Ariakenom has quit [Read error: Connection reset by peer]
sheeldotme has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sheeldotme has joined #nixos
<{^_^}>
[nixpkgs] @grahamc pushed to revert-59945-fix/ova-eval-failure « Revert "nixos/virtualbox: Fixes configuration to evaluate" »: https://git.io/fj3Tp
<{^_^}>
[nixpkgs] @grahamc opened pull request #59994 → (do not merge) Revert "nixos/virtualbox: Fixes configuration to evaluate" to ensure ofborg checks this properly → https://git.io/fj3Th
jluttine has quit [Ping timeout: 244 seconds]
Mr_Keyser_Soze has quit [Ping timeout: 246 seconds]
Rusty1 has joined #nixos
vk3wtf has joined #nixos
Acou_Bass has quit [Quit: byeeeeeeeeeeeeeee]
Acou_Bass has joined #nixos
wfranzini has joined #nixos
<parsnip>
is there a canonical file like /etc/nixos/configuration.nix if you're not on nixos?
<parsnip>
or, what exactly is nix-darwin?
jluttine has joined #nixos
<parsnip>
maybe i should try jumping in
myskran has joined #nixos
winem_ has quit [Ping timeout: 252 seconds]
<jasongrossman>
linarcx: I like it. It makes things discoverable.
<jasongrossman>
linarcx++
<{^_^}>
linarcx's karma got increased to 2
<jasongrossman>
linarcx: I'm not seeing it in unstable though.
jluttine has quit [Ping timeout: 255 seconds]
<{^_^}>
[nixpkgs] @matthewbauer pushed to master « aws-c-event-stream: add libexecinfo on musl »: https://git.io/fj3kL
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/fj3kt
kleisli_ has joined #nixos
qubasa_ has quit [Read error: Connection reset by peer]
<oborot>
I see this syntax used in the release notes, but can't get it to work.
<oborot>
Oh wait, I think I got it
<oborot>
Need to add lib to the {...} import at the top
<infinisil>
kleisli_: It's all tags and commits
<lordcirth>
Trying to run nix-shell /nixpkgs/pkgs/tools/filesystems/ceph/generic.nix in order to fix it, but it won't load, saying "error: cannot auto-call a function that has an argument without a default value ('stdenv')"
<lordcirth>
What arguments does stdenv require?
<infinisil>
lordcirth: You changed it and want to build the changed package?
<lordcirth>
infinisil, yes
<lordcirth>
Or get started fixing it, at least. It's been abandoned a long time
<infinisil>
lordcirth: The file nixpkgs/pkgs/top-level/all-packages.nix defines a mapping from attribute name to file, which is where you'll find ceph referenced
<infinisil>
lordcirth: To build an attribute in nixpkgs, you need to go to the root, and then run `nix-build -A <attribute>`, so `nix-build -A ceph` in this case
<lordcirth>
infinisil, ah, ok, thanks!
<lordcirth>
And now I just have to let make -j 4 run forever...
acarrico has quit [Ping timeout: 244 seconds]
<{^_^}>
[nixpkgs] @jonafato opened pull request #59996 → gnomeExtensions.battery-status: mark as not broken → https://git.io/fj3k0
ris has quit [Ping timeout: 255 seconds]
shibboleth has quit [Quit: shibboleth]
justanotheruser has quit [Ping timeout: 245 seconds]
<oborot>
Anybody familiar with the new programs.firejail.wrappedBinaries option? Is there a way I can verify that the binaries are actually running in firejail?
<oborot>
When I do a which on a binary, I can't really see any indication that it's being wrapped.
hellrazo1 has joined #nixos
acarrico has joined #nixos
<{^_^}>
[nixpkgs] @grahamc closed pull request #59994 → (do not merge) Revert "nixos/virtualbox: Fixes configuration to evaluate" to ensure ofborg checks this properly → https://git.io/fj3Th
<{^_^}>
[nixpkgs] @grahamc pushed 0 commits to revert-59945-fix/ova-eval-failure: https://git.io/fj3ka
jluttine has quit [Ping timeout: 255 seconds]
_fw has quit [Read error: Connection reset by peer]
hellrazor has quit [Ping timeout: 246 seconds]
Acou_Bass has quit [Quit: byeeeeeeeeeeeeeee]
Acou_Bass has joined #nixos
jluttine has joined #nixos
buffet_ has joined #nixos
buffet has quit [Ping timeout: 255 seconds]
bikes has joined #nixos
kreisys has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kisonecat has quit [Ping timeout: 255 seconds]
romildo has quit [Quit: Leaving]
kreisys has joined #nixos
<worldofpeace>
,locate bin overlayroot-disable
<parsnip>
ah nice, nix-darwin makes something like a /etc/nixos/configuration.nix
<parsnip>
that is, ~/.nixpkgs/darwin-configuration.nix
Acou_Bass has quit [Quit: byeeeeeeeeeeeeeee]
Acou_Bass has joined #nixos
alex``` has joined #nixos
fl0_id has quit [Read error: Connection reset by peer]
kisonecat has joined #nixos
fl0_id has joined #nixos
gagbo has joined #nixos
fl0_id has quit [Read error: Connection reset by peer]
silver has quit [Read error: Connection reset by peer]
stepcut has quit [Remote host closed the connection]
stepcut has joined #nixos
stepcut has quit [Remote host closed the connection]
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
<{^_^}>
[nixpkgs] @andersk opened pull request #60002 → pytorch: Move cudatoolkit to nativeBuildInputs → https://git.io/fj3LJ
InFlames has quit [Ping timeout: 268 seconds]
orivej has quit [Ping timeout: 268 seconds]
mkoenig has quit [Ping timeout: 255 seconds]
<{^_^}>
[nixpkgs] @worldofpeace merged pull request #59968 → guake: fix by setting strictDeps to false → https://git.io/fjOhK
<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/fj3L8
<{^_^}>
[nixpkgs] @worldofpeace pushed commit from @markuskowa to release-19.03 « guake: fix by setting strictDeps to false »: https://git.io/fj3LB
tomberek has quit [Ping timeout: 246 seconds]
bikes has quit [Read error: No route to host]
mkoenig has joined #nixos
<kleisli>
hey all, i'm trying to write a shell.nix for a project that uses nodejs among other things
kleisli_ has joined #nixos
<kleisli_>
sorry, got disconnected
o1lo01ol1o has quit [Remote host closed the connection]
<kleisli_>
so i'm trying to combine the output of node2nix with an already existing shell.nix, but for some reason node isn't in my $PATH in the resulting shell
<lewo>
parsnip: Only Docker images with Nix are available. What are you expecting?
<jomik>
Hey guys - I am having trouble with a light-locker systemd service. It seems that it does not get the XDG_SESSION_ID when I start it as a service. How do I pass that along with systemd?
<parsnip>
lewo: nixos
<parsnip>
lewo: hmm, upon reflection, i realize my original question was, "where did the /etc/nixos/configuration.nix go"
<parsnip>
so i probably just need to put stuff in .nix-profile or something
Makaveli7 has joined #nixos
<parsnip>
that is, i'd like to add a nix configuration file adjacent to a dockerfile, that easily configures a docker image for troubleshooting things
Jetien_ has joined #nixos
jmgai has quit [Ping timeout: 250 seconds]
lovesegfault has joined #nixos
jmgai has joined #nixos
Jetien has quit [Ping timeout: 245 seconds]
<lewo>
parsnip: to configure image at buildtime or once the image is running?
<lewo>
parsnip: (the latter is much more complicated)
<parsnip>
first and foremost/ moreso at build time
<parsnip>
does anyone go over what appears to be a paradigm shift with this whole nix-shell, etc stuff
<yorick>
jomik: systemctl --user import-environment somewhere after starting x
<jomik>
Oh, I got some debugging output ! `PID 1913 does not belong to any known session`
<jomik>
So I guess light-locker can't be run by systemd :D
val_ has joined #nixos
<enteee>
I can't disable autoindent in vim using home-manager. I have added nocindent, noautoindent, nosmartindent to programs.vim.extraConfig. But this didn't help so far. Any Ideas how I can get rid of all indentation support in vim?
<jomik>
enteee: Maybe you can set indentexpr to nothing.
jomik has quit [Client Quit]
jasongrossman has quit [Remote host closed the connection]
<enteee>
jomik: did that as well didn't help
<enteee>
i als have filetype indent off and filetype plugin indent off
risson_ has joined #nixos
risson_ has quit [Client Quit]
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos
leat has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @lightbulbjim opened pull request #60008 → librecad: add .desktop file and icon → https://git.io/fj3q5
leat has joined #nixos
zupo has joined #nixos
orivej has quit [Ping timeout: 250 seconds]
m0rphism has quit [Ping timeout: 250 seconds]
<val_>
hey guys
<val_>
how can i download 18.9 version iso?
<val_>
i have problems with 19.3 so i wanna check the previous version
<val_>
i know there's hydra and other stuff, couldn't figure it out
<parsnip>
hmm, so i cloned the olm repo, i ran cmake, make install, etc.
<parsnip>
then `cd python` and `make olm-python2`, and now i get the error about setuptools, and it seems i need to understand how to use python libraries
<parsnip>
and nix environments and derivations, etc
<parsnip>
i guess i can write a nix file? but where?
<parsnip>
weird, even if i do
<parsnip>
nix-shell -p 'python35.withPackages(ps: with ps; [ setuptools ])'
Jetien__ has quit [Remote host closed the connection]
Jetien__ has joined #nixos
<linarcx>
lassulus: Thank you. but when i want to install it, i get an error: warning: name collision in input Nix expressions, skipping '/home/linarcx/.nix-defexpr/channels_root/nixos'
<teto>
my nix-shell is stuck at "instantiated 'builder.pl' -> '/nix/store/zzc2a76p5zsinr9cgq5f55w3nw2dw7sm-builder.pl.drv'" with latest nixos-unstable . (dunno if it's linked, I've just updated). Any idea ?
<lassulus>
linarcx: huh, weird, what exactly did you do?
tmaekawa has joined #nixos
jasongrossman has quit [Ping timeout: 255 seconds]
<linarcx>
lassulus: nix-env -iA nixos.yarn2nix
<teto>
with more verbosity I see "download thread waiting for 10000 ms" x(
<lassulus>
linarcx: does nix-shell -p yarn2nix work?
<lassulus>
linarcx: well, that error seems to bee intended, so I guess you should use upstream. If you don't want to clone the repo, you can install it like this: nix-env -I yarn2nix=https://github.com/moretea/yarn2nix/archive/master.tar.gz -f '<yarn2nix>' -i yarn2nix -i yarn
tboston_ has quit [Ping timeout: 250 seconds]
jmgai has quit [Ping timeout: 246 seconds]
val_ has quit [Quit: Page closed]
jboy has quit [Remote host closed the connection]
Anton-Latukha has joined #nixos
jmgai has joined #nixos
<parsnip>
hmm, i have python libraries installed globally, so shouldn't any running python have those libraries available?
Jetien__ has quit [Remote host closed the connection]
<simpson>
It's a fun theory, but you should probably explicitly bring packages to any Python that wants them. What are you building?
<parsnip>
oh, maybe you mean withPlugins to clean up the code a bit
tboston_ has quit [Ping timeout: 246 seconds]
<edcragg>
hi, has anyone here set up roundcube on nixos before?
<edcragg>
it's pretty trivial to install (on unstable), it even sets up nginx virtual host and postgres, however i'm not sure how to add any users to be able to log in and use it. also, the web based installer at /installer as mentioned on the roundcube docs appears to be a 404
<linarcx>
lassulus: Oh, thank you. it take some times to install! Can we hope that it fix in near future?
jmgai has quit [Ping timeout: 268 seconds]
lopsided98 has quit [Remote host closed the connection]
Ariakenom has quit [Quit: Leaving]
lopsided98 has joined #nixos
zupo has joined #nixos
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<siraben>
I don't understand why I get an attribute "latest" missing error when I try to invoke nix-shell with this as my shell.nix http://ix.io/1GR6
<siraben>
Is RUST_SRC_PATH = "${pkgs.latest.rustChannels.stable.rust-src}/lib/rustlib/src/rust/src"; not a valid expression?
<lassulus>
siraben: pkgs does not have a latest attribute
<lassulus>
maybe you want to add the mozilla overlay to your shell.nix
<linarcx>
lassulus: warning: name collision in input Nix expressions, skipping '/home/linarcx/.nix-defexpr/channels_root/nixos'
<siraben>
Also I'm getting errors from my autocompleter regarding paths: Company: backend company-capf user-error: No such directory: /nix/store/hcszvp7a7y5cz6s15cii98yhxxfq7zy4-rustc-1.33.0/lib/rustlib/src/rust/src. Please set ‘racer-rust-src-path’ or ‘RUST_SRC_PATH’
<siraben>
So will setting it in shell.nix fix this?
<lassulus>
linarcx: shouldn't be a problem I guess. But to be honest, I haven't used nix-env in a long time.
silver has joined #nixos
<{^_^}>
[nixpkgs] @tfc opened pull request #60016 → tup: Let tup passthrough NIX env vars → https://git.io/fj3YD
<Izorkin>
etu: if merged PR #59904 without oci8, need reconfigure PR #59926
risson has quit [Remote host closed the connection]
risson has joined #nixos
winem_ has joined #nixos
<infinisil>
siraben: did you install the rust overlay?
<siraben>
infinisil: doing it right now
<gyroninja>
is it possible to have nix build not query NixOS's binary cache for a personal project
<siraben>
so I should remove rustc, rustfmt, racer from my configuration.nix?
orivej has joined #nixos
<siraben>
and use nix-shell instead and launch emacs from within it?
<siraben>
infinisil: it works now!
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<siraben>
so I'm just confused about best practices, now I have to launch a separate emacs instance for haskell projects, and another for a rust project, can I have both?
[Leary] has joined #nixos
Lears has quit [Ping timeout: 246 seconds]
<adisbladis>
siraben: A nice solution is direnv + direnv-emacs
aanderse has quit [Remote host closed the connection]
aanderse has joined #nixos
jmgai has quit [Ping timeout: 255 seconds]
jmgai has joined #nixos
jmgai has quit [Ping timeout: 250 seconds]
<gchristensen>
a new r13y report is up after a while of it not generating properly: https://r13y.com/ 1278 / 1295 (98.69%) are reproducible. if you're on holiday toady, consider spending a bit on making one of them reproduce :)
hellrazo1 has quit [Ping timeout: 255 seconds]
hellrazo1 has joined #nixos
jmgai has joined #nixos
Xal has quit [Ping timeout: 268 seconds]
o1lo01ol1o has joined #nixos
Xal has joined #nixos
kisonecat has quit [Ping timeout: 244 seconds]
Glider_IRC_ has joined #nixos
jmgai has quit [Ping timeout: 250 seconds]
Glider_IRC__ has quit [Ping timeout: 245 seconds]
jmgai has joined #nixos
<ashkitten>
these numeronyms are awful
o1lo01ol1o has quit [Remote host closed the connection]
<ashkitten>
wow
<gchristensen>
haha
<gchristensen>
what's up, ashkitten?
o1lo01ol1o has joined #nixos
<ashkitten>
oh just complaining to nobody that a11y isn't an accessible word
<ashkitten>
among other things
<gchristensen>
I mean the "wow" part :)
o1lo01ol1o has quit [Remote host closed the connection]
<ashkitten>
i'm in awe of how much mental effort it takes to decode sentences with numeronyms
<gchristensen>
+1
<MichaelRaskin>
I guess with replaceDependency we could have testability: count the number of packages such that replacing them with nonsense fails at least one NixOS test
<ashkitten>
honestly if the programming world could do away with numeronyms altogether that'd make me very happy
<MichaelRaskin>
I am kind of scared to even approximate the amount of computing power that might need, given that each run of a test has quite a lot of boot overhead
<gchristensen>
MichaelRaskin: it would certainly be A Lot
<ashkitten>
is there any way to reduce runtime overhead of tests?
<gchristensen>
even if we assume a test takes 1s, it would still be A Lot
<MichaelRaskin>
On the other hand, each _package_ would be built only once, and most packages are not covered by any test anyway…
fendor_ has joined #nixos
<ashkitten>
i suppose it's hard to cut corners when testing reproducibility
<MichaelRaskin>
And hopefully, many packages only need _one_ test…
<simpson>
ashkitten: To the contrary, it seems like it's quite easy for the typical package author to do, this cutting of corners~
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<ashkitten>
i'm spicy every morning, but i don't have ffxiv today
o1lo01ol1o has joined #nixos
<simpson>
I'm procrastinating rewriting a chunk of a compiler. It's easier to be silly on IRC than to get stuff done.
jmgai has joined #nixos
coconnor has quit [Ping timeout: 244 seconds]
kvda has joined #nixos
coconnor has joined #nixos
orivej has quit [Ping timeout: 245 seconds]
<ashkitten>
i don't really have anything to do till tomorrow when i get packing stuff and get working on that
<ashkitten>
worst time for ffxiv to have datacenter maintenance tbh
o1lo01ol1o has quit [Ping timeout: 255 seconds]
fendor has joined #nixos
carlosdagos has quit [Read error: Connection reset by peer]
rauno has joined #nixos
<gchristensen>
`waf` is a real devil of a package
carlosdagos has joined #nixos
<{^_^}>
[nixpkgs] @aanderse opened pull request #60019 → nzbget: fix broken service, as well as some improvements → https://git.io/fj33O
<gchristensen>
oh cool we broke 60k
jmgai has quit [Ping timeout: 250 seconds]
<lordcirth>
That's a lot of PRs!
o1lo01ol1o has joined #nixos
orivej has joined #nixos
<lordcirth>
So, I changed the version tag to download in ~/nixpkgs/pkgs/tools/filesystems/ceph/default.nix, (to 14.2.0 Nautilus) but when I ran nix-build -A ceph in ~/nixpkgs, it created a derivation with 14.2.0 in the name, but ceph --version says luminous!
tboston__ has quit [Ping timeout: 255 seconds]
zupo has joined #nixos
npmccallum has quit [Quit: npmccallum]
o1lo01ol1o has quit [Remote host closed the connection]
<clever>
lordcirth: if you leave the sha256 the same, thats claiming that the src is unchanged, so nix uses the old src
o1lo01ol1o has joined #nixos
<lordcirth>
clever, ah, so to do Trust on First Use, I need to change the hash to something random, so it doesn't cache?
zupo has quit [Client Quit]
<clever>
yeah, i set my editor to replace mode, and then replace a few digits with 0's
<clever>
tofu only works if the first hash is invalid
romildo has quit [Quit: Leaving]
<lordcirth>
clever, in hindsight that makes sense. Thanks!
<lordcirth>
Ah, It actually showed the git clone this time, great
<clever>
:D
<{^_^}>
[nixpkgs] @bebehei opened pull request #60020 → spotify: add libnotify as dependency → https://git.io/fj33l
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
<gchristensen>
,tofu
<{^_^}>
To get a sha256 hash of a new source, you can use the Trust On First Use model: use probably-wrong hash (for example: 0000000000000000000000000000000000000000000000000000) then replace it with the correct hash Nix expected.
wolfman154 has joined #nixos
winem_ has quit [Ping timeout: 250 seconds]
<{^_^}>
[nixpkgs] @grahamc opened pull request #60022 → waf: make reproducible → https://git.io/fj33K
<gchristensen>
anyone available to test this PR? fetch the PR, run grep REVISION= $(nix-build . -A waf )
kisonecat has joined #nixos
<aanderse>
gchristensen: on it
<wolfman154>
Good morning, yesterday I setup xautolock, it rebuilt without any errors, but when I tested this morning, it is not working, I’m not using bo desktop environment, just xmonad?
<gchristensen>
thanks, aanderse!
<wolfman154>
Should I setup, autolock in home.nix instead of configuration.nix?
<aanderse>
gchristensen: np. REVISION="492eeeb1294054d82d4520c9c5990db3"
<aanderse>
care for a pastebin of full output?
<gchristensen>
yay!
<gchristensen>
nope, that is conclusive :)
<gchristensen>
thank you!
<exarkun>
`nix search gtimelog` shows me `nixpkgs.gtimelog` but `nix-env -i gtimelog` says gtimelog matches no derivations, what's the deal?
<{^_^}>
[nixops] @AmineChikhaoui merged pull request #1134 → do: use region ams3 in example → https://git.io/fjOXR
<{^_^}>
[nixops] @AmineChikhaoui pushed 2 commits to master: https://git.io/fj33X
<wolfman154>
Good morning, yesterday I setup xautolock, it rebuilt without any errors, but when I tested this morning, it is not working, I’m not using a desktop environment, just xmonad?
<wolfman154>
Should I setup, autolock in home.nix instead of configuration.nix?
o1lo01ol1o has quit [Remote host closed the connection]
<gchristensen>
wolfman154: please wait a bit longer, not many people have changed and it was only a few minutes ago.
<wolfman154>
gchristensen: 👍
<gchristensen>
thanks :)
tboston_ has quit [Ping timeout: 258 seconds]
orivej has quit [Remote host closed the connection]
orivej has joined #nixos
kisonecat has quit [Quit: leaving]
InFlames has joined #nixos
<wolfman154>
What is the setting that makes my screen go black, after being idle in nixos?
ddellacosta has joined #nixos
<ar>
wolfman154: if you're not using a full DE (say, KDE or gnome), it's probably logind. man logind.conf | less -p IdleAction
<toppler>
Hey. Can any EXWM folks tell me about their setup? In particular, how do you use per-project nix-shells and use emacs specifically within those shells? Is it terrible to launch an emacs as a child of EXWM's Emacs?
linarcx has quit [Ping timeout: 255 seconds]
<pie_>
Profpatsch, ? ^
linarcx has joined #nixos
<symphorien>
not an emacs user myself, but I have heard about a direnv plugin
gagbo has quit [Ping timeout: 255 seconds]
<{^_^}>
[nixpkgs] @roosemberth opened pull request #60024 → libguestfs-with-appliance: Add package for libguesfs containing libguestfs appliances → https://git.io/fj3G3
fendor_ has quit [Remote host closed the connection]
o1lo01ol1o has quit [Remote host closed the connection]
<pie_>
is anyone other than us stress testing github? :P
<drakonis1>
people cloning the linux repository?
<pie_>
(ok maybe some big private corporate monorepo=
<pie_>
drakonis1, ah good one
<drakonis1>
it has a lot of forks
<Yaniel>
just cloning should be trivial, no?
<drakonis1>
no
<drakonis1>
a shallow clone is trivial
<drakonis1>
a full clone is not
<pie_>
whats pantheon
<drakonis1>
linux has 826440 commits
<drakonis1>
this is a lot.
<symphorien>
pie_: the desktop environment of elementary os iirc
<pie_>
oh
<drakonis1>
pantheon is a desktop environment, it looks a lot like a mac
<drakonis1>
even more than gnome
<drakonis1>
gnome vaguely reminds me of mac's interface but pantheon looks like a near carbon copy
o1lo01ol1o has joined #nixos
<gchristensen>
aminechikhaoui: have you ever hit problems with your ssh agent causing auth problems with nixops?
<aminechikhaoui>
gchristensen don't think so, what kind of problems
<gchristensen>
my agent provides many SSH keys, and then nixops provides another key
<gchristensen>
but by the time SSH tries the nixops key, the other end has given up and refuses me
drakonis_ has joined #nixos
aleph- has joined #nixos
<aminechikhaoui>
oh, never seen that but I also don't have that many keys
o1lo01ol1o has quit [Ping timeout: 246 seconds]
drakonis2 has joined #nixos
<aminechikhaoui>
also I think I don't have ForwardAgent by default anymore
<gchristensen>
ForwardAgent isn't related, it is just my local agent and my local nixops trying to auth
drakonis has quit [Ping timeout: 257 seconds]
<gchristensen>
I wonder if there is a way to specify priority on keys
drakonis2 has quit [Client Quit]
aleph- has quit [Client Quit]
drakonis1 is now known as drakonis
aleph- has joined #nixos
drakonis_ has quit [Ping timeout: 252 seconds]
linarcx has quit [Ping timeout: 250 seconds]
<aminechikhaoui>
gchristensen wonder if IdentitiesOnly would work
<gchristensen>
oh seems promising
<aminechikhaoui>
as nixops does basically ssh -i <key>
o1lo01ol1o has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
Anton-Latukha has quit [Ping timeout: 246 seconds]
linarcx has joined #nixos
jomik has joined #nixos
jomik has joined #nixos
jabranham has joined #nixos
jabranham has quit [Client Quit]
endformationage has quit [Quit: WeeChat 2.4]
jabranham has joined #nixos
jabranham has quit [Client Quit]
stepcut has quit [Ping timeout: 246 seconds]
jabranham has joined #nixos
jtojnar has joined #nixos
jabranham has joined #nixos
jabranham has quit [Changing host]
mexisme has joined #nixos
Xyliton has joined #nixos
<Xyliton>
Tried to run a game using wine today and I ran into some "couldn't initialize opengl" problems. Am I using the wrong drivers or what am I missing?
<jabranham>
One solution would be to use unstable's gnupg in my system config, but that would require building basically everything myself, which isn't great...
<val_>
these pure storages are a bit confusing. i have xmonad and try to add scaling to xserver. xinitrc from nix/store refers to the Xresources file. xrdb is installed. is editing configuration in the xresources file the best way to configure hidpi?
<infinisil>
aanderse: Btw, regarding the perl thing at your $work, it might be a good idea to fork nixpkgs and add the commits your company needs ontop. Then you won't be constrained by nixpkgs slowness
orivej has quit [Ping timeout: 250 seconds]
lord| has quit [Read error: Connection reset by peer]
lord| has joined #nixos
Ralith_ has quit [Remote host closed the connection]
val__ has joined #nixos
<val__>
these pure storages are a bit confusing. i have xmonad and try to add scaling to xserver. xinitrc from nix/store refers to the Xresources file. xrdb is installed. is editing configuration in the xresources file the best way to configure hidpi?
Ralith_ has joined #nixos
<val__>
Sorry I just hit enter and the message was saved
<val__>
anyway, would be awesome to know
val_ has quit [Ping timeout: 256 seconds]
<val__>
modifying .Xresource's Xft.dpi to different values doesn't yield changes. I merged it and also restarted display-manager service
<infinisil>
Well there is the services.xserver.dpi option, not sure whether this will work
<val__>
which file are you talking about?
b1000101 has quit [Ping timeout: 258 seconds]
<infinisil>
configuration.nix
<val__>
oh
<val__>
good to know
<val__>
thx, gonna try it
<val__>
how can i see all the options i could add or modify in configuration.nix?
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
mwilsoncoding has joined #nixos
<wolfman154>
programs.slock.enable = true; Is throwing an error: option ‘home-manager.users.wolf.programs.slock’ defined in ‘/etc/nixos/configuration.nix’ does not exist?
<wolfman154>
Why is this?
o1lo01ol1o has quit [Ping timeout: 246 seconds]
<pie_>
> { a = 1; c = 2; } // { a = 3; b = -1; }
<{^_^}>
{ a = 3; b = <CODE>; c = 2; }
<pie_>
-_-
<pie_>
> { aa = 1; cc = 2; } // { aa = 3; bb = -1; }
<infinisil>
wolfman154: Don't ask questions in multiple channels at the same time
<{^_^}>
{ aa = 3; bb = <CODE>; cc = 2; }
<pie_>
well nevermind about giving someone an example in here heh.
<infinisil>
pie_: :p
<infinisil>
pie_: Like, add :p to make it strict
<pie_>
oh lmao
<pie_>
i thought you were sticking out your tongue :p
lovesegfault has joined #nixos
orivej has joined #nixos
ggpeti has joined #nixos
<{^_^}>
[nixpkgs] @aanderse opened pull request #60033 → automysqlinit: init at 3.0_rc6 (19.03 backport) → https://git.io/fj3Cv
jabranham has quit [Quit: ERC (IRC client for Emacs 27.0.50)]
o1lo01ol1o has quit [Remote host closed the connection]
<infinisil>
anaothmane: Well nix-shell is a Nix concept, usable wherever you have Nix available. It's used to get ad-hoc environment to develop packages. NixOS is the distribution for configuring things via configuration.nix, it's based on Nix, and it ships with Nix installed
<infinisil>
There won't be anything online to explain this difference, but there are manuals for NixOS and Nix
<anaothmane>
Ok thanks!
o1lo01ol1o has joined #nixos
<anaothmane>
I'm currently trying to make MineTime work (minetime.ai). I patched the interpreter but it still tells me that libsecret-1.so.0 is not found (I installed it with nix-env -i libsecret and it exists in my ~/.nix-profile). What am I missing?
<symphorien>
,libraries anaothmane
<anaothmane>
I patched it like this: patchelf --set-interpreter "$(cat $NIX_CC/nix-support/dynamic-linker)" MineTime-1.5.1-x86_64.AppImage
<{^_^}>
anaothmane: Don't install libraries through nix-env or systemPackages, use nix-shell instead. See https://nixos.wiki/wiki/FAQ/Libraries for details.
<infinisil>
Would anybody mind if we used markdown everywhere instead of xml?
<infinisil>
xml docs are painful to view and edit :/
johanot has quit [Ping timeout: 244 seconds]
<simpson>
infinisil: Yikes, that's unfortunate. Which text editor are you using? It might have secret hidden XML support not yet configured.
<infinisil>
simpson: I am using xml support!
<infinisil>
Or do you have an editor that can remove all xml tags somehow?
dbmikus has joined #nixos
dbmikus__ has quit [Ping timeout: 246 seconds]
kleisli has joined #nixos
<simpson>
infinisil: I am using an editor that has special SGML/XML/HTML tag support, yes, but it doesn't do anything fancy with the visuals other than syntax highlighting. This is roughly the same amount of pain with Markdown vs XML.
<infinisil>
The page is littered with 80% xml tags
jluttine has quit [Quit: WeeChat 2.3]
<infinisil>
It's hard to get an overview
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<simpson>
Read the rendered output, then.
<simpson>
Markdown's not even *defined* except in terms of HTML, which would be a major lossage compared to what we've currently got.
<infinisil>
Ehh
<infinisil>
Who cares about that tbh, I just want good docs for Nixpkgs, and everything being in xml is a major hinderance in writing/updating them for a lot of people
<infinisil>
samueldr: It's an upgrade in pretty much every regard. The only thing we lose is the overly specific tags, which we use for.., for what?
<pie_>
its good that i read that now too because things seem to be a bit different than i thought
Neo-- has quit [Ping timeout: 246 seconds]
zupo has joined #nixos
<sphalerite>
infinisil: please stop complaining, we've had this argument more than often enough. Go ahead, write some docs! You know by now that there's a number of volunteers willing to translate it into docbook. At this point you're just trolling and I think it sucks.
<sphalerite>
(write some docs in markdown*)
<infinisil>
sphalerite: I'm seriously not trolling. I've started my text editor and am deciding upon whether to use rst or markdown to rewrite some docs.
<infinisil>
We've had the problem of bad and outdated docs for a while
<sphalerite>
infinisil: no, don't. That's unproductive.
<infinisil>
There's so many people doing PR's to change nixpkgs, but a super low fraction of those are for docs
<infinisil>
Meanwhile I don't know any other project that struggles so much with docs
<sphalerite>
it isn't actually useful just because it's easier than actually improving the docs (just run the stuff through pandoc) and feels like you're doing things because it generates a big diff
<infinisil>
Rewriting docs in a better format will make improving them much easier in the long term
<pie_>
I wonder how rust handles docs. People are usually pretty happy with their infra AFAIK
<infinisil>
Sure we can go another decade with xml docs and keep them as stagnant as now
<infinisil>
Maybe using a different format won't improve things that much to be on par with other projects, but it will help for sure
<pie_>
...looks like rustdoc uses markdown. is there a reasonable escape hatch for things that markdown doesnt support but you might want to do anyway
sigmundv has quit [Ping timeout: 246 seconds]
<infinisil>
Of course, this would never get into nixpkgs because we have a bunch of important people advocating xml, but I still feel like I should give it a try
<pie_>
has anyone talked to anyone that prrofessionally works on technical docs
<infinisil>
pie_: No idea, but thanks for another data point
<pie_>
otoh maybe thats the wrong question, because of course professional technical documentation writers probably use xml
<infinisil>
Rust, a language praised for its great docs (have seen that multiple times on reddit), uses markdown, one of the supposedly worst doc writing formats
<andi->
they also have a fulltime position writing docs.. ("the book" at least)
<infinisil>
Yeah true, they have much more man power in every way. But this just illustrates that we don't need xml/docbook at all
<pie_>
well, maybe we need to think about whats expected
<andi->
I am not set on an opinion here.. Haven't written much docs but docbook never was a bummer for me.. Not sure how others feel about it (that haven't written docs).
<pie_>
more focused documentation development or more drive-by edits
<pie_>
or both
<pie_>
the format doesnt matter as long as the latter gets a frontend that has a low initial cost
<pie_>
the former can just use whatever more involved tooling that they want
<infinisil>
Every time I have to edit our docs I have to look up the tags again
<infinisil>
,xml
<{^_^}>
Nix XML docbook cheatsheet: <link xlink:href="https://example.com"/> <citerefentry><refentrytitle>man</refentrytitle><manvolnum>1</manvolnum></citerefentry> <programlisting>let some = "program"; in some</programlisting> <literal>true</literal> <option>users</option> Simple docbook reference: https://docbook.rocks/
<pie_>
ok so why dont we have an IDE for it yet
<pie_>
and im not trolling, i dont mean why havent you written it yet
<samueldr>
when writing a minor overhaul of a documentation section, and with experience writing markdown-based documentation; with additional "semantic document" bias (e.g. the semantic web) I can say that I much prefer docbook even though it looks sometimes inconvenient for drive-by changes...
<infinisil>
Why I haven't written an IDE for editing markdown yet?
<pie_>
i mean, if this is a problem for drive-by-edits, why havent we made it easier yet
<samueldr>
but, sometimes optimizing for drive-by edits is not the answer :/
<pie_>
my understanding is that a lot of changes to nixpkgs are drive-by contributions
<pie_>
and i could see how that could be true
<samueldr>
my work time will soon be more focused towards nix-dependent things, hopefully, and things I want to look for _for my own benefits_ is better tooling for docbook
<samueldr>
I sure hope it turns out I can do those things
<pie_>
so i guess im kind of with infinisil on this one except I guess my first reaction is why isnt this just solved with tooling
<pie_>
well, 1) because noone has written it yet 2) "how hard could this possibly be"
<pie_>
xD
<samueldr>
a good chunk of (1), easy solutions like markdown have stolen the limelight
<infinisil>
Considering how many tags and stuff docbook has, I'm not surprised to not see much tooling
<infinisil>
Pandoc only implemented a fraction of all of docbooks tags
<samueldr>
seeing how github and github-flavoured markdown probably helped a bunch
<samueldr>
meanwhile, docbook relies more on xml-generic tooling AFAIUI
<pie_>
i mean, xml is for arbitrary semantics right? i dont know anything about this docbook stuff
<sphalerite>
infinisil: in 2019, exactly 2 people have made material contributions to the parts of the manuals that are written in markdown. You are not one of those people ;)
<samueldr>
they have full schemas, meaning that an XML editor will be able to use them
<pie_>
(i kind of hate latex but its also pretty good kind of)
<pie_>
sphalerite, there are parts of the manual written in markdown?
<pie_>
:D :P
<sphalerite>
pie_: yep!
<samueldr>
about (2), pie_ it's most likely "not that hard" to have a 90% docbook useful, 100% for nixpkgs, but lacking the obscure tags
<sphalerite>
doc/languages-frameworks/*.md
<samueldr>
I made some exploratory work earlier this year
<pie_>
samueldr, exactly thats what i meant, a sufficient subset to handle most of nixpkgs might be good enough
<infinisil>
sphalerite: I'm not sure what you're trying to hint at with this. I don't write much docs? I don't contribute enough? Similar I could make a similarly irrelevant point: Maybe the reason I haven't contributed docs in 2019 is because all parts I wanted to were written in xml which discouraged me to do it.
<samueldr>
I believe it should be possible to have a WYSIWYM editor for the manual in a couple week's worth of time
<pie_>
samueldr, on the other hand, im not sure how that could be integrated into github stuff
Jackneillll has quit [Read error: Connection reset by peer]
<pie_>
samueldr, good luck \o/
<infinisil>
I did contribute one piece of docs last year, and that was the idris docs, written in markdown, because they were seriously outdated
<sphalerite>
infinisil: they were seriously outdated? :D
<sphalerite>
even though they're written in markdown which is the solution to docs being outdated :D
<infinisil>
Seriously outdated, for serious this time!
<infinisil>
They might still be outdated if they were in xml..
<sphalerite>
I mean, if you're rewriting docs then go ahead and do it in markdown, I'm sure that would be welcomed
<infinisil>
I'm just saying, here's my point of view: I feel *much* more encouraged to contribute docs in markdown than xml. I'm used to markdown, I use it on GitHub and my projects all the time
<infinisil>
xml is just an hour of "how does this work again" "oh i forgot a closing tag here" "Oh the compilation fails, let's see what I messed up"
<sphalerite>
I just don't think destroying all the semantic information in the existing docs is prodcutive.
<pie_>
samueldr, a whole infrastructure around dealing with translations to and from docbook/markdown sounds kind of sketchy, how could that be handled well? docbook as more expressive base representation with an option to expose markdown to the user via tooling layer?
<infinisil>
sphalerite: I asked this before: What do we use these semantics for?
<pie_>
what are the things we could do with extra semantic information, and why dont we do it yet?
<samueldr>
it can be as dumb as a DOM tree, or as fancy as WYSIWYG with tag-like objects
<pie_>
infinisil, there it is then? :p The Things We Should Be Doing
<samueldr>
there are cross references already, other things are due to lack of work hours, things like testing the docs automatically so that when examples bitrot, the docs fail to build
* pie_
is great at wanting more COOL features and making everything complicated ;~;
<pie_>
samueldr, sounds like that could be nice
<infinisil>
I wouldn't dislike xml as much if we had cool features available due to it
vid has quit [Ping timeout: 245 seconds]
<infinisil>
Not possible with other formats
<samueldr>
when I'll be $working on nix related things I'll (1) have more free time since my free time currently has been/is being consumed with preparatory work for that exciting future
<pie_>
i think gchristensen or someone else mentioned it in a talk and i feel like we could benefit from better community work distribution :/ but maybe thats just my desire to overorganize everythin
<pie_>
like, is there some documentation workgroup i could go to if i wanted to help with improving docs? not really afaict
<infinisil>
pie_: I'd love to be part of one
<samueldr>
I feel nixpkgs is EDD, egoistically driven development (said in a loving caring way)
<pie_>
ive been meaning to look into how rust does community organization because i hear they do it well?
<pie_>
samueldr, <3 :3
<mdash>
pie_: too soon to tell
<pie_>
need to make an orga-orga group and read a couple organizational psychology textbooks :P ENOTIME ... anyway </tangent>
<infinisil>
Ohh also something else: We don't even have proper syntax highlighting in our docs
<pie_>
so....we have established that infinisil would like nicer ways to work with docs, samueldr has done some experiments
<samueldr>
that's a fun thing: the syntax highlighting is added post-compile by a js library
tboston_ has quit [Ping timeout: 245 seconds]
<samueldr>
and I think it's only done on the website :/
<infinisil>
:/
* pie_
ponders opening 5 more workgroup irc channels
<samueldr>
yeah, formats with github integration will always have an almost-unfair leg up :(
<pie_>
infinisil, have you seen agda generated html
<infinisil>
pie_: Nope
<jtojnar2>
How do I run fsck on NixOS manually? I have tried passing fsck.mode=force to systemd-boot line but it asks for manual run
<jtojnar2>
(getting error in stage 1)
<pie_>
samueldr, i guess we should se what we can d oabout that thne
o1lo01ol1o has joined #nixos
<samueldr>
jtojnar2: maybe shell on fail and do it imperatively in stage-1 ?
<infinisil>
I guess if I want to move nixpkgs away from xml, I should collect evidence of xml being worse for contributing docs than other formats
<pie_>
my main reason for not contributing docs is i would have to line break stuff by hand when writing the raw xml with github, if anyone cared about that
<jtojnar2>
samueldr: the prompt only offers r to reboot or * to ignore
<pie_>
well and also not knowing enough to write docs, but.... :p
<samueldr>
jtojnar2: do you know whether it's systemd just as when stage-1 hands over to stage-2? if so not sure exactly how there
<samueldr>
jtojnar2: but I know my first instinct would be shell on fail and debug1devices (iirc)
<infinisil>
pie_: Phew, editing on github would be a pain
<samueldr>
jtojnar2: able to extrapolate from there or should I dig the right parameters to pass to the kernel?
mexisme has joined #nixos
<samueldr>
infinisil: additionally what other formats do better, and that can include "getting more drive-by contributions" and "look at all those toolings... can't hold them all!"
<pie_>
infinisil, and the reason i even wanted to do that was because i didnt know how to use git
<infinisil>
samueldr: Yeah
<pie_>
infinisil, imagine all the noobs out there that want to help but dont know git
__monty__ has joined #nixos
<pie_>
samueldr, ^ :p
<infinisil>
pie_: Hmm yeah, that's another point against xml
o1lo01ol1o has quit [Ping timeout: 246 seconds]
<pie_>
.-. it is?
<jtojnar2>
samueldr: I do not really know how to shell on fail
<pie_>
well ok yeah i guess cuz you could just edit markdown on github
<sphalerite>
jtojnar2: add boot.shell_on_fail to the kernel ocmmand line
<pie_>
i would also prefer not to lose all the semantic info though
<infinisil>
pie_: Yeah because if new people want to contribute without knowing git, they could totally edit markdown (or similar) docs on github
<samueldr>
boot.shell_on_fail boot.debug1devices # both of these, the second one will stop before mounting things in stage-1
<stepcut_>
how do I set restrict-eval / allowed-uris in a release.nix ? I tried , pkgs = import <nixpkgs> { restrict-eval = false; ... }, and that allows the checkout to happen, but then the hydra jobs fail with ==> in job ‘lens-ghc861’: anonymous function at /nix/store/db8qirfxqxavwk1c0nc2mjyg6hllrcz7-source/pkgs/top-level/default.nix:20:1 called with unexpected argument 'restrict-eval', at /nix/store/db8qirfxqxavwk1c0nc2mjyg6
<jtojnar2>
samueldr: I probably still want to mount since I use LUKS
<samueldr>
IIRC this is handled at device
* samueldr
checks
drakonis has joined #nixos
<samueldr>
I'm still pretty sure that yes, luks (and lvm) happens before mount commands
tomberek has joined #nixos
<jtojnar2>
Holy smokes, the number of corrupted directories
<samueldr>
:/
<pie_>
oh deer
<pie_>
backups?
* pie_
gets jittery again because he still hasnt set up backups...
<infinisil>
Another point: I've been reviewing NixOS contributions for a while now, and I've probably never seen a new contributor get xml tags right on the first try. I have to correct them all the time. And then later I learn about new tags myself and have to correct everyone for another tag in the future
<pie_>
infinisil, are you writing all this down
<infinisil>
pie_: Um, it's in my IRC logs, soo yes? xD
<pie_>
infinisil, sounds like that could also be handled by tooling but still
<pie_>
infinisil, that doesnt count :P "its in there" but do you actually want to go back and search it up later to substantiate some argument?
<pie_>
its easier to start pasting snippets in a textfile now :p
<infinisil>
pie_: Can't write tooling to detect that "See man(1) for more info" should be "See <citerefentry><refentrytitle>man</refentrytitle><manvolnum>1</manvolnum></citerefentry> for more info" instead
<pie_>
this is what i do and then i forget everything <pie_> infinisil, that doesnt count :P "its in there" but do you actually want to go back and search it up later to substantiate some argument?
<pie_>
infinisil, thats not correcting bad tags though?
<jtojnar2>
At least the work and configs are on GitHub, but minor things like installed user styles are broken and not really backed up
<infinisil>
pie_: Yeah, I meant correcting also as in making sure people actually use the tags
<pie_>
ah
<infinisil>
Very important for https links
<pie_>
infinisil, could check if "see ..." has any tags after it and warn if not xD
<infinisil>
Ehh...
<pie_>
(yeah thats totally not going to result in semantic whack-a-mole)
<pie_>
infinisil, in the other hand markdown doesnt have that at all?
<pie_>
s/in/on
<infinisil>
Bot: "I noticed your docs contain only 2.5% tags, whereas the average docs contain 3.74%, are you sure you tagged everything"
<pie_>
haha
<infinisil>
s/3.74/20.74
<infinisil>
pie_: Yeah markdown doesn't have anything specific for man pages, you'd just use `man(1)` or so
val_ has joined #nixos
<samueldr>
the thing with markdown (and equivalent systems) is how they're watered-down html with mostly 1:1 mapping (when they're defined well), meaning that yeah, you won't have a tagsoup to cite+ref a manpage, but have to do the hard work of making a link yourself and to _something_
<pie_>
so its kind of not a point against xml, at least with respect to markdown
kleisli has quit [Quit: Leaving]
<infinisil>
samueldr: Yeah I guess
johanot has joined #nixos
<jtojnar2>
samueldr: shell_on_fail was enough
<samueldr>
jtojnar2: good to know
<jtojnar2>
samueldr++
<{^_^}>
samueldr's karma got increased to 78
linarcx has quit [Quit: WeeChat 2.4]
<jtojnar2>
And fsck even fixed the user styles (probably som sqlite database in Firefox)
<jtojnar2>
I should probably have some more robust backup scheme in place
<jtojnar2>
What do you guys use?
<pie_>
restic generally seems good
<samueldr>
I use borg backups, but still haven't made them automated :x
lovesegfault has quit [Ping timeout: 244 seconds]
<pie_>
also dont do the stupid thing that i did where my only copy of my backup decryption key was in the encrypted backup
<samueldr>
borg has amazing deduplication that works across machines, if you use the same.... whatever they name it
<samueldr>
(I have a bunch of duplication across my in-use machines)
<jtojnar2>
So no NASes or cloud?
rauno has quit [Ping timeout: 264 seconds]
<samueldr>
it goes to a remote server
<Ralith>
borg is nice for manual backups against your own infra
<pie_>
restic has encryption by default which i think is a plus
<Ralith>
I wish it could be pointed at backblaze though
<Ralith>
and I wish it had a write-only cryptography setup so you could automate it
<jtojnar2>
Hmm, and now the second drive is throwing smart errors, what a luck
<pie_>
yeah restic doesnt have the write only thing either, not cryptographically anyway
<jtojnar2>
Thanks for the help, got to run now
<pie_>
i havent used borg but https://stickleback.dk/borg-or-restic/ seems relatively thorough, not sure how outdated it is with respect to restic, for example i havent seen the lack of compression ever mentioned and incremental backups never seemed slow to me *shrug*
<pie_>
good luck
jtojnar2 has quit [Quit: AtomicIRC: The nuclear option.]
<{^_^}>
[nixpkgs] @thoughtpolice pushed 6 commits to nixpkgs/eda-updates: https://git.io/fj3l4
<{^_^}>
[nixpkgs] @thoughtpolice pushed 0 commits to nixpkgs/eda-updates: https://git.io/fj3lB
<{^_^}>
[nixpkgs] @thoughtpolice pushed 6 commits to master: https://git.io/fj3l0
<stepcut_>
is the format of /etc/nix/machines documented anywhere? I am not finding it in the manuals
<{^_^}>
[nixpkgs] @dtzWill opened pull request #60043 → <!-- Nixpkgs has a lot of new incoming Pull Requests, but not enough people to review this constant stream. Even if you … → https://git.io/fj3lN
val_ has quit [Ping timeout: 256 seconds]
<wolfman154>
14:01 wolfman154: I tried setting up slock in home.nix but it’s not working. Could someone paste bin their screen lock code in home.nix? Thanks!!!
<{^_^}>
[nixpkgs] @dgarzon opened pull request #60045 → pipenv: missing virualenv in propagatedBuildInputs. → https://git.io/fj3lp
<jcrben>
pie_: you've got the wrong person; I've never given a nixcon talk
<pie_>
oops
<pie_>
i mean to highlight nbp
<pie_>
nbp: <pie_> I'm watching your nixcon talk about overlays but I don't really understand why library functions should come from super and not self
<pie_>
<pie_> I'm not sure if you actually explained that, I don't think you did, you only said that super should be used.
<pie_>
<pie_> btw this talk has helped me a lot more than anything else ive found so far
yuken has joined #nixos
<infinisil>
pie_: I think the reason might be something like, functions from lib could be used do define the overlays themselves
<infinisil>
And if those depend on the overlays (which they do if you were to use self), it would give infinite recursion
<wolfman154>
Never mind I give up, maybe I’ll try later
<Dharco>
Hi, so I'm against a wall, and my searches aren't finding much to help... I've managed to compile the custom application as nixpgks package, and setup the systemd using configuration.nix, etc... however the custom application depends on another executable (zfs in particular), and that executable although in my shell, is not in the custom applications shell/path... could anyone point me to relevant documentation to help me specify runti
<cransom>
am i missing something vital with the 'eval' input type with hydra? to get evaluation time down and eliminate some duplicate builds, i switched one of my git inputs over to a previous hydra evaluation and now the noop evaluation take just as long as the ones where i would have had git changes upstream.
orivej has joined #nixos
<pie_>
emilsp, you can also try the bot / nix-locate to find the binary
<pie_>
,locate dig
o1lo01ol1o has quit [Remote host closed the connection]
<{^_^}>
Found in packages: bind.dnsutils
<emilsp>
I'll try and keep that in mind. I'm sure this isn't the first time I've received such advice
<pie_>
Dharco, can you post your configuration?
drakonis has quit [Read error: Connection reset by peer]
o1lo01ol1o has joined #nixos
<pie_>
emilsp, maybe start taking notes! :P i started too late
<Dharco>
which part? I modified/created numerous files
<pie_>
now im trying to compile a FAQ for things i wish i knew in the beginning but its hard because i dont remember what i didnt know
<buckley310>
can someone else confirm an issue just so i know its not just my machine? this stack-traces for me: nix-shell -p python2.pkgs.pwntools --command 'python2 -c "import pwn"'
johanot has quit [Quit: WeeChat 2.4]
o1lo01ol1o has quit [Ping timeout: 258 seconds]
Ariakenom has quit [Quit: Leaving]
noonien has joined #nixos
<gchristensen>
I have machineA and machineB, and I'm nix-copy-closure'ing from machine A to machine B. machine B has a special cache which machine A does not have. I'm copying something whose closure is already in the special cache. will machineB substitute it all, or will machineA copy files over since machineA doesn't have the special cache?
<clever>
gchristensen: that also means that a copy will initially hang and not say what its copying, because the remote machine is checking the binary cache, to see what has to copy
<ivan>
symphorien: that's just unlocking nightly features in the stable release though, right
<symphorien>
Yes
<{^_^}>
[nixpkgs] @dtzWill opened pull request #60058 → iwd: split ell to separate derivation, update iwd and ell → https://git.io/fj34O
<ivan>
heh I just tried RUSTC_BOOTSTRAP=1 with #![feature(proc_macro_hygiene)] and it worked, so maybe it's good enough, thanks
hoek has quit [Ping timeout: 264 seconds]
sdier has quit [Ping timeout: 264 seconds]
<symphorien>
For now the only thing I renounced to build with this trick is clippy
coot has quit [Max SendQ exceeded]
ajmcmiddlin has quit [Ping timeout: 252 seconds]
teehemkay has joined #nixos
kalbasit has quit [Read error: Connection reset by peer]
silver has quit [Read error: Connection reset by peer]
georgew has joined #nixos
bpierre has quit [Ping timeout: 252 seconds]
jw358 has joined #nixos
ottidmes has joined #nixos
tomberek has quit [Ping timeout: 246 seconds]
oharvey has joined #nixos
rajivr___ has quit [Ping timeout: 258 seconds]
kleisli has joined #nixos
stepcut has joined #nixos
techtangents has joined #nixos
aanderse_ has joined #nixos
tazjin has joined #nixos
drakonis1 has joined #nixos
bpierre has joined #nixos
ajmcmiddlin has joined #nixos
coot has joined #nixos
NemesisD has joined #nixos
aanderse has quit [Ping timeout: 264 seconds]
gonz_ has joined #nixos
sdier has joined #nixos
kalbasit has joined #nixos
dongcarl has joined #nixos
rajivr___ has joined #nixos
hoek has joined #nixos
noonien has joined #nixos
knupfer has quit [Remote host closed the connection]
__monty__ has quit [Quit: leaving]
<ottidmes>
I have packaged the frontend of bitwarden that contains some static web assets (HTML/CSS/JSS/images) and is used as the frontend of the bitwarden_rs package. I have made that static asset package configurable through the NixOS module for bitwarden_rs. Now I wonder about the static asset package, should I put the files in the root its store directory (i.e. $out) or in $out/share/bitwarden_rs/vault?
<ottidmes>
guess I will just go with the /share variant, because other distros follow that pattern and that would also allow installing the frontend via systemPackages, in the module option I will just mention it should be placed under that directory structure
lovesegfault has quit [Quit: WeeChat 2.4]
lovesegfault has joined #nixos
work_ has joined #nixos
aanderse_ has quit []
aanderse has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
Shouou has quit [Remote host closed the connection]
kini has joined #nixos
Shouou has joined #nixos
<Dharco>
how does nix handle package dependencies? like if executable a calls executable b, how would a specify that needs access to the path to b?
aw_ has quit [Ping timeout: 276 seconds]
aanderse has quit [Remote host closed the connection]
teehemkay has quit [Ping timeout: 276 seconds]
dongcarl has quit [Ping timeout: 276 seconds]
hoek has quit [Ping timeout: 276 seconds]
kalbasit has quit [Ping timeout: 276 seconds]
acarrico has quit [Ping timeout: 276 seconds]
retrry[m] has quit [Ping timeout: 276 seconds]
ottidmes has quit [Ping timeout: 276 seconds]
aw has joined #nixos
aw has joined #nixos
aw has quit [Changing host]
teehemkay_ is now known as teehemkay
aanderse has joined #nixos
ottidmes_ has joined #nixos
NemesisD has quit [Ping timeout: 276 seconds]
v0d1ch has quit [Ping timeout: 276 seconds]
lovesegfault has quit [Ping timeout: 276 seconds]
lovesegfault has joined #nixos
<infinisil>
Dharco: You'd specify that in the build by either patching the source code to refer to the binary directly (e.g. replacing /usr/bin/b with ${pkgs.b}/bin/b), or if the package checks for the executable in PATH, you can wrap a with a PATH that contains b
kalbasit has joined #nixos
<{^_^}>
[nixpkgs] @Infinisil pushed commit from @JohnAZoidberg to release-19.03 « libpst: Fix build »: https://git.io/fj3BE
dongcarl has joined #nixos
hoek has joined #nixos
NemesisD has joined #nixos
v0d1ch has joined #nixos
<Dharco>
I think my problem is that the systemd unit file isn't getting its PATH set to the location of the 'zfs' executable
glasserc_ has joined #nixos
<Dharco>
and I cannot for the life of me figure out how to inject that path into the systemd unit file that is being generated
<infinisil>
Dharco: If the package itself depends on zfs though, you shouldn't use this NixOS option, but change the build of the package to refer to the binary directly
octe_ has joined #nixos
slyfox_ has joined #nixos
nahamu_ has joined #nixos
jordansinn_ has joined #nixos
<infinisil>
..Although, upon second thought, zfs depends on the kernel, which might be different for different systems..
gyroninja_ has joined #nixos
andromed1-galaxy has joined #nixos
mconstant has quit [Ping timeout: 245 seconds]
orbekk1 has joined #nixos
vid has joined #nixos
<Dharco>
I'm not the owner/maintainer of the application code, I'm just trying to package it up so that I can use it in a NixOS way..... I can build the application and it runs fine in a shell... but I want to run it as a service from my configuration.nix file
IooNag_ has joined #nixos
moredhel- has joined #nixos
<pie_>
infinisil, i *was* wondering what to do in the case of kernel stuff...i mean theres two things here i guess? one is the kernel side stuff and the other is the interface to it?
<infinisil>
Oh I think the user-land zfs package doesn't depend on the kernel, so embedding it in the build should be fine
glasserc has quit [Ping timeout: 250 seconds]
zfnmxt has quit [Ping timeout: 250 seconds]
cyphase has quit [Ping timeout: 250 seconds]
lstanley has quit [Ping timeout: 250 seconds]
alexherbo2 has quit [Ping timeout: 250 seconds]
moredhel_ has quit [Ping timeout: 250 seconds]
dminuoso has quit [Ping timeout: 250 seconds]
delan has quit [Ping timeout: 250 seconds]
dave_uy has quit [Ping timeout: 250 seconds]
eon` has quit [Ping timeout: 250 seconds]
jb55 has quit [Ping timeout: 250 seconds]
slyfox has quit [Ping timeout: 250 seconds]
irdr has quit [Ping timeout: 250 seconds]
glenndavy has quit [Ping timeout: 250 seconds]
nahamu has quit [Ping timeout: 250 seconds]
glines has quit [Ping timeout: 250 seconds]
mg has quit [Ping timeout: 250 seconds]
mdash has quit [Ping timeout: 250 seconds]
eisbaer-north has quit [Ping timeout: 250 seconds]
migy_ has quit [Ping timeout: 250 seconds]
jordansinn has quit [Ping timeout: 250 seconds]
shazow has quit [Ping timeout: 250 seconds]
ar1a has quit [Ping timeout: 250 seconds]
Jackneill has quit [Ping timeout: 250 seconds]
orbekk has quit [Ping timeout: 250 seconds]
nurelin_ has quit [Ping timeout: 250 seconds]
afics has quit [Ping timeout: 250 seconds]
iMatejC has quit [Ping timeout: 250 seconds]
sveitser has quit [Ping timeout: 250 seconds]
oleks has quit [Ping timeout: 250 seconds]
veverak has quit [Ping timeout: 250 seconds]
szicari has quit [Ping timeout: 250 seconds]
le_jonge has quit [Ping timeout: 250 seconds]
ent- has quit [Ping timeout: 250 seconds]
kriztw has quit [Ping timeout: 250 seconds]
gyroninja has quit [Ping timeout: 250 seconds]
maurer has quit [Ping timeout: 250 seconds]
bsima has quit [Ping timeout: 250 seconds]
octe has quit [Ping timeout: 250 seconds]
andromeda-galaxy has quit [Ping timeout: 250 seconds]
Lears has quit [Ping timeout: 250 seconds]
andi- has quit [Ping timeout: 250 seconds]
marcusr has quit [Ping timeout: 250 seconds]
IooNag has quit [Ping timeout: 250 seconds]
WizzieB has quit [Ping timeout: 250 seconds]
cransom has quit [Ping timeout: 250 seconds]
<infinisil>
pie_: Yeah
zfnmxt has joined #nixos
<Dharco>
so how would 'embed' it into the build?
<Dharco>
+I
cyphase has joined #nixos
phreedom_ has quit [Remote host closed the connection]
dave_uy has joined #nixos
WizBright has joined #nixos
phreedom has joined #nixos
<infinisil>
Dharco: How does it refer to the binary in the source code? If it gets it from PATH, you can use `wrapProgram $out/bin/a --suffix PATH : ${zfs}/bin/zfs`
tdbgamer has joined #nixos
maurer has joined #nixos
cransom has joined #nixos
Jackneill has joined #nixos
afics has joined #nixos
<infinisil>
(after adding makeWrapper to nativeBuildInputs)
iMatejC has joined #nixos
szicari has joined #nixos
marcusr has joined #nixos
bsima has joined #nixos
veverak has joined #nixos
marcusr has joined #nixos
marcusr has quit [Changing host]
dminuoso has joined #nixos
stepcut_ has quit [Ping timeout: 276 seconds]
<Dharco>
I assume it gets it from PATH... I'm not even a golang novice, so I'm not going to try to dig into that code to find out
andi- has joined #nixos
<Dharco>
looking into 'wrapProgram'
ar1a has joined #nixos
nurelin_ has joined #nixos
stepcut has joined #nixos
<pie_>
Dharco, welcome to the rabbit hole, 'xD good luck!
<scott>
does nixpkgs have something akin to ubuntu's build-essential package? specifically something i could `nix use` that would automatically give me commonly expected build tools like gcc, make, etc, without needing to name multiple packages in most cases
<infinisil>
scott: A nix-shell will give you that :)
tobiasBora has quit [Ping timeout: 246 seconds]
<infinisil>
nix-shell -p <list of packages, or empty>
<scott>
ahh, makes sense, thanks
<infinisil>
These standard tools come from stdenv
zupo_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<infinisil>
So if you want to have more flexibility than -p, you'll have to write a shell.nix file with contents like `stdenv.mkDerivation { name = "..."; ...; }`
<scott>
conveniently i can just look at $PATH inside a nix-shell with empty package list
andrewrk has joined #nixos
dbmikus has quit [Quit: WeeChat 2.4]
azdle has quit [Ping timeout: 246 seconds]
<infinisil>
Yeah (but don't hardcode those paths anywhere)
dbmikus has joined #nixos
azdle has joined #nixos
tobiasBora has joined #nixos
<pie_>
i wonder why i feel like we're getting a sudden influx of new people :D
<MichaelRaskin>
Aren't we always?
<Dharco>
pie_ & infinisil thank you very much, I was finally able to get it working end-to-end with your help
<pie_>
gonna have to start writing some faqs real soon (or something)
<scott>
i'm not new ftr, i just hadn't realized the particular trick of an empty nix-shell package list :p
<pie_>
MichaelRaskin, youre probably right :3
<infinisil>
Dharco: Nice :)
<pie_>
or im just on irc too much recently
<infinisil>
pie_: I should record stats in my bot on what the new-people rate is
<pie_>
scott, ok maybe nix just has a lot of dark corners :p
<pie_>
a very large amount of dark corners
<pie_>
with different corners being dark for everyone
<infinisil>
Hehe
<Dharco>
I would not have thought to overwrite/wrap the executable with something to inject the path... even with the hints that was not trivial to understand/do
<scott>
learning more about exactly how stdenv is structured has been on my todolist for a couple years now...
<MichaelRaskin>
Don't worry, if gets restructured once in a couple of years anyway
<scott>
heh
<pie_>
i feel like its kind of hard to half-ass something with nix...depending on what youre half-assing
<pie_>
other times its very easy
<pie_>
x'D
galaxie has left #nixos [#nixos]
* pie_
randomly starts pondering if nix will ever become a thing you dont need to spend months of your life on to be able to work with
<pie_>
i.e. be able to learn nix and still get paid if your job is not devops :p
<scott>
i've described nix to someone as "the first package manager". like, it's the first one that has properties that *really* interest me, but i (optimistically) consider it the earliest stage in its class, so there's lots of room to improve, especially in ergonomics
<pie_>
i dont want to imagine computing without nix
<scott>
nix 2.0 definitely made a lot of stuff feel fancier
<pie_>
i think i have serious stockholm syndrome :D
<MichaelRaskin>
I think the Nix manual can still be read in a few days. And it is kind of enough to use Nix.
<MichaelRaskin>
If you had Stockholm syndrome, you would be praising apt!
<pie_>
i dont thing you can read the nix manual like you read a book
<MichaelRaskin>
At the book-rate I would give it half a day!
<pie_>
"The existence of an overrideScope attribute that creates a complete, modified copy of the package set for the context of the given attribute may have been a mistake in the first place"
<pie_>
though maybe im not immediately understanding what thats talking about
<scott>
there's being able to use nix at all, and there's being able to use nix fluently in a way that works for you on a daily basis
<pie_>
ugh i should stop for today. i feel like i dug into a really complicated corner...
<MichaelRaskin>
And when I first installed NixOS (a bit more than ten years ago) I did indeed just read all the three manuals in a day (they were smaller back then)
<scott>
i feel like i'm still figuring out new ways to use nix that i really like that aren't in any manual
<pie_>
scott, per clever's not...suggestion...but mention of some things, i startedusing nix to run stuff as opposed to just packages and it...works lol
<pie_>
but my stuff im running doesnt need network access or such
<scott>
programming in the nix language? what kind of stuff?
<pie_>
no i mean using the nix sandbox as a runner for ghidra
<scott>
oh ok
<pie_>
im trying to figure out how to make a sane .withPackages for it ;~;
<{^_^}>
#59344 (by deliciouslytyped, 1 week ago, open): Proposal: Engineering a Common "Plugin" Infrastructure
<NemesisD>
do y'all have any suggestions when you need command to generate several nix files? for instance, my package depends on node2nix, which generates 3 .nix files, which my project also needs
<NemesisD>
do i build a derivation with buildInputs of `node2nix` and my config and dumps those .nix files to the `$output`?
<{^_^}>
[nixpkgs] @worldofpeace pushed to master « gnome3.gnome-control-center: hardcode path to gnome-desktop »: https://git.io/fj3R0
statusbot has quit [Ping timeout: 246 seconds]
Lears has joined #nixos
<pie_>
I think that might fall under recursive nix, which is for now generally discouraged I think - I mean you should probably have all the .nix you need availible at the time you run nix-build or whatever
statusbot has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
<pie_>
i mean, if after that you want to call nix on the generated files
<pie_>
but im not sure, maybe you can make it work
[Leary] has quit [Read error: No route to host]
<pie_>
actually....i guess you could just have that derivation be an input to something else and just import its outputs
<NemesisD>
pie_: that's what i had in mind
<pie_>
im suddenly unsure if you can intersperse builds and evaluation
<pie_>
i think you cant
<pie_>
but hey, give it a shot
wolfman154 has joined #nixos
o1lo01ol1o has joined #nixos
<pie_>
it shouldnt be that hard to try a simplified experiment#
<pie_>
just write a derivation that echos a string containing nix code to an output and try to use its result
Dharco has quit [Quit: Page closed]
<NemesisD>
alternatively i could have my script to start a nix shell run a `nix-shell` command first that uses `node2nix` to generate all the nix files since that process is idempotent
<wolfman154>
I’m going to do a fresh install of nixos 19.03, will changing homeState in home.nix to 19.03 cause anything to crash? As long as both configuration.nix and home.nix are set to 19.03, every should work without anything crashing right?
<scott>
that RFC mentions import from derivation in the alternatives section but doesn't go into a lot of detail
o1lo01ol1o has quit [Ping timeout: 246 seconds]
mmlb has joined #nixos
vk3wtf_ has quit [Ping timeout: 244 seconds]
<pie_>
scott, maybe clarification can be asked for
<NemesisD>
also a much more basic question. how do i go from a set to a list of values?
cyphase has quit [Quit: cyphase.com]
<scott>
NemesisD: builtins.attrValues, i think
ayerhart has quit [Read error: Connection reset by peer]
gagbo has joined #nixos
<NemesisD>
scott: thanks!
work_ has quit [Quit: Connection closed for inactivity]
<pie_>
if only we had hoogle! :P
<pie_>
I mean I guess we could manually make something
* pie_
ponders
<pie_>
so yeah what if we just made haskell or something prototypes for a bunch of stuff
<pie_>
there's already some types in comments for functions in nixpkgs, though I dont know if that uses the prototype nix type system or what?
stepcut has joined #nixos
<gchristensen>
it doesn't
<pie_>
infinisil, maybe with good enough documentation type signatures for hoogle could be generated from the documentaiton? :p ^
<pie_>
(and by that i actually mean a bunch of hand written type signatures in the docs or something :p , probably better to just have it in the code instead though)
<infinisil>
pie_: A problem is that we don't have any define nix structure to define stuff
<infinisil>
pie_: lib is just `import <nixpkgs/lib>`, and they're defined in some random attrset structure
<pie_>
eh, its all ad-hoc anyway right? :o
<pie_>
* :p
stepcut_ has quit [Ping timeout: 268 seconds]
<infinisil>
We don't have something like Haskell modules, where all top-level definitions are to be used
<pie_>
could start with at least lib and builtins
sigmundv has joined #nixos
jasongrossman has joined #nixos
<pie_>
I should go to sleep and stop coming up with Ideas.
<pie_>
gnight guys! o/
<wolfman154>
I’m going to do a fresh install of nixos 19.03, will changing homeState in home.nix to 19.03 cause anything to crash? As long as both configuration.nix and home.nix are set to 19.03, every should work without anything crashing right?
<wolfman154>
I’m just double checking?
<gchristensen>
should be good
mbrgm_ has joined #nixos
<wolfman154>
gchristensen: ok, sounds good
<wolfman154>
:)
mbrgm has quit [Ping timeout: 246 seconds]
mbrgm_ is now known as mbrgm
coconnor has joined #nixos
b1000101 has quit [Ping timeout: 252 seconds]
<gchristensen>
infinisil: I did some experiments with an "exports" thing, and public vs. private library functions, and also unit testing.