quit [Ping timeout: 248 seconds]
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
o1lo01ol1o has quit [Ping timeout: 245 seconds]
{^_^} has quit [Ping timeout: 246 seconds]
o1lo01ol1o has joined #nixos
gchristensen has joined #nixos
<stites>
last shot: so its not possible just to nixos-rebuild build and inspect the binaries? or are you messing around with the network-stack itself instead of just networking?
<stites>
jared-w: when you say "nuking everything" do you mean overwrite your OS?
viric has quit [Read error: Connection reset by peer]
<stites>
or just reinstall nix on nixos?
viric has joined #nixos
<jared-w>
I just remembered; nixos-rebuild is a shortcut for a few commands
<jared-w>
You can do it manually and do nixos-rebuild build which will output a /nix/store/your-new-system result symlink
<jared-w>
so just chroot into the result
cosimone has joined #nixos
<jared-w>
(I say "just" because I haven't figured out how to nicely chrooot in nixos yet; it keeps barfing on missing /run/current-system/* stuff)
cosimone has quit [Client Quit]
codedmart_ is now known as codedmart
<jared-w>
stites: essentially just reinstall nix on nixos. `rm -rf /nix && "curl install nix"` but without breaking everything
pigeonilla has joined #nixos
<DigitalKiwi>
duairc: could you use a screen session nixos-rebuild switch sleep and --rollback ?
viric has quit [Read error: Connection reset by peer]
<pigeonilla>
Hi, because Hydra is super slow for me I ask here: does dovecot-pigeonhole compiles for you on master? It looks like someone updated it to 0.5.8 but without checking whether it compiles.
<ivan->
what's the actual attribute
<DigitalKiwi>
duairc: and alternatively. use a screen session, and nixos-rebuild test -I nixpkgs=~/test/config.nix && sleep 60 && nixos-rebuild test
<ivan->
oh dovecot_pigeonhole
<DigitalKiwi>
screen or tmux if that's your poison
Profpatsch has joined #nixos
<Profpatsch>
Do I have to do something special so that a systemd service is started when it’s added? I set wantedBy = [ "default.target" ];
<Profpatsch>
And Restart = "always"
juxiemaotu has joined #nixos
<infinisil>
Profpatsch: That should work, is it the default Type?
<ivan->
pigeonilla: /nix/store/m8ih8gw7qnd2c7p0qgv3cn4b6abkbanb-binutils-2.31.1/bin/ld: ../../src/lib-sieve/.libs/libdovecot-sieve.so: undefined reference to `mailbox_alloc_delivery'
viric has joined #nixos
<infinisil>
Profpatsch: Usually people use `multi-user.target`, but that should be the same as default.target
werner291 has joined #nixos
<duairc>
DigitalKiwi: ah, yes, something like that could work, thanks
<Profpatsch>
infinisil: Oh, hm. AWS image, so not sure.
<juxiemaotu>
i wanna to install nixos on cloud host, it need nixos image format raw/vhd/qcow2/vmdk,but nixos homepage where i only can get iso file. what shuld i do to get that format file?
<DigitalKiwi>
duairc: oh and nixpkgs=~/test/config probably isn't the one you want is nixos-config=
* DigitalKiwi
uses -I nixpkgs=~/mygitclone for PRs so much it's just the one I'm used to typing
<duairc>
Yeah, I figured that much :)
<DigitalKiwi>
ok cool :D hope that works
viric has quit [Read error: Connection reset by peer]
tsrt^ has joined #nixos
viric has joined #nixos
<duairc>
DigitalKiwi++
<infinisil>
gchristensen: {^_^} not present
viric has quit [Read error: Connection reset by peer]
<gchristensen>
aye, network troubles in my datacenter
Sargun has joined #nixos
Sargun has quit [Changing host]
{^_^} has joined #nixos
<{^_^}>
[nix] @edolstra pushed to master « Add priority setting to stores »: https://git.io/Je71L
<Profpatsch>
infinisil: Does `default.target` have to be listed in systemctl if it exists? Because I’m only seeing multi-user.target
rauno has joined #nixos
<pigeonilla>
Thank you ivan- it sucks folks bump packages without even checking.
<infinisil>
Profpatsch: From `man systemd.special`: "The default unit systemd starts at bootup. Usually, this should be aliased (symlinked) to multi-user.target or graphical.target"
<Profpatsch>
infinisil: “usually”, “should be”
viric has joined #nixos
<stites>
oh, hey, cool! zfs saves the day jared-w
<infinisil>
should be graphical.target for user units, multi-user.target for systemwide
<Profpatsch>
That’s the specificity I’ve come to expect from systemd :)
<infinisil>
Profpatsch: The following paragraph mentions that it can be changed
<infinisil>
So I think there's nothing wrong with the terms used
viric has quit [Read error: Connection reset by peer]
<stites>
my first snapshot was also corrupt, but it looks like just had to roll things back a bit further (I was about to try pulling in a remote backup)
pigeonilla has quit [Remote host closed the connection]
<Profpatsch>
infinisil: cool, multi-user.target seems to do the trick
<gchristensen>
I'm trying to debug a crashwhich I can't easily reproduce, where the crashing program didn't have debugging symbols enabled nor separate debug symbols. is there a way to get symbols after the fact?
sigmundv has quit [Ping timeout: 245 seconds]
<infinisil>
I very much doubt it
<gchristensen>
me too :|
<DigitalKiwi>
I was robbed of karma D:
<DigitalKiwi>
tell your datacenter I am very displeased
Ariakenom has quit [Quit: WeeChat 2.6]
<gchristensen>
is there a reason we don't debug symbols for everything?
<DigitalKiwi>
and also it's funny everytime I'm robbed given my name :<
jaeckel has joined #nixos
<DigitalKiwi>
do debug symbols bloat/slow down binaries?
Profpatsch has left #nixos ["WeeChat 0.4.3"]
orivej has quit [Ping timeout: 265 seconds]
<evils>
gchristensen: would giving gdb another binary that's not stripped, and the original coredump, work?
viric has joined #nixos
<evils>
DigitalKiwi: i've got a binary here of 128K, after running `strip`, it's 35K
viric has quit [Read error: Connection reset by peer]
<DigitalKiwi>
that seems like a a reason people might want them stripped then heh
<evils>
kicad unstripped, 84M, stripped, 2M
<DigitalKiwi>
now the final test; does it crash faster stripped
<jared-w>
stites: nice! I've used btrfs for years on my old laptop and never made a single snapshot
<jared-w>
Finally wised up and stopped doing that, only to realize that I picked nixos to stop using btrfs when a inode-less FS is a compelling advantage to nixos. So, bah
<stites>
jarend-w: if you're every feeling like taking a leap of faith, I know that there are some services to toggle autosnappshotting for zfs (which is actually pretty pleasent to set up on nixos). : D
viric has quit [Read error: Connection reset by peer]
viric has quit [Read error: Connection reset by peer]
<DigitalKiwi>
(my fork of it enabled encryption differently than adam did i think if you want to look at that too it may be up...if not out of date)
<DigitalKiwi>
it saved me a *lot* of time
ixxie has joined #nixos
viric has joined #nixos
<jD91mZM2>
Hi! Is there an easy way to find out why "warning: error(s) occurred while switching to the new configuration" is printed? I seem to get that after each initial `nixos-rebuild`, but then when I try again it always goes away. I don't see anything suspicious in `systemctl status`
<betaboon>
could i use hydra to deploy nixops deployments? (only when all builds required for that deployment succeeded)
phreedom has quit [Remote host closed the connection]
viric has quit [Read error: Connection reset by peer]
phreedom has joined #nixos
Ariakenom has joined #nixos
<NemesisD>
is it possible given a nix file and an attribute name to get a list of all dependencies, sub dependencies, and so on that go into that derivation? preferably a tree. i need upgrade a package that i don't depend on directly and i don't know what is asking for it
viric has joined #nixos
justanotheruser has quit [Quit: WeeChat 2.6]
<pbb>
NemesisD: runtime or build-time dependencies?
<{^_^}>
[nixpkgs] @grahamc opened pull request #75825 → slack: add xdg_utils to the PATH → https://git.io/Je7Mf
<evils>
NemesisD: could you be looking for `nix why-depends`?
justanotheruser has joined #nixos
<jared-w>
stites: sounds fun. I'm trying out bcachefs because why not
mcfrank has left #nixos ["WeeChat 1.9.1"]
viric_ has joined #nixos
<{^_^}>
[nixpkgs] @jonringer closed pull request #69382 → dotnet-runtime{,_2}: init at 3 and 2.2.7 → https://git.io/JeGtW
<jared-w>
speaking of bcachefs, linux_latest is 5.4.3 but linux_testing_bcachefs is only 5.2? Is that for a reason or?
viric has quit [Read error: Connection reset by peer]
Profpatsch has joined #nixos
erasmas has joined #nixos
<stites>
say, what is the difference between nativebuildinputs and buildinputs again? is the former just for compile-time and the latter is for run-time?
klimi has joined #nixos
<Profpatsch>
Is nix-store -qR sorted topologically?
<gchristensen>
pretty sure yes
is_null has quit [Ping timeout: 265 seconds]
<betaboon>
any suggestion on how to update nixops deployments following a hydra build ?
viric has joined #nixos
is_null has joined #nixos
<jD91mZM2>
stites: nativeBuildInputs is for stuff that runs on the machine compiling the code (the native machine when cross-compiling), so yes compile time. buildInputs is the machine that runs the code, so yes runtime
<bennofs_>
stites: seems about right. nativeBuildInputs is for things that need to be executed during build, so they need to be compiled for the same architecture that the build is performed on (this matters when cross compiling)
endformationage has joined #nixos
viric has quit [Read error: Connection reset by peer]
lukash_ is now known as lukash_|away
<klimi>
Do you guys use home-manager and manage your keyboard layout there or where do you manage your layout?
viric has joined #nixos
<jared-w>
klimi: if I want it managed systemwide I set it in /etc/nixos/configuration.nix; if I want it per-user I put it in home-manager
<jared-w>
although root can use home-manager too, so I'll probably eventually migrate quite a bit of stuff over to home-manager
<Profpatsch>
So has anybody figured out a one-liner to display the transitive sizes of all transitive dependencies of a store path, sorted topologically?
<Profpatsch>
Something like ncdu, but for nix store paths
<gchristensen>
sounsd cute, Profpatsch
<klimi>
jared-w i just put it into home manager... then i changed it... and the configuration hasn't changed at all so i am confused. yes i switched the generation
bvdw has quit [Read error: Connection reset by peer]
bvdw has joined #nixos
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
<LnL>
Profpatsch: what about nix path-info -rhS?
<genesis>
gchristensen : the best test language for ofborg is nix isn't it ? i can start some nix hook for test some ideas and discuss later how integrate this in the bot ?
<gchristensen>
genesis: w.r.t. testing Desktop files, the proper place to add this test is in Nixpkgs itself, not ofborg
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
<genesis>
oki, like a apps wrapper i guess
<gchristensen>
yeah
<genesis>
i've to write some hook i guess.
<gchristensen>
exactly :)
<genesis>
let have a try.
stites has quit [Remote host closed the connection]
<jared-w>
I figured. Is it set in the /etc/nixos/configuration.nix? I don't think that would overwrite it but I'm not sure; I haven't set keyboard layout in home-manager myself
<klimi>
It's not set and home-manager should overwrite it
<gchristensen>
IMO dmenu/etc. programs should get a different path than my user path. the yhave no business running coreutils programs, so why grant them the authority to
<gchristensen>
also, we should probably get all the `.wrapped*` programs out of the PATH
gyroninja__ has joined #nixos
gyroninja_ has quit [Ping timeout: 245 seconds]
<qyliss>
gchristensen: you perhaps want j4-dmenu-desktop
chloekek has joined #nixos
<duairc>
This isn't NixOS specific, but I've been trying to create a bridge interface to eth0 on my server, and I've found that the only way to make it routable is by explicitly setting the bridge interface to have the same MAC address as eth0. I don't really know much about networking. Does that sound like a reasonable thing to do?
<gchristensen>
seems good, though I'd favor revoke the ability for dmenu to execute the other ones :)
<drakonis>
that's a fairly good argument for a permission system
<equivrel>
did builtins.fetchgitPrivate disappear in a recent update?
atlas_ is now known as atlvs
<gchristensen>
fetchgitPRivate was never a builtin, you should just use builtins.fetchgit
atlvs is now known as atlas_
<equivrel>
gchristensen: ah, I see
tdeo_ has quit [Quit: Quit]
tdeo has joined #nixos
tdeo has quit [Changing host]
tdeo has joined #nixos
<NoctisLabs>
A package is failing to compile and stopping an upgrade, how can I keep the current working version of that package and still upgrade? (Usually I would just remove the broken package but this one is important to me)
mexisme has joined #nixos
halfbit has quit [Ping timeout: 252 seconds]
viric_ has joined #nixos
viric has quit [Ping timeout: 252 seconds]
viric_ has quit [Read error: Connection reset by peer]
viric has quit [Read error: Connection reset by peer]
viric has joined #nixos
<equivrel>
will specifying all of rev, ref, sha256 in builtins.fetchGit help with not redownloading something that has already been fetched?
mexisme has quit [Ping timeout: 268 seconds]
<jared-w>
NoctisLabs: I think overriding with a pinned version is the most straightforward? Idk what "allowBroken" does but that might also solve your problem
<infinisil>
jared-w: allowBroken is only for packages that are explicitly marked as broken with `meta.broken = true`, which would throw an error pointing this out when somebody tries to build it
<infinisil>
(but it would try to build it anyways with `allowBroken = true`)
<rembo10>
I just updated to the latest nixos-unstable and I'm having some trouble with libvirt. I'm getting this 'Failed to initialize a valid firewall backend' error. I looked through the latest commits and couldn't really pinpoint what may have changed. I found some info relating to this, that installing ebtables and dnsmasq seem to fix this - but I saw that libvirt already pulls those packages in, so I'm not
<evanjs>
How might I copy out all libraries from an arbitrary drv in an installPhase? e.g. `cp "${pkgs.libwebsockets}/lib/*so*" $out/runtime/x86_64/debug/lib`
<rembo10>
really sure what's going on
<waleee-cl>
juxiemaotu: if you used the last I wrote?
<evanjs>
But that isn't working
<betaboon>
evanjs: shouldnt that be `*.so` ?
<waleee-cl>
some libs have the pesky standard of putting version numbers after .so, eg *.so.6
<evanjs>
betaboon: yah that’s what I tried initially but no worky still
<evanjs>
waleee-cl: yah I was trying to account for all those. And libwebsockets is definitely one of those libraries
<juxiemaotu>
waleee-cl: when use pip3 it work, but pip cannot
wildtrees has joined #nixos
tbenst_ has quit [Ping timeout: 268 seconds]
<waleee-cl>
juxiemaotu: Do you have pip2 installed?
<juxiemaotu>
waleee-cl: i have pip2 and pip3
<waleee-cl>
juxiemaotu: ok, that may be one of the things they added in pip3
<evanjs>
Basically I'm just trying to copy over lws libs when sensor_tester is enabled (target is a gentoo system that might or might not have the req libs, so put them in the lib folder we place our other libs for the app in)
<waleee-cl>
juxiemaotu: python3X.withPackages is what I use instead of using pip manually
<juxiemaotu>
i had try to solve prombles about python used in nixos on manual,but less improve
<evanjs>
man I'm excited about getting this to a nixops or similar-managed system someday
the_pumpkin_man[ has joined #nixos
<juxiemaotu>
ok , i will try about python3x follow you , help it work well
<waleee-cl>
juxiemaotu: substitute the 'x' with the minor version(s) in the nixpkgs repo atm
<evanjs>
I could probably explicitly copy the libs... but I'd rather glob if possible
<juxiemaotu>
waleee-cl: any example?
<juxiemaotu>
waleee-cl: about how to use your method in nixos
<the_pumpkin_man[>
(sorry for the spam, Im also new to IRC and dont know if there are spoilers)
nuncanada has joined #nixos
<evanjs>
the_pumpkin_man[: don't worry about that, it cuts off long messages and gives us links to the full messages via matrix
<the_pumpkin_man[>
Ok
<evanjs>
the_pumpkin_man[: of course, often times, pastebin or etc is preferrable for longer code snippets or whatever. at least compared to screenshots and etc :P
<juxiemaotu>
waleee-cl: for example, i wanna install python-language-server on my nixos machine, what should i do
<waleee-cl>
juxiemaotu: I had apparently changed it in my configuration.nix to python3Full at some point, but otherwise, in eg. environment.systemPackages , pkgs.python37.withPackages (ps: with ps; [ numpy ]) should give you a python3.7 with numpy
<waleee-cl>
juxiemaotu: add python-language-server in the [ ... ] and you should probably be peachy. (Although it looked like it might not be available for python3 in nixos-19.09)
<waleee-cl>
oh, it was
<juxiemaotu>
waleee-cl: you mean ,add it in configuration.nix in environment.systemPackages use pkgs.python37.xxxx, and then run nixos-rebuild ? do this when i need a python moudle?
<waleee-cl>
or use a nix-shell if you only want it temporary
<waleee-cl>
or in a default.nix if you want to use nix build
erictapen has joined #nixos
* waleee-cl
or in ~/.config/nixpkgs/config.nix (the python manua
* waleee-cl
or in ~/.config/nixpkgs/config.nix (the python manua
<waleee-cl>
* or in ~/.config/nixpkgs/config.nix if you want to define a python package that you can install using nix-env -iA myPythonEnvironment
<waleee-cl>
the manual section on python lists multiple ways
nornagon2 is now known as nornagon
<betaboon>
evanjs: remove the quotes around `${pkgs.libwebsockets}/.....`
<evanjs>
omg...
<betaboon>
evanjs: and btw in the cases where you use `if sensor_tester == true ...` you could use `pkgs.lib.optional` `optionals` and `optionalString`
<evanjs>
betaboon: totally forgot about optional(s) yeah, thank you
<evanjs>
Is there a section in the manual or wherever that details that functionality between e.g. "${}" and ${}?
<evanjs>
I'd assume it's basically just path vs a string?
<juxiemaotu>
waleee-cl: thank you , i got it.
<betaboon>
evanjs: that doesnt have anything to do with nix. thats how bash globs work
<betaboon>
evanjs: globbing doesnt work in quoted strings
<betaboon>
evanjs: in your installPhase you could use `mkdir -p $out/{lib,bin}` also XD
<evanjs>
betaboon: derp. Yeah just normal shell globbing
<dooms>
betaboon: nice
<evanjs>
betaboon: right right, don't worry this has some heavy refactoring in store anyway
<evanjs>
dooms knows haha
werner292 has joined #nixos
<betaboon>
evanjs: no shame there. sometimes happens to me to. its like forgetting a semicolon. you just dont notice because everything surrounding that stupid mistake is so complicated xD
domogled has joined #nixos
Neo-- has quit [Ping timeout: 248 seconds]
demize has quit [Quit: Do Androids Dream of Electric Sheep?]
<evanjs>
betaboon: can't help when python and especially TS is used at work, either. One doesn't need semicolons... the other sometimes needs them, but sometimes doesn't (care)...
werner291 has quit [Ping timeout: 268 seconds]
werner292 is now known as werner291
<betaboon>
evanjs: yeah thats what you get from jumping between multiple languages xD
demize has joined #nixos
<evanjs>
betaboon: yep. Can't do anything about it. Almost as bad as going from Rust to other languages (at the very least, in terms of linting/error messages)
<sshow>
Trying to setup a wireguard server. Unit-script fails with "fopen: No such file or directory". Any experience with this error?
imdoor has joined #nixos
zeta_0 has joined #nixos
<sshow>
I am running a custom kernel (latest_hardened), but it does not seem to make a difference whether I include the wireguard kernel module explicitly or not
thc202 has quit [Ping timeout: 248 seconds]
<zeta_0>
sorry for being a little off topic, but i installed the nix package: `protonvpn-cli`, and when i try to initialize, it prompts me for a OpenVPN username and password? do i need to use the my username of and password that i use to sign into the protonvpn website, or do i need to create a OpenVPN account?
<sshow>
the private key did not exist. Way to go debugging :D
<{^_^}>
[nixpkgs] @matthewbauer pushed 2 commits to master: https://git.io/Je794
psyanticy has quit [Quit: Connection closed for inactivity]
MightyJoe has joined #nixos
<{^_^}>
[nixpkgs] @pasqui23 opened pull request #75832 → nodePackages.dat :init at 13.13.1 → https://git.io/Je79H
cmacrae has joined #nixos
cyraxjoe has quit [Ping timeout: 248 seconds]
<cmacrae>
Hee people! o/ I'm getting "attempt to call something which is not a function but a set" on something I've built out. I'm pretty fresh on modules, so I must be doing something silly. I could really use a hand working it out. I've chucked up an example NixOps structure here: https://git.io/Je79Q
<cmacrae>
Any help is greatly appreciated! I mustn't be using 'options' and 'config' correctly :(
mexisme has joined #nixos
<erictapen>
cmacrae: hey, with the redacted lines the error message with the line number is pretty much useless. could you tell which line corresponds to line 61 in some-module.nix?
cosimone has joined #nixos
<cmacrae>
erictapen: Hello! Yeah that's fair enough - perhaps I should just share the real code
<infinisil>
cmacrae: Some comments: For modules, always use `{ pkgs, lib, ... }` with all the arguments you need, instead of not using "...". You can't use `//` to merge configs, you need mkMerge instead. Usually it's `import <nixpkgs> {}` not <nixos>
<cmacrae>
Thank you!
<jared-w>
yeah, go for it; if you're concerned about it, using a private gist and then deleting it after you get help is pretty safe as far as things go. Technically this server's chat history is logged publicly but people won't be able to view the gist after it's deleted so
<infinisil>
jared-w: Anything ever on the internet should be assumed to not be secret anymore
<exarkun>
that is weird advice, yea
<exarkun>
maybe it's fine but it depends what's being kept secret and why
<jared-w>
infinisil: true, but if they redacted something and then were fine with making it public it probably wasn't something that was _super_ secret, so I don't think it was that out of line
<exarkun>
the best case, anyway, is to create a self-contained, complete, minimal example that reproduces the problem
<infinisil>
^^
<cmacrae>
Thanks for the advice so far guys - it's nothing super secret! Just my home network config I'm in the middle of building out :)
<cmacrae>
Let me just fix stuff up with the advice so far, then I'll share
dirkx_ has joined #nixos
philr has joined #nixos
iqubic has joined #nixos
<iqubic>
I'd like to delete all of my zfs snapshots, if possible.
<iqubic>
How would I go about doing that?
<infinisil>
iqubic: Doesn't sound Nix specific, I can answer in #zfsonlinux
<jared-w>
I am both unsurprised that #zfsonlinux is a thing and that you're in it, infinisil :p
<infinisil>
jared-w: Only there when I need to, so I just joined now for this :)
<jared-w>
nice
<jared-w>
hmm. Really hoping my laptop is finally stable. I've been reinstalling nixos on it for like 12 hours now and it's getting ridiculous
<jared-w>
I think I've finally gotten it to realize I'm more stubborn than it and now it's slowly cooperating
<cmacrae>
Oh! I think the call to mkMerge instead of using // worked :D I've got some other problem now, but it looks like Python (so, NixOps)
<cmacrae>
lovesegfault has joined #nixos
<cmacrae>
Thanks for the pointers infinisil++
<{^_^}>
infinisil's karma got increased to 169
<infinisil>
Nice :)
asheshambasta has quit [Ping timeout: 246 seconds]
iqubic has quit [Remote host closed the connection]
<jared-w>
yehhh it rebooted twice in a row. #success
<infinisil>
jared-w: What was the problem?
<jared-w>
It's a dell precision 7740 and I needed an encrypted hard drive. I don't think the encryption was the problem, but it certainly wasn't terribly happy about the GPU it seems like
<jared-w>
infinisil: _still_ can't get it to recognize any plugged in monitors though, ugh
<infinisil>
jared-w: Do they not show up in xrandr?
dirkx_ has quit [Quit: dirkx_]
CMCDragonkai has quit [Quit: Connection closed for inactivity]
<jared-w>
infinisil: I've blacklisted i915 (intel iGPU) and set "amdgpu" to the boot.initrd.kernelModules so that it's there
<jared-w>
xrandr only shows one monitor (the internal laptop monitor)
<jared-w>
xrandr --listproviders shows two providers, and lspci | grep VGA shows an intel device and the AMD dGPU
dirkx_ has joined #nixos
<jared-w>
so blacklisting the i915 doesn't seem to actually be blacklisting it...
selfsymmetric-pa has joined #nixos
<selfsymmetric-pa>
Hi! What does `sw` stand for in `/run/current-system/sw/`?
<Yaniel>
software?
<simpson>
switched?
<Yaniel>
or system-wide
<selfsymmetric-pa>
ah, shame, I thought it was "scotch whisky"
<jared-w>
superrifically wonderful?
<selfsymmetric-pa>
Secretary of War?
<jared-w>
"ah, that bash, what a superrifically wonderful piece of software"
<selfsymmetric-pa>
that's why we don't put it in the /bin/ like everyone else
<selfsymmetric-pa>
it deserves better
o1lo01ol_ has joined #nixos
hmpffff has quit [Quit: nchrrrr…]
o1lo01ol1o has quit [Ping timeout: 246 seconds]
hmpffff has joined #nixos
hmpffff has quit [Client Quit]
hmpffff has joined #nixos
hmpffff has quit [Client Quit]
hmpffff has joined #nixos
hmpffff has quit [Client Quit]
selfsymmetric-pa has quit [Remote host closed the connection]
selfsymmetric-pa has joined #nixos
o1lo01ol_ has quit [Remote host closed the connection]
ivan- is now known as ivan
lordcirth has quit [Remote host closed the connection]
selfsymmetric-pa has quit [Remote host closed the connection]
selfsymmetric-pa has joined #nixos
v0|d has joined #nixos
o1lo01ol1o has joined #nixos
<infinisil>
Yaniel: Oh system-wide sounds nice, I always thought of it as "software"
<infinisil>
Hmm
<infinisil>
Though "software" makes more sense I think
<infinisil>
Or not
lovesegfault has quit [Ping timeout: 265 seconds]
selfsymmetric-pa has quit [Remote host closed the connection]
selfsymmetric-pa has joined #nixos
<kraem>
is the nix daemon building packages in parallel by default?
lopsided98_ has quit [Remote host closed the connection]
<gchristensen>
yes
o1lo01ol1o has quit [Ping timeout: 252 seconds]
lopsided98 has joined #nixos
<kraem>
any way i can figure out which package is failing to build with that option on? or would i disable that feature somehow to see what build is failing?
<gchristensen>
oh an individual package in parallel?
o1lo01ol1o has joined #nixos
<gchristensen>
`nix show-config | grep cores` will tell you. 0 means all cores, N means N cores
<tilpner_>
kraem: I sometimes add -j1 to keep it from building things in parallel
<infinisil>
kraem: I usually do --max-jobs=1 to deug
knupfer has quit [Ping timeout: 248 seconds]
<infinisil>
Hehe, three answers, three different names for parallelism, cores, j and max-jobs
<gchristensen>
I'm pretty sure max-jobs and -j isn't the thing kraem is asking about
<infinisil>
I'm pretty sure it is
<gchristensen>
why would building 2 packages independently cause one to fail?
<kraem>
gchristensen: actually it is :)
<kraem>
thanks!
<tilpner_>
gchristensen: It doesn't, it's just easier to tell if output is not interleaved
<gchristensen>
then I must know the answer to my question! this is a very serious concern!
<gchristensen>
ooh I see, the problem of output
<gchristensen>
kraem: usually I'll run the build with --keep-going, and then a second time with -j1
tilpner_ is now known as tilpner
<gchristensen>
otherwise you might end up building a ton of packages in the build graph before getting to the one which failed
<kraem>
yeah it was hard telling from the logs which one was failing. but i figured it out. it was idea-ultimate :)