<energizer>
dgpratt: i prefer to avoid nix-env altogether, in favor of configuration files
<energizer>
take a look at home-manager for example
<dgpratt>
thanks, I've seen it energizer -- at present it's just one more thing I don't want to try to understand just at the moment, but I expect to get there sooner or later
<ldlework>
If it's a problem with my configuration, how am I supposed to figure out what to fix? The traceback is not helpful.
<bqv>
if you want the latest nix from master, build it yourself
<bqv>
nixpkgs will not help you there
<multun>
ldlework: it's probably a bug, nixos unstable is very unstable
<ldlework>
multun: how do i roll back my channel to whatever i was at before?
<cole-h>
man nix-channel
<multun>
I _think_ sudo nix-channel --rollback
<cole-h>
There's a --rollback flag, yep ^
<ldlework>
oh nice
<simpson>
Is there a lib.license for GitHub's default rights granted to repositories without a COPYING?
<cole-h>
lib.licences.unfree
<cole-h>
You should ask them to license their work.
<simpson>
There's an outstanding issue for it. It's a single file of Go that hasn't been altered in a few years, and indeed it probably would be rewritten in bash before heading to nixpkgs.
<simpson>
I have a situation where avahi-daemon is not reloading after live-switching generations with nixos-rebuild. I changed services.avahi.extraServiceFiles to add a new service file, which shows up as an XML file in /etc, but avahi-daemon needs a SIGHUP to reload those files and it's not getting it.
<simpson>
What can I read to learn about how to properly send the signal and reload the daemon? Alternatively, is there a cruder way to just make changing that setting force a reload?
<bqv>
guess i'll settle for downgrading to ghc882 then
<bqv>
with the amount of people that complain about the idea of things not being in nixpkgs, i'm surprised hie managed to not be in nixpkgs
<quinn>
simpson: might not be the answer you're looking for but a service i use (asterisk) has live reload off, presumably because it has smoother live-reload functionality built into it. might be the same story with avahi
<infinisil>
simpson: You probably want systemd.services.<name>.reloadIfChanged
<simpson>
quinn: It doesn't appear so, but I'll consider that, thanks.
jumper149 has quit [Quit: WeeChat 2.8]
<simpson>
infinisil: That sounds promising. I'll see if that improves things, and make a PR if appropriate.
<infinisil>
Oh also specify `ExecReload = "${pkgs.coreutils}/bin/kill -HUP $MAINPID";`
<quinn>
infinisil: is that just for the service file changing or will it catch other changes?
<infinisil>
quinn: Only for the service file changing pretty sure
<infinisil>
Oh, maybe systemd.services.<name>.restartIfChanged needs to be disabled as well for this to work
<infinisil>
(it's turned on by default)
<infinisil>
quinn: You can add additional trigger files to systemd.services.<name>.restartTriggers though
<infinisil>
(which underneath just inserts a dummy entry into the service file)
rardiol_ has quit [Read error: Connection reset by peer]
<multun>
I remember coming across a website that compared the size and freshness of various package repositories
<quinn>
infinisil: that's cool. and if config file gets changed the service file will get changed cause config file path changes when it changes too
<quinn>
multun: repology
<multun>
quinn: awesome thanks!
<quinn>
:)
<multun>
ohoh
<multun>
nix's getting close to aur
<infinisil>
In terms of?
<multun>
number of packages available
<multun>
it's probably way over it terms of packages that actually build
rardiol has joined #nixos
bdju has quit [Read error: Connection reset by peer]
bdju has joined #nixos
drakonis has quit [Ping timeout: 260 seconds]
ShaRose has quit [Quit: I appear to have left for some reason.]
<{^_^}>
[nixops] @cole-h opened pull request #1371 → default.nix: don't use builtins.fetchGit → https://git.io/JfH9A
iqubic` has joined #nixos
iqubic` has left #nixos [#nixos]
iqubic has quit [Ping timeout: 256 seconds]
<quinn>
multun: yeh i do not trust the AUR in the least
<multun>
it's actually handy
<multun>
not always maintained, but a good starting point
<quinn>
i'm aware it just made me uncomfortable when i was using it (not that it stopped me though)
<multun>
yeah 100% agree
<quinn>
honestly custom packages always put me off, it probably would have been fine if i was better at managing and debugging pkgbuilds but i only really learned packaging after switching to nix
<multun>
a lot of stuff is super unmaintained and not being able to contribute a fix is annoying
<quinn>
oh is it maintainer-only by default?
<multun>
I think so
<multun>
(for the AUR only)
<multun>
after all, it's the User Repository
<quinn>
oh true. that would really annoy me too (though it's a similar situation now with nixpkgs, i don't have the wherewithall to deal with the PR process)
<quinn>
wherewithal*
jbetz has quit [Ping timeout: 260 seconds]
omnigoat has quit [Read error: Connection reset by peer]
pkral has quit [Ping timeout: 260 seconds]
ProofTechnique has quit [Read error: Connection reset by peer]
claudiii_ has quit [Ping timeout: 256 seconds]
jlpeters has quit [Read error: Connection reset by peer]
MarkRBM has quit [Read error: Connection reset by peer]
betawaffle has quit [Read error: Connection reset by peer]
georgyo has quit [Read error: Connection reset by peer]
bitonic has quit [Read error: Connection reset by peer]
emilazy has quit [Read error: Connection reset by peer]
higherorder has quit [Ping timeout: 256 seconds]
felixfoertsch23 has joined #nixos
felixfoertsch has quit [Ping timeout: 258 seconds]
felixfoertsch23 is now known as felixfoertsch
pingveno has quit [Read error: Connection reset by peer]
sbetrium has joined #nixos
parsnip has quit [Ping timeout: 272 seconds]
wildsebastian has quit [Ping timeout: 264 seconds]
higherorder has joined #nixos
NekomimiScience has quit [Ping timeout: 272 seconds]
teej has quit [Ping timeout: 272 seconds]
angerman has quit [Read error: Connection reset by peer]
<rickhull>
I'm pretty new to NixOS, trying to use it for a development (and build) environment, at the moment. I came across a very helpful blog post and adapted for my needs. I'm looking for feedback, mostly concerning README.md here: https://github.com/rickhull/hello-nix
ProofTechnique has quit [Ping timeout: 272 seconds]
<dkjii>
hey, how would I got about having 2 expressions in the same nix with? I do something like this: with import <nixpkgs/nixos> {}; with config.nix.package; [ pkgs.gnutar ] && with config.system.build [ nixos-install ]
<dkjii>
but it's not working
<multun>
you can do with (import ... {}) // (import ... {});
<ldlework>
multun: how do I get a nix-shell where that nixpkgs fork is active?
<ldlework>
apparently there's some problem with the build process in rpcs3 master, and they are asking me for the content of a file produced by some cmake shit
<multun>
dkjii: why not doing two with?
<ldlework>
and i can't get at it when it's just building in a derivation
<ldlework>
so i think i need to try building this in a nix-shell first
<dkjii>
multun: I have my config in the {} and I believe u need to return only one list of derivations?
<energizer>
dkjii: fwiw a lot of people consider nested `with`s poor style
<multun>
you can either activate the two scopes and build a list, or bring one scope, create a list, exit the scope, bring the other scope, make the other list, and add the two lists
<dkjii>
multun: I tried nix-shell --option substituters "" -E "$(cat expression.nix)", but it's trying to rebuild the whole thing
<dkjii>
multun: whereas if I do nix-env --option substituters "" -iE "$(cat expression.nix)" it only builds sharutils and fails
<multun>
dkjii: your expression looks weird, why do you import nixpkgs/nixos?
<clever>
dkjii: that would be the same as `nix-shell expression.nix` or `nix-shell -E 'import ./expression.nix'`
<dkjii>
although I have to remove the _: to make the nix-shell work
Acou_Bass has quit [Ping timeout: 246 seconds]
<clever>
dkjii: and `nix-env -f expression.nix -i`
<dkjii>
oh ill do that instead^
waleee-cl has quit [Quit: Connection closed for inactivity]
<dkjii>
i'm following the nixos manual for installing nixos install tools
<multun>
huh ok
<multun>
Is it your end goal?
<kandinski>
I just followed the nixos manual's instructions at https://nixos.org/nixos/manual/#sec-linux-config-developing-modules to compile the btusb module with a couple of new magic numbers, and I get the following error: "insmod: ERROR: could not insert module ./drivers/bluetooth/btusb.ko: Invalid module format"
Acou_Bass has joined #nixos
<dkjii>
multun: It's asking for this "nix-env -iE "_: with import <nixpkgs/nixos> { configuration = {}; }; with config.system.build; [ nixos-generate-config nixos-install nixos-enter manual.manpages ]""
<dkjii>
and yeah right now I'm just trying to bootstrap nix without relying on binary caches
<dkjii>
sharutils is just an indirect dependency of nixos-install and manpages
<multun>
interesting
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos
<multun>
I don't think it implies building nixos though
Acou_Bass has quit [Ping timeout: 256 seconds]
<multun>
I think you'd need to 1) build nix 2) get a channel / copy of nixpkgs 3) build some packages, maybe not nixos
<multun>
I may be wrong, I'm not a nix guru
Acou_Bass has joined #nixos
<dkjii>
multun: I got nix installed and a channel selected; this step is for install the tools needed to install nixos
<dkjii>
usually you wouldnt build it, but I'm trying out building it from scratch
<dkjii>
and I have useImpure flags which will mostly force rebuild everything i believe
orivej has quit [Quit: No Ping reply in 180 seconds.]
<{^_^}>
[nixpkgs] @ehmry pushed to master « nixos/lib/make-disk-image.nix: VDI support »: https://git.io/JfH57
<kandinski>
Following the manual's instructions at https://nixos.org/nixos/manual/#sec-linux-config-developing-modules to compile the btusb module with a couple of new magic numbers, and I get the following error: "insmod: ERROR: could not insert module ./drivers/bluetooth/btusb.ko: Invalid module format". I think I may not have ran the configuration properly. Can anybody help?
<dkjii>
kandinski: what's file ./drivers/... ?
<dkjii>
multun: got it to work, seems like the build environment for sharutils is broken somehow, just ended up CTRL+Z and fixing the test scripts
<chessai>
shlevy: Ericson2314: in case tagging in snippets is broken
<dkjii>
any way to inspect the configuration that was used to build an expression? I got /nix/store/HASH-bash-p23 which I'd like to know where it came from
rickhull has quit [Ping timeout: 245 seconds]
<clever>
dkjii: run `nix show-derivation` on that path
amerigo has joined #nixos
<dkjii>
clever: and is there a way to compare it to the original nix expression in nixpkgs?
<clever>
dkjii: that is the final output, after all of the functional expressions have been ran
<clever>
depending on whats happening, you can just grep for one of those strings in nixpkgs to find the expr
<dkjii>
I'll try it out
<dkjii>
clever: Somehow the whole build system is using a busybox bash to compile, not sure why its not using a normal bash
<clever>
thats probably the bootstrap phase, where its building gcc from a gcc in a tarball
<kandinski>
dkjii: the file is a relative path from the kernel tree root, it's btusb.ko, the driver for usb bluetooth adapters.
<clever>
it will then use that bootstrap gcc to build the bash everything else uses
orivej has quit [Ping timeout: 246 seconds]
<dkjii>
gcc is already built which is why i find it weird
orivej has joined #nixos
<clever>
dkjii: the bootstrap may build 2 gcc's, and unpack a 3rd from a tar
<dkjii>
kandinski: file gives you more info about file like arch and module type
<dkjii>
clever: how is the bootstrap phase represented in the nixpkgs?
<dkjii>
like how does sharutils know to use the bootstrap bash vs normal bash?
<{^_^}>
[nixpkgs] @dfordivam opened pull request #90223 → Add 'wasm32-unknown' to known targets → https://git.io/JfHdz
<clever>
dkjii: each stage will apply some overrides, to change a few packages, which will dynamically impact everything that depends on those packages
fragamus_ is now known as fragamus
<wedens[m]>
sorry for asking again, but I just can't let it go :D what may cause a problem that nix-build (or nix run) requires nixos config that can be evaluated with the current channels?
<clever>
dkjii: line 144 will override binutils, coreutils, grep, and gcc, and point all of them towards the bootstrap tools from the tar, thats enough to give a stdenv that can build some things
xelxebar has quit [Remote host closed the connection]
<dkjii>
clever: alright thanks, its crearer to me now
<wedens[m]>
otherwise it fails with things like "undefined bla bla" due to new option in some nixos module
xelxebar has joined #nixos
<dkjii>
clever: what i find odd is that I already have gcc-9.3.0/ built
<clever>
dkjii: 185-216 will then begin building things from that, but force glibc, binutils to stay behind
<clever>
dkjii: what directory is gcc-9.3.0/ at?
<dkjii>
clever: /nix/store/HASH-gcc-9.3.0/
<dkjii>
clever: I also have /nix/store/HASH-glibc-2.30/
<clever>
dkjii: it may not be the version that nixpkgs wants though
<clever>
dkjii: nix allows you to have many glibc's, based on what nixpkgs is saying to build
asheshambasta has joined #nixos
<dkjii>
clever: makes sense, but wouldnt it use the already built glibcs to build the other ones?
kenran has joined #nixos
<clever>
dkjii: it will use the bootstrap glibc, so nothing depends on outside sources
eoli3n_ has joined #nixos
<dkjii>
clever: any way to show a dependency tree as opposed to a list of dependency when using nix-env -i?
<dkjii>
clever: im finding the dependencies very weird, I already have a few programs installed and it's wanting to build a whole new bootstrap based on musl instead of glibc
<clever>
works on both outputs and .drv files
<dkjii>
clever: ok
<dkjii>
clever: I'm trying to reproduce a bug I had with sharutils, but when I try rebuilding with nix-build -E instead of nix-env -iE, it wants to rebuild a whole bunch of stuff
zenxhd has joined #nixos
appleclusters has quit [Quit: Connection closed for inactivity]
<clever>
dkjii: run `nix-diff` on the 2 .drv files its showing at the end
<dkjii>
clever: how do I install this into my path? doesnt seem to be a package
<wedens[m]>
how do I make `debug-shell.service` available, but not enabled? it's listed in example for `systemd.additionalUpstreamSystemUnits`, but I don't want to enable it
simba1 has quit [Ping timeout: 256 seconds]
<wedens[m]>
the doc says "Additional units shipped with systemd that shall be enabled."
asheshambasta has quit [Remote host closed the connection]
<pinkieval>
out of curiosity: the nixos manual mentions " If allowReboot is true, then the system will automatically reboot if the new generation contains a different kernel, initrd or kernel modules.". Does this include cpu microcode?
alp has quit [Ping timeout: 272 seconds]
dingenskirchen1 has joined #nixos
dingenskirchen has quit [Ping timeout: 256 seconds]
<ldlework>
error: getting status of '/home/ldlework/ext/rpcs3/default.nix': No such file or directory
<wedens[m]>
pinkieval: microcode is included in initrd (with `hardware.cpu.*.updateMicrocode`), so I'd suppose that it does
<clever>
ldlework: you told it what <nixpkgs> maps to, but not what file to actually open a shell for
<clever>
ldlework: so its looking for either shell.nix or default.nix in the current dir
<ldlework>
I don't usually have to do that.
<ldlework>
So I didn't expect it.
<clever>
ldlework: you usually have a shell.nix or default.nix in the current dir?
<ldlework>
clever: nix-shell doesn't require either does it?
<clever>
it does when you dont tell it what else to do
<kandinski>
seems to me that btusb.ko is what it should be: "btusb.ko: ELF 64-bit LSB relocatable, x86-64, version 1 (SYSV), BuildID[sha1]=363ddb2bbe84b22efda842f19c2a18c461b94527, not stripped
asheshambasta has joined #nixos
<pinkieval>
wedens[m]: cool, thanks
klntsky has quit [Remote host closed the connection]
klntsky has joined #nixos
<asheshambasta>
anyone seen "Daemon startup without any loaded modules ..." in pulseaudio? I'm attempting to use pa to output to jack sinks which means pa shouldn't be enabled as a service for reasons I don't understand.
<asheshambasta>
I'm trying to start pa with `pulseaudio -v`
<asheshambasta>
I've also purged the ~/.config/pulse and ~/.pulse dirs.
vandenoever has joined #nixos
vandenoever has quit [Changing host]
vandenoever has joined #nixos
user_0x58 has quit [Quit: Leaving]
vandenoever has quit [Ping timeout: 246 seconds]
orivej_ has quit [Ping timeout: 256 seconds]
q3k has quit [Ping timeout: 240 seconds]
<ldlework>
In nix-shell is there a way to invoke the standard builder? Like if you were to make a derivation with src = ./. whatever would happen by default.
<stteevveen>
Hi! I'm trying to understand why my .o files generated simply with make contains debug info, but not when I use nix-build that will use this same Makefile. When I use nix-build I see there is a "pos-installation fixup". And for all my .o ffiles it writes: shrinking RPATHS of ELF executable and libs: Shrinking myFile.o - wrong ELF type. There is
<stteevveen>
also a line about stripping by executable. Do you know of any Nix-build mechanism that remove debug info in the post-installation fixup ? I'm only wondering, trying to explain why I can debug in one case and not in the other witht the same Makefile.
mekeor has quit [Ping timeout: 260 seconds]
mekeor has joined #nixos
<clever>
stteevveen: add dontStrip = true; to the derivation
<stteevveen>
@clever great news
<stteevveen>
@clever but that made me come to a question: should I use nix-build at any tiny recompilation, or should I you make, and nix-build only from time to time to update the package/derivation ?
<stteevveen>
should I you = should I use
<clever>
stteevveen: for fast interation, you usually want to run make in nix-shell
<clever>
then it can reuse any other .o files your project isnt changing
orivej has joined #nixos
mallox has joined #nixos
<{^_^}>
[nixpkgs] @ehmry opened pull request #90226 → kristall: init at 0.2 → https://git.io/JfHbr
ncakerlist has quit [Ping timeout: 240 seconds]
m0rphism has joined #nixos
chloekek_ has joined #nixos
fragamus has quit [Quit: Connection closed]
<{^_^}>
[nixpkgs] @vbgl opened pull request #90227 → coq_8_12: init at 8.12+β1 → https://git.io/JfHbX
<eyJhb>
I am trying to mount a LUKS encrypted volume, where I have some ZFS datasets. Currently trying to do it like so - https://termbin.com/9116
<eyJhb>
But for some reason, it seems like it won't open the LUKS crypt, at least that is was journal says (NOTHING about cryptsetup), so the rtank/data will fail
<eyJhb>
Any ideas why? I would expect to see cryptsetup some place
od1n has joined #nixos
luc65r has joined #nixos
<od1n>
how does one get over the Pillow-7.1.2 issue preventing upgrade?
<od1n>
not an issue, I get that it's a good thing that old python was deprecated, just wondering if there's any steps to follow to be able to upgrade --switch now
<{^_^}>
[nixpkgs] @FRidh pushed commit from @r-ryantm to staging « libmbim: 1.22.0 -> 1.23.900 »: https://git.io/JfHNE
mallox has joined #nixos
uidnull has joined #nixos
<uidnull>
can nix be used as a replacement to apt in ubuntu?
<uidnull>
if so, what's the benefit of using nixos over ubuntu with nix
is_null has quit [Ping timeout: 264 seconds]
sbetrium has joined #nixos
<srk>
uidnull: it can, but you're missing the integration with the system and ability to configure e.g. services or other aspects of your system via nix
_ris has joined #nixos
<uidnull>
ah, so its value is the native integration
<tyrion-mx>
Is `nix build` the same as `nix-build`?
<{^_^}>
[nixpkgs] @vbgl pushed commit from @anmonteiro to master « ocamlPackages.tls: 0.12.0 -> 0.12.1 »: https://git.io/JfHN6
<srk>
uidnull: yup
<tobiasBora>
Hello, how can I get the configuration value of a package outside of nix? I tried nix eval --raw config.services.postgresql.dataDir but it does not work
<srk>
tyrion-mx: the result is the same, UI is different
<uidnull>
i heard blm linux (black lives matter linux) has really good integration as well
ris has quit [Ping timeout: 246 seconds]
<tyrion-mx>
tobiasBora: you can do `nix repl '<nixpkgs/nixos>'` and then type `config.services.postgresql.whatever`
is_null has joined #nixos
<chreekat[m]>
nix and nix-build aren't the same. The first is experimental software, the latter is part of the latest official release. (This might only matter if you're using it in production though)
<tobiasBora>
tyrion-mx: great, thanks! and do you have a non interractive version?
magnetophon has quit [Ping timeout: 246 seconds]
<tyrion-mx>
tobiasBora: nope, I have no idea (I am not an expert). But if you find out please tell me :D
<tyrion-mx>
maybe you can pass stuff from stdin to nix repl?
orivej has quit [Ping timeout: 260 seconds]
orivej_ has joined #nixos
<tobiasBora>
tyrion-mx: sounds a bit dirty, especially when you want to parse the output (you will need to remove all the text given by nix repl)
agsdheidjd has joined #nixos
codygman has quit [Ping timeout: 265 seconds]
codygman has joined #nixos
<tyrion-mx>
tobiasBora: I guess you can do it with `nix eval` but I am not able
<tobiasBora>
tyrion-mx: since you seem to understand more than me nix repl, any idea how to turn nix eval --raw nixpkgs.postgresql_11.outPath into nix repl? I'd love to have tab completion has I can't find much documentation about that
<clever>
tobiasBora: `nix repl '<nixpkgs>'` then `postgresql_11.outPath`
<luc65r>
Hi! Maybe here my call for help will have a little bit more success
<uidnull>
luc65r: the answer is: don't jump
<luc65r>
I'm tring to build my system with gcc10 and binutils 2.34: https://github.com/NixOS/nixpkgs/pull/89793. That broke systemd too. Most of the services (like journal or udev) fail with code 127. I can launch them manually without any problem. I tried to build systemd with gcc9 and gcc10, and it seems to be related to binutils-2.34. Does anyone have an
<luc65r>
idea of what could have gone wrong?
<{^_^}>
#89793 (by luc65r, 4 days ago, open): Bump gcc to gcc10 and binutils to 2.34
<{^_^}>
[nixpkgs] @vbgl pushed commit from @sternenseemann to master « ocamlPackages.dns(-client): 4.5.0 -> 4.6.0 »: https://git.io/JfHA3
marusich has quit [Quit: Leaving]
<tobiasBora>
clever: good point, thanks! and in that nix repl shell, any idea how to query config.services.postgresql.dataDir? Or do I need to quit the shell and start a new shell with file "<nixpkgs/nixos>"?
<clever>
tobiasBora: yeah, you need to switch over to nixos
markus1189 has joined #nixos
<tyrion-mx>
tobiasBora: in the nix repl you can use `:l <nixpkgs/nixos>` and then simply write config.whatever
sputny has joined #nixos
<tobiasBora>
ok thanks!
sputny has quit [Client Quit]
<tyrion-mx>
tobiasBora: I suggest you read the nix pills if you haven't. I read only the parts that were interesting to me and I found them pretty useful
<{^_^}>
[nix] @mpscholten opened pull request #3692 → Don't stop user from installing the nix store on an unencrypted volume → https://git.io/JfHAu
<tobiasBora>
tyrion-mx: yeah, I did that a while ago, but I forgot everything, I should start again ;-)
codygman has quit [Ping timeout: 256 seconds]
m4ts_ has quit [Quit: bye]
uidnull has left #nixos [#nixos]
<tyrion-mx>
tobiasBora: take notes! I use zettlr to write down notes when I read stuff. It helps me remember it and more importantly I don't have to read stuff again when I forget everything
<asheshambasta>
so it seems to me that if one starts pulseaudio via `pulseaudio -D --start`, there is no way to have clients discover the pulseaudio server.
<asheshambasta>
it seems to rely on environment variables, configuration values etc. as fallbacks but none of them are populated, and what `default-server` should be in client.conf is also not documented.
<tobiasBora>
tyrion-mx: yeah, it's what I'm doing now ;)
<asheshambasta>
srk: following up on our discussion earlier, how are you starting pulseaudio in your system?
<asheshambasta>
would you mind sharing your config. if any?
<srk>
alsa or loopback is probably not needed since steam games work with pulse
orivej_ has quit [Ping timeout: 246 seconds]
<asheshambasta>
I see, I'll try replicating that and trying
<srk>
asheshambasta: I haven't tested this properly as I don't want to re-login or reboot right now but I was able to su - srk, stop and start pulseaudio manually and got sound via jack sink
alexherbo2 has joined #nixos
orivej has joined #nixos
<srk>
asheshambasta: one tweak this needs is to start pulse after like suggested in the issue you've posted yesterday or adjust its modules
m4ts has joined #nixos
<srk>
*after jack
never_released has quit [Quit: ZNC 1.7.5 - https://znc.in]
knupfer has quit [Quit: knupfer]
knupfer has joined #nixos
<asheshambasta>
so rebuilding this as you suggest, and having hardware.pulseaudio.enable = true, my pa clients are still unable to connect to pa.
knupfer has quit [Client Quit]
knupfer has joined #nixos
<asheshambasta>
there is something that definitely looks wonky here
knupfer has joined #nixos
<asheshambasta>
and overall, this whole thing still looks brittle and maybe nix can help
never_released has joined #nixos
codygman has joined #nixos
<asheshambasta>
rebooting, sigh.
asheshambasta has quit [Remote host closed the connection]
orivej has quit [Quit: No Ping reply in 180 seconds.]
asheshambasta has joined #nixos
orivej has joined #nixos
<asheshambasta>
for some reason my network seems extremely slow now.
asheshambasta has quit [Remote host closed the connection]
__monty__ has joined #nixos
asheshambasta has joined #nixos
<asheshambasta>
I hate computers :-)
shabius has quit [Read error: Connection reset by peer]
<asheshambasta>
(now all these reboots/nix-rebuilds etc.) have broken networking. Things are really slow on my desktop now.
shabius has joined #nixos
<srk>
oO weird
asheshambasta has quit [Remote host closed the connection]
asheshambasta has joined #nixos
noname_ has joined #nixos
alexherbo28 has joined #nixos
alexherbo2 has quit [Ping timeout: 260 seconds]
alexherbo28 is now known as alexherbo2
asheshambasta has quit [Remote host closed the connection]
cosimone has joined #nixos
alp has quit [Ping timeout: 246 seconds]
dkjii has quit [Quit: Lost terminal]
<noname_>
.list csgo
Darkmatter66 has joined #nixos
<wedens[m]>
how can I do `nixos-install --system` if I can pre-build system's closure except disks configuration?
<clever>
wedens[m]: you cant really, but if you copy the pre-built nixos to /mnt/, it will have a lot less to rebuild when it builds from the new configuration.nix
<clever>
if you keep the nixpkgs rev the same
Darkmatter66_ has quit [Ping timeout: 256 seconds]
<wedens[m]>
clever: so I can do `NIX_PATH=.... nixos-install --no-channel-copy` after `nix copy`?
zarel has joined #nixos
<clever>
wedens[m]: yeah, that should work
zarel_ has quit [Ping timeout: 265 seconds]
CMCDragonkai1 has quit [Quit: CMCDragonkai1]
alp has joined #nixos
<eyJhb>
How can I get a zfs filesystem to mount before a LUKS encrypted filesystem? Currently I have neededForBoot on the ZFS part, but initrd.luks.devices.... will still be before my normal filesystem?
<eyJhb>
In this case it won't help, as it is boot.initrd.luks.devices that needs to understand, that it should do a filesystem mount before it will function :/
<srk>
looks like you need stage1 systemd and proper deps :)
<eyJhb>
With most of Nix
<eyJhb>
I don't know if that is a thing :p
<wedens[m]>
is there an alternative to overlays-compat trick to have the same overlays everywhere? (https://nixos.wiki/wiki/Overlays#Using_nixpkgs.overlays_from_configuration.nix_as_.3Cnixpkgs-overlays.3E_in_your_NIX_PATH)
<eyJhb>
Lets try this solution...
andi- has quit [Remote host closed the connection]
<eyJhb>
Apparantly "right before LVM" is untrue, damn it
never_released has quit [Quit: ZNC 1.7.5 - https://znc.in]
avaq has joined #nixos
jakobrs has left #nixos ["WeeChat 2.7.1"]
asheshambasta has joined #nixos
never_released has joined #nixos
<asheshambasta>
any ideas on how to troubleshoot slow networking on nixos?
yasu has joined #nixos
<asheshambasta>
I rebooted my system numerous times to get jack/paudio to work together and I seem to have broken something with networking where my network now seems extremely slow. It seems unrelated to me but you never know.
<yasu>
This line used to work on my old sever but not in the new one...services.wakeonlan.interfaces = [ { interface = "enp4s0"; method = "magicpacket"; }] ;
<yasu>
The problem line in my /etc/nixos/configuration.nix is:
<asheshambasta>
ping takes a long time to start but the latency seems ok once it gets going.
<eyJhb>
No.. Sadly, and I couldn't even do the idea I wanted
<eyJhb>
I am missing a "run this before you try to unlock the device"
orivej has quit [Ping timeout: 256 seconds]
<eyJhb>
Because both are running at somewhat the same time
orivej has joined #nixos
<srk>
yup
luc65r has quit [Remote host closed the connection]
<betawaffle>
does anyone know how to get opencl working on nixos? (i've got an amd 5700 xt, and i want to use it for blender)
<srk>
but it retries
yasu has quit [Remote host closed the connection]
<srk>
betawaffle: I've had to recompile blender with opencl enabled and it just worked, with nvidia tho
<srk>
oh, no, that's cuda support actually
<betawaffle>
well, opencl is enabled for blender, but it's not detecting my card
<betawaffle>
and clinfo reports zero platforms
<betawaffle>
so i feel like there's something i'm missing
<srk>
,locate clinfo
<{^_^}>
Found in packages: clinfo
luc65r has joined #nixos
<evils>
betawaffle: fwiw, i've got nixos-rocm set up, blender detects my wx 3100, but it segfaults when trying to render (it ran fine for foldingathome though)
<evils>
it's been a while since i tried amdgpu-pro and i'm not sure what's the right way to set that up anyway
<betawaffle>
do i need amdgpu-pro?
<evils>
amdgpu-pro includes a closed source opencl driver for amd cards i think, ROCm is the open source alternative
<betawaffle>
i'm _ok_ with closed source, if it works. but it's not clear how to use it on nixos
Thra11 has quit [Ping timeout: 256 seconds]
chloekek_ is now known as chloekek
<evils>
yea i suspect if you can get it working, amdgpu-pro has a better chance of actually working :P
zarel_ has joined #nixos
<asheshambasta>
srk: actually, switching to a local dns server seems to "fix" the issue, I was using google dns earlier.
<asheshambasta>
Same issue with opendns.
zarel has quit [Ping timeout: 256 seconds]
<eyJhb>
srk: F it, lets patch it
orivej has quit [Quit: No Ping reply in 180 seconds.]
liminal18 has joined #nixos
orivej has joined #nixos
ramses_ has joined #nixos
luc65r has quit [Ping timeout: 272 seconds]
<ramses_>
Hi guys. I have the impression that IPv6 is not properly accounted for by the networking.firewall options. I added some custom rules in extraCommands, but they only affect the IPv4 firewall and not the IPv6 one. Is this expected behaviour?
<ramses_>
I don't see any IPv6-specific options
<ramses_>
Since there are no options for specific protocol versions, I was expecting the rules to be applied to both
<srk>
ramses_: they are shared but you need to use ip46tables I guess
<liminal18>
how do I get rid of this issue: warning: name collision in input Nix expressions, skipping '/home/liminal18/.nix-defexpr/channels_root/nixos'
<eyJhb>
If anyone known of a better way, than to execute `mount -t zfs /tmp/keypar`, so that I can extract the key for my LUKS volume
<liminal18>
also not sure how I caused it but I am assuming I did something wrong with a build.
<eyJhb>
- my encrypted LUKS volume, where the key is located in a encrypted dataset. I can get stage1 to import the pool, so the datasets are ready, but it will only mount them after the LUKS par have been openened
orivej has quit [Quit: No Ping reply in 180 seconds.]
<srk>
looks legit
orivej has joined #nixos
<asheshambasta>
srk: in my case when paudio is running as a service (hardware.pulseaudio.enable=true) it is unable to find the dbus daemon; it tries to launch one and fails
Thra11 has joined #nixos
luc65r has joined #nixos
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<asheshambasta>
and when it isn't running as a service, I can run it via pulseaudio -D --start but that daemon is not discovered by any client.
<ramses_>
srk: I am such an idiot, of course I was just using "iptables" *facepalm*
<srk>
asheshambasta: not sure about that :(
<ramses_>
The defaults seem to differ though, IPv4 has DROP policy for FORWARD, IPv6 has ACCEPT. But this could also be docker changing the default for IPv4
<srk>
ramses_: hehe, it's good to have a copy of nixpkgs around so you can check how stuff is implemented when needed
ncakerlist has joined #nixos
<asheshambasta>
it could be in your case that jack is running as a server
<asheshambasta>
in my case I cannot start the server via qjackctl: server options are greyed out
<srk>
asheshambasta: yup, it is, and pulse is just a client for it. is your user part of audio group ?
luc65r has quit [Ping timeout: 264 seconds]
<srk>
asheshambasta: server options in Advanced tab?
zarel has quit [Ping timeout: 256 seconds]
zarel has joined #nixos
<asheshambasta>
yes x2
<srk>
ok
<srk>
(not grayed out for me, only few of them)
<asheshambasta>
for me it is entirely greyed out
<srk>
asheshambasta: you're on 20.03?
<asheshambasta>
yeah
<srk>
running out of ideas but I can try with 20.03 later
ramses_20 has joined #nixos
<asheshambasta>
I must thank you for being so helpful and patient
ramses_ has quit [Ping timeout: 258 seconds]
<srk>
yw!
ramses_20 is now known as ramses_
<{^_^}>
[nixos-homepage] @github-actions[bot] pushed commit from GitHub Actions to master « Update flake.lock and blogs.xml [ci skip] »: https://git.io/JfHjH
<asheshambasta>
I do have jackd.enable set to true; not sure if that is the issue.
<srk>
I'm also interested in having this working without too much trouble
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos
sputny has joined #nixos
<asheshambasta>
yeah
knupfer has quit [Quit: knupfer]
knupfer has joined #nixos
knupfer has quit [Read error: Connection reset by peer]
knupfer has joined #nixos
knupfer has quit [Read error: Connection reset by peer]
<srk>
I'll install a vm and pass it another USB soundcard so I can reboot and work with clean environment
knupfer has joined #nixos
<asheshambasta>
thanks a lot
<asheshambasta>
to me it seems like qjackctl is letting the systemd service take precedence but still presents an interface represeting the current state.
<srk>
have too much stuff open to reboot the desktop and the config is alos quite large so hard to tell what's actually needed
davidv7 has joined #nixos
noname_ has quit [Ping timeout: 246 seconds]
<asheshambasta>
yeah, and this stuff is quite messy
<asheshambasta>
also, stopping the jack service and attempting to start it with qjackctl leads to a permission denied: something with futexes.
<asheshambasta>
and there is a sum total of one SO issue on the same error message: the user claims to have fixed it on ubuntu via a purge-install
<asheshambasta>
on the outset it seems to me that the working of pulseaudio is really contorted
<asheshambasta>
it relies on env. vars, configuration, etc; it is extremely hard to reason about
orivej has quit [Ping timeout: 256 seconds]
orivej_ has joined #nixos
ramses_ has joined #nixos
<asheshambasta>
I wouldn't mind getting rid of paudio and just using jack as my primary sound driver, but it seems like most common applications are built with pa in mind, I
<asheshambasta>
this is what I gather from reading up on this, it would seem to me that other software would be built using ALSA in mind, which, as per my understanding is the lowest level of sound in linux from an application standpoint.
<srk>
this. ok, got jack working in vm with another soundcard, will take a break and play with pa on top of it
knupfer1 has joined #nixos
<srk>
there's also pipewire which could possibly replace both, but I don't know much about it
<asheshambasta>
supercollider seems to exclusively support jack
<srk>
like in theory interop should be possible, converting jack apps to pulse apps and vice versa
vandenoever has quit [Ping timeout: 260 seconds]
knupfer has quit [Ping timeout: 260 seconds]
knupfer1 is now known as knupfer
<srk>
but who knows, maybe the APIs are completely different
<asheshambasta>
in your vm test case, I
orivej_ has quit [Ping timeout: 260 seconds]
<asheshambasta>
I'd be keen in finding out that iff you run jackd + pulseaudio as services, can you i.) use qjackctl to control jack; ii.) connect pa to jackd
<pinkieval>
I feel like I'm missing something obvious
ncakerlist has quit [Ping timeout: 260 seconds]
zarel has quit [Ping timeout: 264 seconds]
mgdm has joined #nixos
<srk>
eyJhb: :)
<mgdm>
Hello. I'm looking into packaging a Node CLI tool I'm using. I've added it to node-packages.json and run generate.sh, but the diff it's created touches a lot of unrelated packages. I assume I'd only want to PR the bits that were relevant? Or should I commit the whole node-packages.nix with all the changes?
<srk>
pinkieval: you can cross compile it yourself with nix-build -A pkgsCross.aarch64-multiplatform.ubootRockPro64
<pinkieval>
srk: I don't have nix installed yet :|
<pinkieval>
(well I do, but on a machine I don't trust to bootstrap other machines)
<srk>
the first command creates uboot folder with idbloader.img and u-boot.itb
orivej has joined #nixos
dermetfan has joined #nixos
<pinkieval>
hmm, still requires a nix machine
<pinkieval>
what format is .nar?
<srk>
nix archive ;)
<pinkieval>
oh no
<srk>
you can install nix on any distro
dramforever has joined #nixos
<pinkieval>
hmm right, a single-user install should do
<pinkieval>
oh but it still requires root, damn
<pinkieval>
nvm, I forgot I can just chown /nix
ramses_ has quit [Ping timeout: 258 seconds]
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos
noname_ has joined #nixos
luc65r has quit [Read error: Connection reset by peer]
<srk>
asheshambasta: yup, I'm getting errors trying to load jack sink and also the dubs-daemon related one
luc65r has joined #nixos
noname_ has quit [Client Quit]
<asheshambasta>
okay, interesting
<srk>
asheshambasta: what's your desktop session?
<asheshambasta>
i think if you run jack as a service but pa as a daemon; you will be able to connect pa to jack, but nothing can any longer discover the pa server.
<asheshambasta>
I'm using xmonad
<srk>
my desktop setup runs xmonad with home-manager but I'm on more recent commit, vm has sddm+plasma
<srk>
connect(47, {sa_family=AF_UNIX, sun_path="/dev/shm/jack_default_1000_0"}, 110) = -1 ENOENT (No such file or directory)
shatsky has joined #nixos
shatsky has left #nixos [#nixos]
<asheshambasta>
is there a way to get rid of pulseaudio altogether and just use jack?
<srk>
not sure, depends on apps you need I guess
<gchristensen>
I'm trying to debug a Go program and in the backtrace it is showing some stack frames as just #8 0x00007fb857c67700 in ?? () having ??
luc65r has quit [Read error: Connection reset by peer]
orivej has quit [Quit: No Ping reply in 180 seconds.]
<gchristensen>
I see location information for other stack frames, though. uniquely, I see information for lines like /build/go/pkg/mod/github.com/golang/protobuf@v1.3.3/proto/table_unmarshal.go and /nix/store/26l01arw0gxm1d883ni5z4ygairsyvn7-go-1.12.17/share/go/src/runtime/proc.go but no stack frames show location information for actual the program's sorce itself, leading me to think the ??'s are in the program's
<gchristensen>
source. any tips?
orivej has joined #nixos
amerigo has quit [Quit: Connection closed for inactivity]
luc65r has joined #nixos
<asheshambasta>
srk: just to be sure: pastebin.com/ZWftygGB
vandenoever has joined #nixos
vandenoever has quit [Changing host]
vandenoever has joined #nixos
<srk>
gchristensen: maybe check the executable with objdump --debugging
tobeportable has joined #nixos
<srk>
asheshambasta: yup, same, something prevents pulseaudio user service the access to the socket, not sure what
civodul has joined #nixos
Chiliparrot has joined #nixos
<asheshambasta>
maybe the default params are incorrect in the module and can be overriden?
<asheshambasta>
but I'm not sure what the params can be for module-jack-sink
vandenoever has quit [Read error: Connection reset by peer]
<srk>
hmm, looks good, not sure why gdb fails, haven't debug any go yet
ddellacosta has joined #nixos
vandenoever has joined #nixos
vandenoever has joined #nixos
vandenoever has quit [Changing host]
<srk>
asheshambasta: default params should be good, it tries to access the right path, socket is owned by jackaudio
<asheshambasta>
how did you check?
<dramforever>
How do I get a nix-shell with cross compiler and qemu? I got this far, I can't seem to convince it to use the native qemu: https://dpaste.org/txVf (removing qemu makes everything work)
<srk>
asheshambasta: strace -f -o tr1 <pulseProcess> .. found this - connect(47, {sa_family=AF_UNIX, sun_path="/dev/shm/jack_default_1000_0"}, 110) = -1 ENOENT (No such file or directory)
orivej has quit [Ping timeout: 246 seconds]
luc65r has quit [Read error: Connection reset by peer]
orivej has joined #nixos
luc65r has joined #nixos
drakonis has joined #nixos
luc65r has quit [Read error: Connection reset by peer]
luc65r has joined #nixos
zebrag has quit [Quit: Konversation terminated!]
_deepfire has joined #nixos
jumper149 has joined #nixos
<asheshambasta>
the path is different in my case, but I guess that is expected.
luc65r has quit [Read error: Connection reset by peer]
luc65r has joined #nixos
<asheshambasta>
and I did a bit more research; turns out most applications on linux are built with pulseaudio in mind; spotify has an issue reporting a feature request which has been marked stale and has seen no activity for a few years.
<{^_^}>
[nixpkgs] @FRidh opened pull request #90240 → WIP: buildPythonPackage: perform tests in separate derivation → https://git.io/JfQJ5
<srk>
asheshambasta: hmm, no it actually tries to access the wrong path, if I symlink jack_default_0 to jack_default_1000_0 it gets a bit further
leungbk has joined #nixos
knupfer has quit [Quit: knupfer]
admicos has left #nixos ["Leaving"]
knupfer has joined #nixos
drakonis_ has joined #nixos
drakonis has quit [Ping timeout: 272 seconds]
sputny has quit [Quit: sputny]
sputny has joined #nixos
Izorkin has joined #nixos
<srk>
mpv uses the correct one with -ao jack
<srk>
maybe pulse assumes that jack is running in user session and tries that
kenran has joined #nixos
leungbk has quit [Remote host closed the connection]
<asheshambasta>
I'm unable to start jack via qjackctl, I get permission related errors.
arahael1 has joined #nixos
Arahael has quit [Ping timeout: 246 seconds]
<bqv>
infinisil: what even is the use case of having multiple versions of ghc
<bqv>
do you personally need it? i'm amazed people put up with that
<asheshambasta>
srk: if I do systemctl stop jack.service, I see a bunch of "Server is not running" errors in journalctl
sputny has quit [Quit: sputny]
sputny has joined #nixos
<srk>
asheshambasta: so it looks like the problem is with the initialization - pulse trying to connect to wrong path and creating a wrong path which can't be accessed by jackd. trying to compare the code with what mpv does
vuko has quit [Quit: leaving]
vuko has joined #nixos
drakonis_ has quit [Read error: Connection reset by peer]
drakonis_ has joined #nixos
<asheshambasta>
tbh, at this point I
<asheshambasta>
I'd be happy to live with just one sound engine, if only I can make the rest of my applications work with Jack
<srk>
jack module sets it to JACK_PROMISCUOUS_SERVER = "jackaudio"; but it won't propagate to pulseaudio user service, only DISPLAY does (which is why mpv works but not pulse user service)
pamplemousse has joined #nixos
is_null has quit [Remote host closed the connection]
<srk>
asheshambasta: fix for that is systemd.user.services.pulseaudio.environment.JACK_PROMISCUOUS_SERVER = "jack";
<srk>
asheshambasta: but now it fails on permission denied, which is kind-of expected :D
<asheshambasta>
I believe this is also mentioned in the jack nixos wiki; in my case however; I'm using hardware.pulseaudio which doesn
<asheshambasta>
doesn't accept 'environment
<srk>
asheshambasta: hardware.pulseaudio just sets up the systemd.user.services.pulseaudio for you
<asheshambasta>
ah okay
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos
proofofkeags has quit [Remote host closed the connection]
<asheshambasta>
yeah I also see the permission denied error, perplexing that the error is issued by jackd though
meh` has joined #nixos
<srk>
asheshambasta: yup, it's how the sockets are created, pulse connects to jack and also creates its own - .. srw-rw---- 1 jane users 0 Jun 13 17:38 jack_PulseAudio JACK Sink_0
<asheshambasta>
I'm afraid I'm still getting permission denied
<srk>
restart jack manually
<srk>
it doesn't restart automatically
<asheshambasta>
aha, yes!
<asheshambasta>
YES YES YS
<asheshambasta>
you're the man, and I owe you beer.
<Yaniel>
oh right I was going to look into that pulse-on-jack type setup
<phirsch>
Taking a note as well... Thx!
<srk>
asheshambasta: \o/
<srk>
will update the wiki I guess and try to PR this
<phirsch>
srk++
<{^_^}>
srk's karma got increased to 19
<evils>
srk ++ asheshambasta++
<{^_^}>
asheshambasta's karma got increased to 1
<evils>
srk++
<{^_^}>
srk's karma got increased to 0b10100
dermetfan has quit [Ping timeout: 256 seconds]
<phirsch>
asheshambasta++
<{^_^}>
asheshambasta's karma got increased to 1.0
<asheshambasta>
thanks, wow :-) all credit goes to srk though
dramforever has quit [Quit: Leaving]
proofofkeags has joined #nixos
<phirsch>
It was time someone figured it out - it took me a few hours to figure out how to run pa on jack on-demand, but I gave up on jackd as a system service.
<srk>
fun investigation, strace is the best tool :D
<phirsch>
+1
proofofkeags has quit [Ping timeout: 260 seconds]
<asheshambasta>
I will need to learn how to use it after this
<asheshambasta>
(btw, I do make music at present; but doing it in Haskell has really interested me)
orivej has quit [Ping timeout: 246 seconds]
fragamus_ has joined #nixos
xelxebar has quit [Remote host closed the connection]
pamplemousse has quit [Ping timeout: 258 seconds]
xelxebar has joined #nixos
orivej has joined #nixos
<srk>
cool, I'm mostly toying with tidal and VCVRack but was missing visual feedback for the former, wish feedforward had vim bindings :D
zebrag has quit [Quit: Konversation terminated!]
fragamus_ is now known as fragamus
<asheshambasta>
I think emacs also has some decent support for supercollider
<srk>
maybe vim could be used as a renderer, looks like feedforward is a curses interface, haven't tried it yet but I'm going to soonish
<srk>
yeah, or that
<asheshambasta>
and for tidal for that matter; but then I'm writing my own framework on top of tidal while stealing some handy stuff from BootTidal.hs; to set things up to support common stuff I do
<{^_^}>
[nixpkgs] @fps closed pull request #90019 → xonotic: fix a couple of bugs → https://git.io/JfStU
cole-h has joined #nixos
<srk>
I was using it with vim/screen/ghci before, could be nice with emacs with scel
<infinisil>
bqv: HIE doesn't work if you have a mismatched GHC version
zebrag has joined #nixos
<infinisil>
Meaning a project with GHC 8.6.4 doesn't work with a 8.6.5 HIE
<asheshambasta>
srk: yeah, tidal looks interesting; another library that I looked into a while back was csound-expression
zupo has joined #nixos
<asheshambasta>
that, targets csound instead of supercollider as its backend
<infinisil>
bqv: So if not all versions are supported, this means that if HIE updates to a new GHC version, you also need to update your project with that
<asheshambasta>
(has nice tutorials as well)
<Yaniel>
btw are y'all using realtime kernels with jack?
<Yaniel>
what are the risks in doing so?
<srk>
sometimes :)
<srk>
no risks but incompatible with nvidia drivers
tokudan has quit [Remote host closed the connection]
<srk>
asheshambasta: first time I hear about csound, thanks for the tip
<bqv>
infinisil: that's ok, i'm more confused by why you have multiple GHCs on your system at any point - i.e. why not just sync hie with whatever "ghc" is in nixpkgs
<infinisil>
bqv: Because projects use pinned nixpkgs
<infinisil>
And those have varying GHC versions
<bqv>
hm, of course they do
<asheshambasta>
Yaniel: I'm very new to Jack (credits to srk just now), I'm yet to see what the latencies look like before deciding on the kernel stuff
<infinisil>
And if it wasn't for the glibc issue, this would've worked fine
<Yaniel>
I'm interested in setting it up as well, in particular for renoise
<asheshambasta>
srk: csound is very interesting too
<Yaniel>
since atm it has to take over the entire sound card, which is inconvenient
<bqv>
you know, i'm starting to think i can live without hie
<{^_^}>
#42379 (by erictapen, 1 year ago, closed): stack overflow (possible infinite recursion) depending on type of git checkout
<bqv>
that's relatively amusing
<bqv>
i love weird bugs like that
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos
drakonis has joined #nixos
vandenoever has joined #nixos
vandenoever has quit [Changing host]
vandenoever has joined #nixos
knupfer1 has joined #nixos
knupfer has quit [Quit: knupfer]
knupfer1 is now known as knupfer
ris has joined #nixos
_ris has quit [Ping timeout: 256 seconds]
orivej has quit [Ping timeout: 264 seconds]
orivej_ has joined #nixos
alp has joined #nixos
orivej_ has quit [Ping timeout: 264 seconds]
orivej has joined #nixos
orivej has quit [Read error: Connection reset by peer]
orivej has joined #nixos
chloekek has quit [Ping timeout: 246 seconds]
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
Knuxfan has joined #nixos
<Knuxfan>
hi guys, i have a question with nixos
ramses_ has joined #nixos
<Knuxfan>
i recently built an iso image using nixos generator but when I boot it, in virtualbox or on a real machine, crashes once it reaches the second stage
<Knuxfan>
ohhh i think i figured it out nevermind, im using the 20.03 version of home-manager instead of 20.09
<Knuxfan>
that's probably the issue
pbb_ is now known as pbb
knupfer1 has joined #nixos
<cole-h>
I don't think that should crash a machine...
knupfer has quit [Ping timeout: 256 seconds]
knupfer1 is now known as knupfer
<simpson>
Crashing at second stage? Does it crash right around the time that the screen flickers and the font changes; i.e. when modesetting kicks in?
<betaboon>
Knuxfan: maybe you could try getting the full log. maybe by adding a serial port to the vm and getting the text-log that way
<betaboon>
Knuxfan: i think in the virtualbox ui you can find that in: (selected VM) -> Settings -> Serial Ports -> Port 1 -> Enable + Port Mode=Raw File + provide a file-path
<betaboon>
Knuxfan: do you know what the kernel-cmdline is?
<Knuxfan>
you mean pressing tab in isolinux?
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos
<betaboon>
Knuxfan: dunno, maybe? i think it would be good to add something like `console=ttyS0 console=tty0` and maybe `ignore_loglevel` to get all the good stuff
<Knuxfan>
yeah i tried that
cosimone has quit [Ping timeout: 256 seconds]
<Knuxfan>
i typed in console=lp0
cosimone_ is now known as cosimone
<cole-h>
Why lp0? That's not a serial port -- that's a printer...
<betaboon>
Knuxfan: did you completly replace the cmdline? or just append to it ?
<betaboon>
i wouldnt want any package to put stuff in my home-directory :/
<cole-h>
Which is why alacritty doesn't -- we don't want that either :^)
<MtotheM>
I'm aware, but on other distros this example is in the ../doc/ path
<MtotheM>
suppose I could just copy it from Github
<cole-h>
Don't use the one from master
<betaboon>
MtotheM: i got nix-expressions creating a config and symlinking to in /etc/xdg/alacritty.yaml thats it.
<cole-h>
Use the one from the tagged release
<betaboon>
i think alacritty is not packaged to include the default-config in share
<cole-h>
Currently, no.
philr_ has quit [Ping timeout: 246 seconds]
<alexfmpe>
anyone here managed to connect to a 3d printer on nixos?
<alexfmpe>
pronterface on my mac detects my prusa mk3s, but the pronterface on my nixos doesn't
bdju has quit [Ping timeout: 246 seconds]
bdju has joined #nixos
ramses_ has quit [Ping timeout: 264 seconds]
<alexfmpe>
huh turns out I needed to run it with sudo
<srk>
alexfmpe: you can add udev rule
<betaboon>
i have a problem using overlays and crosspkgs. i normally pull in some packages from unstable using an overlay like in the attached link, but that breaks cross-compilation for those packages: https://gist.github.com/betaboon/c73064271609d7080e53f7a650c7895a
<betaboon>
i think i would need to set `crossSystem` for that `import <nixpkgs-unstable>` but as i dont have access to `config` in the overlay i dont know how
<alexfmpe>
srk: makes sense, I have some cargo-culted udev rules for other stuff
<alexfmpe>
thanks, will look into that
<{^_^}>
[nixpkgs] @LnL7 pushed to staging-next « libxslt: fix darwin build »: https://git.io/JfQt4
<srk>
alexfmpe: I think dialout group works as well but better to give it a name instead of just /dev/ttyUSB or something
orivej has quit [Quit: No Ping reply in 180 seconds.]
<alexfmpe>
betaboon: where do you get the idVendor / idProduct ?
<betaboon>
alexfmpe: lsusb is your friend :)
<alexfmpe>
oooh I see
gustavderdrache has quit [Quit: Leaving.]
<srk>
or udevadm info
vandenoever has quit [Read error: Connection reset by peer]
vandenoever has joined #nixos
vandenoever has quit [Changing host]
vandenoever has joined #nixos
pingiun has joined #nixos
<pingiun>
I'm trying to do erase your darlings, but having trouble with redis not wanting to start when /var/lib/redis is a symlink
<pingiun>
I set the permissions on /persist/var/lib/redis the same as was previously on /var/lib/redis but the service errors with status=238/STATE_DIRECTORY
mananamenos has quit [Ping timeout: 256 seconds]
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos
<energizer>
pingiun: do you know the names of the files inside?
alp has joined #nixos
<energizer>
alternatively you can bind-mount the directory
pingiun has quit [Ping timeout: 258 seconds]
cosimone has quit [Quit: Quit.]
pingiun has joined #nixos
<pingiun>
I'm sorry for leaving so suddenly my computer crashed
<pingiun>
I was wondering if anybody knows why the redis service refuses to start when /var/lib/redis is a symlink (as described aboev)
<energizer>
pingiun: if you know the names of the files inside you can link those. alternatively you can bind-mount the directory.
<pingiun>
I think redis normally only has one file in the directory but I'm not sure that will always be the case
<pingiun>
so bind mounting seems like the best option then
<energizer>
i want to wrap a program by piping its stderr through another program. is there an idiom for that?
doomlist3 has joined #nixos
whatisRT has quit [Ping timeout: 256 seconds]
zarel has joined #nixos
dongcarl has quit [Ping timeout: 256 seconds]
mmchen has quit [Remote host closed the connection]
<Thra11>
evils: The trouble with just picking stuff at random is that 90% of the time I don't have the hardware to test it on / don't know what it is / don't have the complex config required to use it / don't know how the language or framework is supposed to be packaged / have no interest in it. I don't mind building random stuff and confirming "Yep, seems to run", but I feel like it's more useful (and
<Thra11>
interesting) to review stuff which I use or know how it should work.
<{^_^}>
[nixpkgs] @Ericson2314 opened pull request #90249 → php < 7.4: Fix validation of PKG_CONFIG var → https://git.io/JfQqD
<Thra11>
I wonder if you could take your system config and search for PRs which affect the packages you use.
<energizer>
that would be a nice feature
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos
alexherbo2 has joined #nixos
fragamus has quit [Quit: Connection closed]
alp has quit [Ping timeout: 272 seconds]
<Thra11>
Even just being able to "subscribe" to a package (without having to be the maintainer), so you get notifications about issues and PRs would be good.
pingiun has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<srk>
Thra11: well that's exactly the point of meta.maintainers for both modules and packages, even if you're interested in a package
<MarcWeber>
What is determining the action on lid close ? I'd like to switch to hibernate cause my ryzen7 notebook cannot resume amdgpu from suspnd/s2idle
<Unode>
if I have "nix-build -E 'with import <nixpkgs> {}; enableDebugging package'", how can I translate this to nix-shell? Tried "nix-shell -E ..." but the built package isn't added to PATH.
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Thra11>
srk: I feel like it generates a lot of noise if I add myself as a maintainer to a bunch of packages I'm vaguely interested in (especially when I get too many emails about some of them and want to take myself off again). It's like if you subscribe to an issue on github: I'm not distracting from the issue by just subscribing, unlike if e.g. I had to comment "cc me" to get notifications.
<srk>
I see
zupo has joined #nixos
whatisRT has joined #nixos
codygman has quit [Read error: Connection reset by peer]
<Thra11>
Also, if the author has taken the text from the BSD 3-clause license and added a paragraph of their own, what would I put in meta.license? Possibly licenses.free as it still looks sorta free, but who am I to judge?
<Thra11>
evils++: Cool, thanks
<{^_^}>
evils's karma got increased to 10
slack1256 has quit [Remote host closed the connection]
mallox has quit [Quit: WeeChat 2.8]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<sss>
.|If U believe in the only creator of heaven & earth [the only one] [monothiesm) that moses believed in him and Jesus and mohammed u will be safe *; For he is: (Allah) the only God; our refuge; follow islam The true religion of God & jewel of THE SEAS to be lifted, that is, be true muslimS to be lifted ****]
<sss>
[, that is, be true muslimS to be lifted ****]
waleee-cl has joined #nixos
<{^_^}>
[nixpkgs] @risicle opened pull request #90259 → [r20.03] opencv3, opencv4: use openblasCompat → https://git.io/JfQ3t
marusich has joined #nixos
orivej has quit [Quit: No Ping reply in 180 seconds.]
kleisli has quit [Ping timeout: 256 seconds]
orivej has joined #nixos
<sss>
[Moses believed in Allah as his creator so did Jesus and mohammed *; For he is: (Allah) the only God; our refuge; follow islam The true religion of God & jewel of THE SEAS to be lifted, [that is, be true muslimS to be lifted ****]
is_null has joined #nixos
<quinn>
has anyone developed with elixir + mix on nixos?
<liminal18>
scons: Reading SConscript files ...
<liminal18>
Python 3.5 or greater required, but you have Python 2.7.18
<sss>
[Moses believed in Allah as his creator so did Jesus and mohammed *; For he is: (Allah) the only God; our refuge; follow islam The true religion of God & jewel of THE SEAS to be lifted, [that is, be true muslimS to be lifted ****]
<cransom>
hrm. where's that banhammer when you need it.
<liminal18>
been getting this while trying to build mongo 4.2.5
<FRidh>
update the mongodb expression to use python3
FRidh has quit [Quit: Konversation terminated!]
elibrokeit has quit [Quit: A random quit message]
elibrokeit has joined #nixos
elibrokeit has quit [Client Quit]
<liminal18>
Fridh it should be python 35 actually
kenran has joined #nixos
elibrokeit has joined #nixos
<{^_^}>
[nixpkgs] @maralorn opened pull request #90260 → [haskellUpdates] Fix reflex, reflex-dom and dependencies. → https://git.io/JfQse
<sss>
-[Moses believed in Allah as his creator so did Jesus and mohammed *; For he is: (Allah) the only God; our refuge; follow islam The true religion of God & jewel of THE SEAS to be lifted, [that is, be true muslimS to be lifted ****], [that is, be true muslimS to be HaPPY ****] . M-
muslimm has joined #nixos
<liminal18>
how do I set scons to python3?
muslimm has quit [K-Lined]
muslimm has joined #nixos
<quinn>
anyone who develops applications uses postgresql, do you add a test server in configuration.nix or do you create one with a socket and everything in the source directory
gurmble has joined #nixos
grumble has quit [Killed (tepper.freenode.net (Nickname regained by services))]
gurmble is now known as grumble
muslimm has quit [Max SendQ exceeded]
tomaw has joined #nixos
muslimm has joined #nixos
muslimm has quit [Max SendQ exceeded]
muslimm has joined #nixos
muslimm has quit [Max SendQ exceeded]
kline has joined #nixos
<mgdm>
quinn: I've created them using shell.nix before
<mgdm>
worked well enough for my purposes
<quinn>
mgdm: can you explain with a bit more detail? that sounds like exactly what i want!
ncakerlist has joined #nixos
shabius has quit [Read error: Connection reset by peer]
alp has quit [Remote host closed the connection]
<mgdm>
quinn: sure. One sec while I write it up
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kenran has quit [Ping timeout: 260 seconds]
codygman has quit [Read error: Connection reset by peer]
shabius has joined #nixos
codygman has joined #nixos
muslimm has joined #nixos
muslimm has quit [Max SendQ exceeded]
Fuchs has joined #nixos
muslimm has joined #nixos
codygman has quit [Read error: Connection reset by peer]
<mgdm>
I'm sure I could do the whole lot in shell.nix but that was neough to get me going
codygman has quit [Ping timeout: 260 seconds]
codygman has joined #nixos
<ConradMearns>
Noob here, I'm trying to make a nixpkg for SeaweedFS and I've got it to build, but I can't seem to launch the binary after using nix-shell. I was wondering if anyone would be willing to take a look at my default.nix and offer suggestions?
orivej has quit [Quit: No Ping reply in 180 seconds.]
codygman has quit [Ping timeout: 264 seconds]
codygman has joined #nixos
Negher has quit [Ping timeout: 256 seconds]
sss has quit [K-Lined]
orivej has joined #nixos
Fuchs has left #nixos ["Part"]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<astylian>
Hello all!! I am thinking to switch to NixOS and I have two questions. 1. Is NixOS a rolling release? If not how difficult is to keep up with the latest Kernel? 2. Does it supports SELinux? If yes, how well is supported?
<notgne2>
astylian: keeping your kernel updated is as simple as having `boot.kernelPackages = pkgs.linuxPackages_latest` in your config, I don't think SELinux is fully supported but I may be wrong
<simpson>
astylian: 1. Yes, somewhat; it is easy enough to be stable and to not update kernels, but also straightforward to get new kernels packaged within a week of upstream. 2. Not really; what did you need from it?
<cole-h>
1) nixos-unstable is rolling, but we also have standard release branches (current is 20.03). Keeping up with the kernel on a release branch is as easy as copying in the derivation and using that instead.
<{^_^}>
[mobile-nixos] @samueldr pushed 2 commits to master: https://git.io/JfQGz
<astylian>
simpson, about SELinux, I just want to have a system that runs SELinux. lets say is a requirment :)
knupfer has quit [Ping timeout: 258 seconds]
<astylian>
and what about nvidia drivers, is it a big hassle?
MightyJoe has joined #nixos
<simpson>
astylian: I think that, very technically, the basic parts of SELinux can be enabled. However, policy information isn't copied into the Nix store, and there's no userspace support. Do you have policies for your custom applications, or custom policies for standard userland?
<simpson>
nVidia and nouveau drivers are both available.
<simpson>
ConradMearns: I just finished building the derivation you posted. It seems to work; I know nothing of this daemon, but I get a result/bin/weed and it responds to `weed --help`.
cyraxjoe has quit [Ping timeout: 256 seconds]
<infinisil>
astylian: I've seen people use apparmor and audit on NixOS, maybe those could be enough for you (unless you specifically need SELinux which I suspect)
<astylian>
simpson, I have seen that, but I am warried what happens in practice
<simpson>
cole-h++ for historical context.
<{^_^}>
cole-h's karma got increased to 68
<cole-h>
1 more karma... >:)
KarlJoad has joined #nixos
noreferences has quit [Quit: Connection closed]
<simpson>
astylian: I don't run nVidia, but in practice, I hear that older cards work alright with nouveau, and newer cards work alright with nvidia, and nvidia's still just as painful to debug as it ever has been.
<cole-h>
simpson: What is a "newer" card? 1080? Or the 2000 series and newer?
<cole-h>
Because nouveau works just fine (for my mostly non-gaming workloads) on my 1080
<cole-h>
(Minecraft works fine)
<mgdm>
quinn: if you reload that page you'll see how I got it to work
<simpson>
cole-h: Shifting; I normally imagine a half-decade sliding window. I never have had a strong grasp of which cards we fully understand, and which cards we don't really grok yet but kind of know how to boot and program. All I can say is that laptop/integrated units tend to lag behind discrete chipsets in terms of support, 'cause the laptop chips are often cut down for power-saving.
<astylian>
simpson, thats my basic usage as well, I just don't want to find my self spending 2+ days on installing a linux distro that seems nice but I am unfamiliar with it
<KarlJoad>
Or, is this something that can only be done by using `nix-env` and `nix-channel`
<quinn>
mgdm: do you mean ~mgdm/psql-shell.md.txt?
<simpson>
astylian: Good news! You will only spend ~30min installing NixOS. You will still be confused, because Nix is quite different from classical package managers, but you can be confused *faster*.
<mgdm>
quinn: yep
<quinn>
mgdm: i think this will work, tyvm
<notgne2>
this reminds me, iirc nixos actually has a slight annoyance when it comes to nvidia drivers
<colemickens>
KarlJoad: a number of us do this. You can do it by setting NIX_PATH to a specific nixpkgs checkout, or a specific tar.gz of a specific commit
<mgdm>
quinn: works on my machine (TM) at least :)
<notgne2>
rebuild switch will actually put the new driver files into place in /run, even though they depend on a newer kernel module being loaded
<quinn>
mgdm: FYI you can add at least some of it to shellHook
<notgne2>
making it so you have to reboot your system to get graphical applications working again
<notgne2>
this only happens once every few months though
<simpson>
KarlJoad: system.autoUpgrade.channel can be used to declare the system channel. I'm not sure if it's possible to have a *list*, or probably an attrset, of channels.
<colemickens>
KarlJoad: alternatively, you can write Nix files that use builtins.fetchTarball to retrieve a fixed nixpkgs and then build a system configuration with it
<mgdm>
quinn: I intend to experiment with that once I figure out how to detect if it's already running, etc
orivej has quit [Ping timeout: 256 seconds]
Zetagon has quit [Quit: WeeChat 2.8]
<notgne2>
I don't see why you can't juse use `<unstable>` in place of `fetchTarball ...`, and then add unstable using nix-channel
<KarlJoad>
simpson: Then what is the proper syntax for switching your `system.autoUpgrade.channel`? For the 20.03 release train, is it "20.03-unstable" or just "unstable"?
magnetophon has quit [Ping timeout: 264 seconds]
<ConradMearns>
simpson: How do I include the package into my system configuration? I have also found the executable in my nix-store and results folder
<KarlJoad>
notgne2: It's not a matter of not being able to do it that way, but I want to minimize the amount of things I have to do to get a system up and running. I want to be able to just download the system config files and do a nixos-install and be done.
<cole-h>
Yes, nixos-20.03 is the 20.03 channel (or nixos-20.03-small if you want faster updates but potentially more rebuilding for yourself)
<simpson>
ConradMearns: Like, into environment.systemPackages? You'd incorporate the derivation into your /etc/nixos/configuration.nix.
<notgne2>
KarlJoad: oh that makes sense, I definitely don't enjoy typing out the nix-channel commands on new installs, twice...
<KarlJoad>
colemickens: I have done that, but it's just one package that I use that is out of date in the 20.03 release. So long as that package gets above a 1.4 release, I don't care about the minor version.
<simpson>
I am messy, so I'd go with `{ ... }: let myPkg = pkgs.stdenv.mkDerivation { ... }; in { environment.systemPackages = [ myPkg ... ]; ... }` etc.
gustavderdrache has joined #nixos
vpfeiffer[m] has joined #nixos
<colemickens>
I don't really understand what you mean about a minor version. By default the nix tooling, etc, expects you're using channels.
<colemickens>
If you want, you can have your system follow nixos-20.03, and then optionally install a single package from nixos-unstable.
<KarlJoad>
colemickens: I want the mu package to be above version 1.4. The standard nixos-20.03 channel is stuck on 1.2 right now. The update to mu to 1.4 happened after the 20.03 release, so I need to bump that version up. I have done it manually, with a nix expression to fetch the repo and build. But the unstable channel is also up-to-date.
<colemickens>
KarlJoad: if you're using channels, you can do something like `nix-channel --add https://nixos.org/channels/nixos-unstable unstable`, `nix-channel --update` `nix-env -iA unstable.mu`
<{^_^}>
[nixpkgs] @jonringer pushed commit from @siriobalmelli to master « vimPlugins: update »: https://git.io/JfQGN
<colemickens>
If you want to keep everything declarative and avoid `nix-env`, you can do the equivalent in a nix file and pull the `mu` package in from a different nixpkgs than the system is built with.
<colemickens>
KarlJoad: for some reason, newer versions of mongodb segfault on rpi4, so that file pulls in an old nixpkgs version, then references the mongodb package from it. Everything else on that machine is built with a different nixpkgs.
<colemickens>
that reddit post looks roughly the same, yeah
<KarlJoad>
cole-h: Honestly, I am a bit unsure what the difference is between the 2 channels, other than the fact that just Nix will not use nixos channels.
<KarlJoad>
colemickens: Ok. That's what I figured. There doesn't seem to be a good way to declaratively change channels.
<cole-h>
nixos-20.03 -> release (think of it as an Ubuntu release, if that helps). nixos-unstable -> rolling release (think of it as Arch, if that helps).
smatting has quit [Ping timeout: 256 seconds]
<colemickens>
hm, imo this is pretty much what I'd expect it to look like. it seems like the minimal amount of info one would need to fully declaratively pull in your deps.
<KarlJoad>
cold-h: I understood that part. As your question asked, I am not sure what the difference is between nixpkgs-unstable and nixos-unstable.
<colemickens>
flakes will make this a bit easier and built-in feeling though.
<cole-h>
Oh
agsdheidjd has quit [Ping timeout: 256 seconds]
<colemickens>
AFAICT: nixos-unstable is more conservative, it has to pass nixos verification tests that nixpkgs-unstable doesn't have to pass.
<{^_^}>
[nixpkgs] @jonringer pushed commit from @danieldk to master « python3Packages.somajo: 2.0.5 -> 2.0.6 »: https://git.io/JfQZJ
<cole-h>
nixpkgs-unstable is just packages; it updates when a certain core package set succeeds certain tests. nixos-unstable waits for more NixOS-specific tests to succeed as well.
<cole-h>
At least that's my understanding.
coderobe has joined #nixos
jgeerds_ has quit [Ping timeout: 260 seconds]
<KarlJoad>
Ok, that clears it up a bit. Either way, this package is out-of-date on the 20.03 channel. I have no problems switching this VM to a completely unstable channel to run. At that point, the question is, how do I change the system's channel, preferably declaratively?
pamplemousse has joined #nixos
kline has left #nixos ["Leaving"]
<{^_^}>
[nixpkgs] @prusnak opened pull request #90261 → images: remove original files when using zstd for compression → https://git.io/JfQZO
<ConradMearns>
simpson: incorporating the derivation into any kind of environment is where I am currently stuck
<mgdm>
quinn: expanded that into a blog post now, thanks for the inspiration :-)
<ConradMearns>
I'm trying to just get it to work in nix-shell, I have a let `seaweedfs = pkgs.callPackage ./seaweedfs/default.nix {};` wrapped around a stdenv.mkDerivation but I get errors about seaweedfs being a set, not a function
<simpson>
ConradMearns: callPackage expects to call expressions which are functions that take various pkgs as arguments. `{ stdenv, fetchFromGitHub, buildGoModule }: ...` sort of situation.
_ris has joined #nixos
ris has quit [Ping timeout: 258 seconds]
<Thra11>
If a package has vendored dependencies (eww), should we try to check their licenses when packaging it? A package might e.g. be BSD licensed, but have a vendored dependency with an unfree license, in which case the "thing" that we end up building is unfree, but risks being labelled as BSD licensed.
vandenoever has quit [Read error: Connection reset by peer]
<n3t>
Time to make an alternative for Nix and call it "Nut". Then make a blog post titled "PostgreSQL in a nut-shell".
<quinn>
mgdm: oho, very cool. bookmarked. funny enough i'm actually using this stuff because i'm trying to write a blog in phoenix/elixir
whatisRT has left #nixos ["ERC (IRC client for Emacs 27.0.90)"]
seku57 has joined #nixos
<mgdm>
n3t: :D
<mgdm>
quinn: nice! I had a play with elixir a while ago but never really used it for anything interesting
<mgdm>
I may have another look
seku has quit [Ping timeout: 258 seconds]
<quinn>
mgdm: i'm a FP guy, so for me it's just appealing because dynamic site + fp + has a markdown parser/templating engine
<quinn>
i draft and sometimes write papers in markdown and export through pandoc, so the idea of having a templating language where i can just drop markdown files into a folder and have a working blogpost from an article is very appealing
<ConradMearns>
simpson: i'm still confused, should I change my seaweedfs default.nix? I thought that it was formatted correctly, so that it could eventually be added to nixpkgs. I don't understand how to get the derivation to connect with nix-shell, I can get everything to run if I just put ./seaweedfs in by buildInputs, but then I can't use the app
<quinn>
ConradMearns: nix-shell (unless you're using it with -p) is generally for making a shell with the inputs for building a project. you probably want to run nix-build and then run the binary from the result folder that appears
<ConradMearns>
my goal is to include the application in another machine's configuration, i figured that if I could get the derivation to work with that, then pushing it to my other machine would be a breeze
<ConradMearns>
but there's probably a lot that i'm misunderstanding
<quinn>
ConradMearns: you want to test that it works with nix-build probably, then you can add a package overlay to the other machine's configuration.nix, and once that is set then you can add your package to environment.systemPackages in that same configuration.nix
seku57 is now known as seku
pamplemousse has quit [Ping timeout: 265 seconds]
<euank>
ConradMearns: You can format it akin to how you'd want it in nixpkgs (i.e. have `pkgs/seaweedfs/default.nix` which has at the top `{ stdenv, fetchurl, dependencies, etc }:` (which is suitable for callPackage). You indeed now can't point nix-shell at that. However....
<euank>
You can then have an additional file that will not be suitable for upstream which glues nixpkgs to your package. Specifically, it would be something like `let nixpkgs = import <nixpkgs> {}; in { seweedfs = nixpkgs.callPackage ./pkgs/seaweedfs {}; }`
<amanjeev>
is anyone else seeing issues with python2.7-apsw derivation? My upgrade fails: builder for '/nix/store/nlhqm39177wxk9ll6072im7nm3haiv2b-python2.7-apsw-3.30.1-r1.drv' failed with exit code 1
<euank>
At that point you could point nix-build or nix-shell or such at that file you don't intend to upstream (shell.nix or such), and that file would specifically be there as the place where you plumb the nixpkgs dependency into your package.
civodul has quit [Quit: ERC (IRC client for Emacs 26.3)]
<jonreeve[m]>
Does anyone know what it means on this wiki page: https://nixos.wiki/wiki/Qt that "The package will need to be fixed to use [the new https://github.com/NixOS/nixpkgs/issues/65399 wrapQtAppsHook]. The hook wraps every qt application with adding QT_PLUGIN_PATH and XDG_DATA_DIRS as well as XDG_CONFIG_DIRS.See wrap-qt-hook.sh in nixpkgs" ?
<fresheyeball>
but not establish a passive sftp connection
mbrgm_ has joined #nixos
mbrgm_ is now known as mbrgm
nek0 has quit [Remote host closed the connection]
<euank>
so if you can use libsForQt5.callPackage, that _might_ do the right thing already. If you can't use that, you may have to dig around a bit more. That's about the depth of my qt5-in-nix knowledge
<euank>
sshow: you could; have you tried to yet and run into an issue?
pamplemousse has joined #nixos
<sshow>
euank: I'm not sure how to tar something in nix. and I'm not even sure that's a thing :P
<sshow>
in my head I'm thinking `archive = tar fetchGit {...}`
nek0 has joined #nixos
<jtojnar>
euank it is actully done by qt5.mkDerivation (which gets passed as mkDerivation argument by libsForQt5.callPackage)
<euank>
sshow: the naive way that comes to mind for me would be to build two derivations. You could use the runCommand derivation helper to make that pretty easy. So something like so... (one sec)
<sshow>
fwiw, I'm building a package with node2nix, but I'm trying to build the package using remote sources :)
<euank>
sshow: something like `gitSrc = fetchgit ....; tarSrc = runCommand "tar-src" {} "${pkgs.gnutar}/bin/tar -cf $out ${gitSrc}"` I think would work. Untested though.
<euank>
If the thing comes from github, btw, you can use the archive url (github.com/.../archive/commit.tar.gz) + fetchurl to get a tarball to begin with.
<euank>
And other git servers also sometimes provide commit archives for download.
<sshow>
euank: thanks. i'll give it a go. sources are from private git server :)
rajivr has joined #nixos
nek0 has quit [Remote host closed the connection]
<euank>
jonreeve[m]: iiuc, usually setup hooks just end up being in `nativeBuildInputs`, and apply that way, rather than explicitly being used via shellHook
magnetophon has joined #nixos
user_0x58 has quit [Remote host closed the connection]