<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/JecXf
o1lo01ol1o has quit [Ping timeout: 240 seconds]
t58_ has joined #nixos
michaelpj has quit [Read error: Connection reset by peer]
michaelpj has joined #nixos
Gopal[m] has quit [Ping timeout: 252 seconds]
gudea[m] has quit [Ping timeout: 252 seconds]
groggy[m] has quit [Ping timeout: 252 seconds]
Dan[m]8 has quit [Ping timeout: 252 seconds]
jak[m] has quit [Ping timeout: 252 seconds]
kxlljoyy[m] has quit [Ping timeout: 252 seconds]
ndarwincorn has quit [Ping timeout: 252 seconds]
peterhoeg[m] has quit [Ping timeout: 252 seconds]
MiloIgnis[m] has quit [Ping timeout: 252 seconds]
Sisyphe[m] has quit [Ping timeout: 252 seconds]
slabity has quit [Ping timeout: 252 seconds]
Seb[m] has quit [Ping timeout: 252 seconds]
luftmensch[m] has quit [Ping timeout: 252 seconds]
rycee has quit [Ping timeout: 252 seconds]
pinage404[m] has quit [Ping timeout: 252 seconds]
sylvie[m] has quit [Ping timeout: 252 seconds]
domenkozar[m] has quit [Ping timeout: 252 seconds]
groggy[m] has joined #nixos
gudea[m] has joined #nixos
Dan[m]8 has joined #nixos
kxlljoyy[m] has joined #nixos
peterhoeg[m] has joined #nixos
Gopal[m] has joined #nixos
jak[m] has joined #nixos
ndarwincorn has joined #nixos
Sisyphe[m] has joined #nixos
MiloIgnis[m] has joined #nixos
Seb[m] has joined #nixos
slabity has joined #nixos
luftmensch[m] has joined #nixos
rycee has joined #nixos
sylvie[m] has joined #nixos
domenkozar[m] has joined #nixos
pinage404[m] has joined #nixos
t58 has quit [Ping timeout: 276 seconds]
mexisme has quit [Ping timeout: 246 seconds]
jluttine has joined #nixos
<jackdk>
I am trying to print to my brother hl31250cdn. I hacked up a derivation for the driver based on some of the other brother drivers, and pushed to https://github.com/endgame/nixpkgs/tree/hl3150cdn , and added it to services.printing.drivers, but it does not show up as a driver I can select in the CUPS web UI. anyone have ideas?
<clever>
hpfr[m]: if you set openDefaultPorts, it just opens 2 ports, one tcp, one udp
<clever>
if you have changed the ports, you need to set allowedTCPPorts and allowedUDPPorts yourself, to whatever you changed them to
<buckley310>
if i have a local package, with a `src` attribute pointing to a local zip file, does nix rehash that file every time i rebuild?
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nikola_i has quit [Ping timeout: 245 seconds]
kvda has joined #nixos
mmlb has quit [Ping timeout: 276 seconds]
<jackdk>
I am trying to print to my brother hl31250cdn. I hacked up a derivation for the driver based on some of the other brother drivers, and pushed to https://github.com/endgame/nixpkgs/tree/hl3150cdn , and added it to services.printing.drivers, but it does not show up as a driver I can select in the CUPS web UI. anyone have ideas?
<hpfr[m]>
clever: oh, I thought there was some kind of user set of ports but I guess that doesn't make much sense. it just means each user would need its own set of ports that you would need to manually configure on all devices to match users or whatever. so I'll use the defaults, gotta do it manually cuz I'm using the home-manager module
<hpfr[m]>
clever: in hindsight maybe the home manager module was not a good idea? I have to be logged in on my server to start syncthing there
<hpfr[m]>
But on the other hand permissions were weird with the nixos module since the directory I’m syncing is under my user home directory
mmlb has joined #nixos
abathur has quit [Ping timeout: 276 seconds]
Rusty1 has quit [Quit: Konversation terminated!]
ardumont has quit [Ping timeout: 276 seconds]
dansho has quit [Ping timeout: 268 seconds]
endformationage has quit [Ping timeout: 252 seconds]
Ariakenom has quit [Remote host closed the connection]
Ariakenom has joined #nixos
ddellacosta has quit [Ping timeout: 240 seconds]
Ariakenom has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @orivej-nixos pushed 2 commits to master: https://git.io/Jec1j
Cale has joined #nixos
ardumont has joined #nixos
Cale has quit [Ping timeout: 245 seconds]
infinee has joined #nixos
nikola_i has joined #nixos
Cale has joined #nixos
palo1 has joined #nixos
palo has quit [Ping timeout: 240 seconds]
palo1 is now known as palo
AluisioASG has quit [Ping timeout: 276 seconds]
kini is now known as fs
fs has joined #nixos
fs has quit [Changing host]
drakonis_ has quit [Ping timeout: 265 seconds]
fs is now known as kini
kini has quit [Changing host]
kini has joined #nixos
drakonis_ has joined #nixos
<{^_^}>
[nixpkgs] @vbgl opened pull request #70366 → OCaml: make version 4.07 the default one → https://git.io/JecMl
slack1256 has quit [Remote host closed the connection]
Jackneill has joined #nixos
nexgen has quit [Remote host closed the connection]
nexgen has joined #nixos
noudle has quit []
<dminuoso>
Nix reports an error `value is a string with context while a set was expected,`
<dminuoso>
What is a "string with context"?
<sphalerite>
dminuoso: a string with dependencies on store paths
<dminuoso>
Ah I see.
graf_blutwurst has joined #nixos
<davidtwco>
hey folks, I want to contribute a new version of a package and I’m not sure how best to go about it. nixpkgs currently has an intel-ocl package that’s pretty old (5.0), it only supports OpenCL 1.x. Intel have a new version (18.1) that supports OpenCL 2.1 but I’m not entirely sure if it’s backwards compatible. They’ve got an even newer experimental version (2019.9.9.0.0901) that I’ve been using successfully
<davidtwco>
and has support for things I need. I don’t know which version to put a package up for, or if it should be a new package.
<dminuoso>
Okay some packages are really scary. Finding options that cannot clearly work because they are configured for the wrong type, don't people actually try out their modules? :/
nexgen has quit [Remote host closed the connection]
<dminuoso>
What's the protocol for overriding nixos modules? Are nixos module overlays?
ee194350 has quit [Read error: Connection reset by peer]
ee194350 has joined #nixos
o1lo01ol1o has joined #nixos
philr has joined #nixos
civodul has joined #nixos
dermetfan has joined #nixos
dermetfan has quit [Client Quit]
asymptotically has joined #nixos
dermetfan has joined #nixos
dermetfan has quit [Client Quit]
<{^_^}>
[nix] @edolstra pushed to flakes « Merge release.nix, shell.nix and release-common.nix into flake.nix »: https://git.io/Jecyi
Okinan has joined #nixos
<zfnmxt>
What version do you specify for a package packed off the HEAD of a repo instead of a given release?
asymptotically has quit [Remote host closed the connection]
<Yaniel>
depends on why it was packaged off the HEAD
knupfer has joined #nixos
<Yaniel>
next.version-pre1 is something I did while waiting for a new release of renderdoc that fixed some issues with the nix build
asymptotically has joined #nixos
<Taneb>
I'm trying to set up a service which cals "hydra-queue-runner --status" and it fails due to not being able to find a home directory (because it's a DynamicUser systemd service). It feels odd to me that a home directory is needed for this, though
<zfnmxt>
Yaniel: I just chose HEAD because the last release was 6 months ago and there have been bug fixes since.
fendor has joined #nixos
<zfnmxt>
I saw a few packages with something like "version = "git-2019-08-05"".
<Yaniel>
I'd prefer to use something like the output of `git describe`
ng0 has joined #nixos
<Yaniel>
I don't think the version comparison logic in nix handles your format very well
<zfnmxt>
That seems better. Is there no standard for this sort on thing for nixpkgs?
<zfnmxt>
s/on/of
<Yaniel>
AFAIK no
<zfnmxt>
I'll just go with `git describe` then. Thanks for the help.
<etu>
zfnmxt: I often see "unstable-2019-10-04" as a suggested format
o1lo01ol1o has quit [Remote host closed the connection]
psyanticy has joined #nixos
o1lo01ol1o has joined #nixos
infinee has quit [Ping timeout: 268 seconds]
<bahamas>
hello. I installed the gdal library and I'm looking for the .so file. I checked the outPath but when I try to list that directory, I get "No such file". what am I missing?
o1lo01ol1o has quit [Remote host closed the connection]
graf_blutwurst has quit [Ping timeout: 264 seconds]
o1lo01ol1o has joined #nixos
mexisme has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
sigmundv_ has joined #nixos
l33 has quit [Remote host closed the connection]
l33 has joined #nixos
spacefrogg has quit [Quit: Gone.]
o1lo01ol1o has quit [Remote host closed the connection]
<hyperfekt>
Are any still at Delft like I read somewhere?
<{^_^}>
[nixpkgs] @marsam opened pull request #70379 → gitAndTools.delta: init at 0.0.12 → https://git.io/Jec9q
Okinan has quit [Quit: Konversation terminated!]
o1lo01ol1o has quit [Remote host closed the connection]
Okinan has joined #nixos
logzet has joined #nixos
<{^_^}>
[nixpkgs] @zfnmxt opened pull request #70380 → variety: init at 0.7.2-96-g3afe3ab → https://git.io/Jec9s
Okinan has quit [Quit: Konversation terminated!]
avn has quit [Ping timeout: 265 seconds]
<tetdim>
Ericson2314, found another error in your original attempt
__monty__ has joined #nixos
<tetdim>
libexpr doesnt import a bunch of stuff
MarcWeber has quit [Remote host closed the connection]
<Taneb>
It's a little annoying that the "manage machines" admin page of a Hydra server has always-greyed-out checkboxes so you can't actually manage the machines from there
Okinan has joined #nixos
nexgen has quit [Remote host closed the connection]
avn has quit [Read error: Connection reset by peer]
avn has joined #nixos
Izorkin has quit [Ping timeout: 240 seconds]
Izorkin has joined #nixos
jacobat has joined #nixos
<jacobat>
I'm trying to install nixos on a machine with only wifi but my wifi card isn't recognized. So I'm trying to use my iphone to tether but nixos isn't picking up my iphone either. How can I get nixos to use my iphone as a network connection?
bahamas has quit [Quit: leaving]
<inkbottle>
I've been trying to get a pdf version of the manual through chromium/print
<inkbottle>
But the conversion is not very good, like "example 2.4" is not transformed correctly.
knupfer has quit [Read error: Connection reset by peer]
knupfer has joined #nixos
Ariakenom has quit [Ping timeout: 246 seconds]
Ariakenom__ has joined #nixos
xkapastel has joined #nixos
<{^_^}>
[nixpkgs] @matthiasbeyer opened pull request #70390 → tmate-ssh-server: init at 2.3.0 → https://git.io/JecQm
phreedom has quit [Remote host closed the connection]
phreedom has joined #nixos
Ariakenom_ has quit [Ping timeout: 240 seconds]
<l33>
just going over the nix pills examples, 12.3. Packaging graphviz. if i do nix-build graphviz.nix .... i get ... /nix/store/86k2z0mc023y9qxn8gln4a6kj2a759gx-builder.sh: line 7: tar: No such file or directory
<l33>
if i run nix-shell graphviz.nix ... echo $src is /nix/store/y0wl6h78qx234bv3jnkffnha4nh8cpgh-graphviz-2.38.0.tar.gz
<l33>
..... hmmm so what am i missing here?
<l33>
ls -lha $src
<l33>
-r--r--r-- 1 root root 25M Jan 1 1970 /nix/store/y0wl6h78qx234bv3jnkffnha4nh8cpgh-graphviz-2.38.0.tar.gz
<inkbottle>
I've converted the manual to pdf using https://www.web2pdfconvert.com and the output looks exactly as the original (not affiliated in any way; but very useful; chrome conversion was just not good)
<niso>
jacobat: you should be able to setup tethering like on any other machine
knupfer has quit [Remote host closed the connection]
<grw>
jacobat: you need services.usbmuxd.enable = true
<grw>
or maybe there is another way to do it, but this works for me
<jacobat>
thanks
<exarkun>
Alright so yesterday clever and tilpner helped me understand that https://github.com/LeastAuthority/privacypass/blob/master/ristretto.nix#L26-L29 is bogus and the Darwin hash there is correct and the other is junk. But my NixOS system still likes the junk. I don't know what I need to do to ... clean up my nix store, I guess? So that my stale cached something goes away and I get the same behavior as a fresh
<exarkun>
nix store.
<exarkun>
If I change the cargoSha256 to something obviously wrong I just get a complaint that the obviously wrong value doesn't match the wrong value encoded in that expression. It doesn't try to fetch anything new.
<{^_^}>
[nixpkgs] @romildo merged pull request #70174 → Update numix icon themes to version 19.09.20 → https://git.io/JenP6
<{^_^}>
[nixpkgs] @doronbehar opened pull request #70392 → ###### Motivation for this change → https://git.io/JecQ2
equivrel has quit [Ping timeout: 240 seconds]
<exarkun>
Also, it seems like crates are indexed in a git repository so I wonder how likely it is that the hypothesized explanation - that someone uploaded new code w/ the same version number - is correct
captn3m0 has quit [Ping timeout: 240 seconds]
<qyliss>
What's the idiomatic way to have an option that can either be a path or a string (where the value of the string would just be written to a path). Is it to just have two options?
<exarkun>
gchristensen: I don't really understand the Rust/cargo toolchain well enough to say. It's a cargoSha256 that mismatches. That means that ... Cargo.lock is different from expected? And that covers all of the dependencies?
<exarkun>
I'm reading the Nix cargo support implementation now to try to figure out how to trace it further but I'm not making a lot of progress
<exarkun>
so any further tips welcome :)
<exarkun>
Okay, also I was just running an experiement
<tilpner>
exarkun: Try to manually cargo-vendor on both systems, then nix-hash both outputs
<exarkun>
and the cargoSha256 matches if I build against current nixpkgs-unstable
<tilpner>
exarkun: AIUI cargoSha256 guards the whole process from Cargo.lock to vendored dependencies, which has a lot of potential for impurities and indeterminism
<exarkun>
tilpner: I don't know how to cargo-vendor things, I'm not actually a Rust programmer
<tilpner>
exarkun: So a different cargo-vendor version might already break things
<exarkun>
It seems like I should bisect nixpkgs to see where this goes from fail to succeed, though
<exarkun>
is `cargo-vendor` a program?
silver has joined #nixos
vesper11 has quit [Ping timeout: 246 seconds]
<tilpner>
It's apparently integrated into cargo now
<tilpner>
So you could just try running "cargo vendor" in your project directory
<exarkun>
What do you think about the fact that changing the version of nixpkgs changes the behavior
<tilpner>
exarkun: Changing the nixpkgs version can also change every other version that matters here
<tilpner>
exarkun: So cargo and dependencies
<tilpner>
If your nixpkgs version change does not alter the cargo input hash, that would worry me
<exarkun>
well, I guess the point is that changing nixpkgs changes the hash to *exactly* the other value
<exarkun>
not just some random 3rd value
<tilpner>
Were darwin and nixos using the exact same pinned nixpkgs?
<exarkun>
also am I kinda hearing that cargo support in nix is fragile and half baked? :)
<tilpner>
Well, this is exactly the reason I recommended not using it yesterday (in favor of crate2nix)
<exarkun>
okay, I'm happy to stop messing with this solution if crate2nix is better
<tilpner>
It's not necessarily, but it's worth a try
<tilpner>
(It worked on 1 out of 3 attempts for me, so...)
<exarkun>
up 'til a minute ago I thought the problem was I had screwed up my nix store but now it seems that's probably not the case
<exarkun>
ha.
<tilpner>
No, while I wouldn't exclude store corruption, I'd put it at the very end of my list of things to investigate
<{^_^}>
[nixpkgs] @lheckemann pushed 2 commits to release-19.09: https://git.io/Jec7O
Makaveli7 has quit [Quit: WeeChat 2.6]
bahamas has joined #nixos
<exarkun>
well it built
<exarkun>
I installed 0.5.1 but it calls itself 0.5.0, not a promising start :)
<exarkun>
welp
<exarkun>
it generated something
<inkbottle>
Plasma Panel configure system tray shows empty transparent window
<wucke13>
WHat could cause requireFile to fail, even though the file is part of the store?
emacsomancer has quit [Ping timeout: 268 seconds]
<exarkun>
but I think I might have found the problem with my previous nix expression ... there was a stray edit to the version passed to buildRustPackage...
jacobat has quit [Quit: leaving]
jacob has joined #nixos
<exarkun>
little bit confused why that leads to a cargoSha256 mismatch (and why different nixpkgs treat it differently) but fixing it gets rid of the mismatch :/
<exarkun>
unless this nix-shell is screwing with things
emacsomancer has joined #nixos
magnetophon has quit [Ping timeout: 245 seconds]
<exarkun>
aaaa now it succeeds locally with either hash
fendor has quit [Read error: Connection reset by peer]
infinee has joined #nixos
vesper11 has joined #nixos
freeman42x has joined #nixos
Okinan has quit [Quit: Konversation terminated!]
<exarkun>
it fails with 1111.... and complains that it wants the value that originally worked on my system. but it's also totally happy with the different value that worked in my darwin tests, instead.
<tilpner>
exarkun: The problem with fixed output derivations is that Nix will no longer second-guess what you say
<tilpner>
exarkun: If you say "the cargoSha256 of this is foo", and there is something with that hash in your store (regardless of how it got there), it will just take that
<exarkun>
okay, so it's not really succeeding with the original hash, it's just noticing that it succeeded at some point in the past and skipping all the work
<exarkun>
but of course I can't tell the difference between this case and success
<tilpner>
You can delete it carefully with nix-store (but never ignore liveness!)
<exarkun>
because I don't know how to clean anything up on nixos yet
<tilpner>
Then you need to ensure it's dead before using nix-store --delete on it
<exarkun>
Okay I know the hash of it, what do I do with that?
<exarkun>
The old cargoSha256 is 1qbfp24d21wg13sgzccwn3ndvrzbydg0janxp7mzkjm4a83v0qij but there is no /nix/store/1qbfp24d21wg13sgzccwn3ndvrzbydg0janxp7mzkjm4a83v0qij*
<tilpner>
exarkun: sqlite3 /nix/var/nix/db/db.sqlite 'select path from validpaths where ca="fixed:r:sha256:1qbfp24d21wg13sgzccwn3ndvrzbydg0janxp7mzkjm4a83v0qij";
<tilpner>
:c
<exarkun>
oof
<exarkun>
tilpner: sqlite3 thinks that select is a write and won't let me :\
<exarkun>
meeting :/ bbiab
<Ox4A6F>
<exarkun "Okay I know the hash of it, what"> exarkun: nix-collect-garbage --help
<tilpner>
exarkun: Put -readonly in. And you added closing ', right?
__Sander__ has joined #nixos
<exarkun>
tilpner: I did both of those, yea, no difference :/
<tilpner>
Huh
<tilpner>
exarkun: This is already so ugly, you might as well copy db.sqlite to your home directory and try it there
<tilpner>
That should make sure no nix-daemon tries to access it concurrently
<nakkle>
Hey there! If I have a default.nix, which evaluates to an argument set, I can do `nix-build ./default.nix`, and it builds all the elements that directly evaluate to a derivation, but if one of the elements of this argument set evaluates to another argument set, it does not build anything from that set. Is there some kind of flag to enable such "recursive" nix-builds?
civodul has quit [Quit: ERC (IRC client for Emacs 26.3)]
selfsymmetric-mu has quit [Remote host closed the connection]
<inkbottle>
I have the following error with Application Launcher, just after removing the previous panel and creating a new one: https://paste.debian.net/1104417
<inkbottle>
like: Could not load: file:///nix/store/z3b3s19726zjcg2zwvdafgd141761...
<inkbottle>
It was working before; don't know what got mixed up
<exarkun>
tilpner: cool. that gave me 3 paths, all which nix-store --delete was able to delete
<inkbottle>
I had to reboot, something was mixed up with plasma, logout wasn't working anymore; that might explain
<{^_^}>
[nixpkgs] @stigtsp opened pull request #70400 → perlPackages.DeviceMAC: init at 1.00 → https://git.io/Jec5g
<exarkun>
though deleting them doesn't appear to make any difference :)
infinee has quit [Ping timeout: 268 seconds]
infinee has joined #nixos
<tilpner>
If those paths re-appeared, that means the derivation actually builds to that hash, which is good
dansho has quit [Quit: Leaving]
domogled has joined #nixos
<exarkun>
none of them re-appeared
<exarkun>
I went ahead and deleted the store paths associated with the other hash too, and they also did not reappear :o
<exarkun>
there's some *other* store path listed in result/nix-support/propagated-native-build-inputs
<tilpner>
Ahh, that would be because you updated nixpkgs
<tilpner>
Only with the exact same input woulds those paths reappear
<tilpner>
*inputs would
<exarkun>
I did my new-nixpkgs test w/ nix-build -I nixpkgs=... so I think these builds should be using the same nixpkgs as the earlier failing ones
<exarkun>
but it is possible my nixpkgs changed at some point in time over the whole course of development on this thing, I can't be quite sure
<exarkun>
is there a reverse-depends search I can do to see what is keeping this newest store path alive?
<exarkun>
--referrers I guess
<exarkun>
heh, 20 versions of my python project that depends on the rust thing
domogled has quit [Quit: domogled]
<{^_^}>
[nixpkgs] @bzizou opened pull request #70401 → Quast: init at 5.0.2 → https://git.io/Jec5M
rmk35 has joined #nixos
<exarkun>
`nix-store --query --referres x` returns just `x` but `nix-store --delete x` says `x` is in use
<rmk35>
Probably a dumb question, but I couldn't google the keywords better: In NixOS is there a philosophy for installing things with "environment.systemPackages" vs "programs.<name>.enable"?
<gchristensen>
rmk35: some programs require your system to be changed in order to function. those need a `programs.X.enable` option.
<clever>
when the bash it started exits, sudo will then exit with the same error code
<clever>
so its not really leaking a process
<bahamas>
I'm working on creating a dev setup with nix+nixos for Django. should I use a shell.nix inside the nixos vm where I start the development server or should I start the dev server as a nixos service?
<exarkun>
clever: huh cool, thanks.
<bahamas>
a problem with the first approach is that the project depends on some libraries: gdal, geos. in order for django to find them, I need to define some env vars that point to them
<tilpner>
exarkun: It does make some sense, I think, it's just really unexpected
horner has quit [Ping timeout: 240 seconds]
AluisioASG has joined #nixos
<tilpner>
exarkun: If you do it randomly, there's a chance enough reboots will get your system booting, so that you can fix the problem
<{^_^}>
[hydra] @edolstra pushed to flake « Use the overlay from the nix flake »: https://git.io/JecAk
<exarkun>
in my experience its not really random, it's always NetworkManager ;)
<exarkun>
it's too bad that there's no consistency checking when creating a systemd configuration so that you could just skip having this problem entirely
philr has quit [Ping timeout: 276 seconds]
<tilpner>
sysinit.target: Job local-fs.target/start deleted to break ordering cycle starting with sysinit.target/start
alex`` has quit [Quit: WeeChat 2.6]
<tilpner>
That at least doesn't mention NM
alex`` has joined #nixos
<tilpner>
Now I can finally get back to debugging my kresd config, yay
<esclear>
Is it okay to add the pizzaware license to licenses.nix or shall I use a placeholder?
<qyliss>
esclear: is it commonly used? has it been evaluated by the FSF or OSI?
<ivan>
does anyone know which commit I need to revert to get Qt4 apps to not look like the ancient GTK theme on 19.09?
rsoeldner_ has quit [Ping timeout: 246 seconds]
<esclear>
qyliss: I've seen multiple uses of it, but could not find any evaluation. It's functionally equivalent to the beerware license (which is already in licenses.nix), but proposes to buy the author a pizza instead of a beer, if the software is deemed useful. But it is a free license.
Jackneill has quit [Remote host closed the connection]
<infinisil>
Ohh!
<infinisil>
I saw that PR's title but didn't think much of it
ericdallo_ has joined #nixos
MichaelRaskin has joined #nixos
horner_ has quit [Remote host closed the connection]
ericdallo_ has quit [Client Quit]
Chiliparrot has joined #nixos
<ivan>
my /run/current-system/sw/share/themes has an Adwaita with a gtk-2.0 but lxappearance sees only Raleigh
<ivan>
hmm I can at least override the theme with GTK2_RC_FILES=/run/current-system/sw/share/themes/Adwaita/gtk-2.0/gtkrc
<arcnmx>
semi-related point that's maybe been discussed to death but it is fairly frustrating how often everything gets stuck/broken; are there any plans/opinions on things like homu/bors style merging?
misuzu has joined #nixos
<infinisil>
arcnmx: What's that?
<clever>
infinisil: bors is a bot, that will generate a merge commit, then run CI (in this case, a full hydra build??) and only push that merge to master if CI passes
magnetophon has joined #nixos
<clever>
infinisil: if you tell bors to merge several branches at once, it will merge all of them together, then test them as a single batch
<clever>
it can solve problems like PR1 passes CI, PR2 passes CI, but if you merge both, CI fails
<infinisil>
I see
<infinisil>
So it would be like an extended @ofborg
<clever>
and travis building the merge branch cant catch the above, because it only tests the merge commit when you push
rsoeldner_ has joined #nixos
<clever>
so if you push PR1, and it passes, then PR2 gets merged, then PR1+master now fails
<qyliss>
clever: does it queue?
<clever>
and travis doesnt re-test every time master moves
<clever>
qyliss: if a PR merge is pending checks, any further attempts to merge one enter a queue
<qyliss>
so, if I say "merge PR1; merge PR2", it'll test PR1 first, then if it fails test PR2?
<qyliss>
oh I've been looking for this :D
<clever>
qyliss: and once the current job is done, it will create a merge commit, containing everything from the queue, and do them in a second batch
<clever>
in your case, it would test pr1, and then pr2, as seperate things
<clever>
but if you tell it to merge pr1, pr2, and pr3, it will start testing pr1
<{^_^}>
[nixpkgs] @timokau pushed 3 commits to release-19.09: https://git.io/Jecpc
<clever>
and when done, it will test pr2+pr3, as a second batch
<clever>
it only waits for a batch if tis already busy doing something else and must wait
ris has joined #nixos
<arcnmx>
infinisil: atomic merging basically? the tl;dr is instead of merging PRs you approve them and get put into a queue, applied to master, then once all tests pass gets merged
<exarkun>
It's too bad GitHub doesn't just give you that.
<clever>
exarkun: it kind of does have some tools to support it
<arcnmx>
and yeah, can do things like batch changes for efficiency
<exarkun>
Oh?
<clever>
exarkun: if you try to fetch the pull/42/head branch, you get the tip of the PR
<emily>
you can block merging on CI passing and make it so that only a bot can merge i think
ericdallo` has quit [Ping timeout: 268 seconds]
<clever>
exarkun: if you fetch pull/42/merge, you get a commit that is the result of merging the pr into master (but that commit wont be pushed to master)
<emily>
dunno if it lets you integrate clicking the merge button to putting it into a queue but like, people can just do @ofborg merge or whatever
m1cr0m4n has quit [Ping timeout: 240 seconds]
<clever>
exarkun: so, you could configure hydra to build pull/$PR/merge, on every PR
<exarkun>
"run a bot to do your merges" makes it impractical for 95% of projects to bother
<clever>
exarkun: BUT, hydra will now rebuild EVERY SINGLE PR, on EVERY SINGLE PUSH TO MASTER
<{^_^}>
[nixpkgs] @7c6f434c pushed to master « xtreemfs: fix build with openssl-1.1 »: https://git.io/Jecpz
<exarkun>
So yea, all that stuff is fine and cool, but it would be better if the big green button just did the right thing :)
<emily>
exarkun: you need CI to support this inherently
<emily>
but yeah, it'd be nice if GitHub offered more of the glue for it
<exarkun>
emily: sure. and github supports CI. it even knows if CI passed or failed!@
<emily>
right
<emily>
well, that's quite new!
<emily>
give them time ^^
<exarkun>
well, ~10 years or something...
m1cr0m4n has joined #nixos
<clever>
exarkun: the problem, is that travis merge builds, are if the pr passed with the old value of master
<emily>
yeah, maybe :<
<clever>
exarkun: and if master changes after that test, things can break
<clever>
which is why many projects demand that you rebase ontop of master before merging
<exarkun>
clever: I guess I'm not clear on the point here. It still seems to me like it would be nice if github gave you this workflow instead of the one it gives you.
Zer0xp has joined #nixos
<clever>
exarkun: the problem, is that github has to re-trigger CI, on the result of merging the commit into master
<clever>
exarkun: bors does that, by just pushing the merge commit to a bors/staging branch
Mr_Keyser_Soze has joined #nixos
<exarkun>
clever: are you saying it wouldn't be nice, or that it would be hard to implement, or that it is impossible to implement?
<clever>
and then if the status checks are green, bors will do it
<clever>
exarkun: the bors logic could be baked into github, just have github itself push the merge to a dummy branch, and wait for ci to pass on that dummy branch
<exarkun>
I think it is probably "hard to implement" and that's why github should do it instead of every single project on github doing it
<clever>
and if it passes, fast-forward master to that merge commit you just tested
<clever>
but, if any naughty admin pushes directly to master, you cant fast-forward, and ci has to start over
<exarkun>
the idea of a workflow where if you break it, it breaks, doesn't bother me _too_ much
arjen-jonathan has joined #nixos
<exarkun>
completely academic, not like github is going to do anything I want them to
rsoeldner_ has quit [Ping timeout: 252 seconds]
<clever>
exarkun: i have been thinking of rewriting bors as well, while it does work, the errors when it doesnt are horid, typically just ruby backtraces, lol
iqubic has quit [Quit: heading to college]
<exarkun>
I love software that, when it fails, does so in a coherent and easily understood manner
<exarkun>
Or, well, maybe not that. But I hate software that doesn't do that.
<Ericson2314>
bendlas: Hey, so if you have cycles to burn, I think the next step is to modify Clang to do incremental preprocessing
<Ericson2314>
I suspect this shouldn't be too bad
<Ericson2314>
Because of dynamic scope, and lack of being able to tell what is a macro identifer other than it being in scope, there is very little one can do but pause and resume
<Ericson2314>
and pausing and resuming will fit an existing sequential implementation
<{^_^}>
[nixpkgs] @Lassulus pushed commit from @doronbehar to master « luaPackages.pulseaudio: init at 0.1 »: https://git.io/JechM
<selfsymmetric-mu>
Hi! I think it would be great to have scoped logs on nixos-rebuild. That is, I would like to be able to see which derivation is responsible for each line of log output. Now, I don't think that that's exactly _controversial_, but I can't find a tracking ticket for it.
<infinisil>
I only used hnix to find urls, manually implemented string replacement
<selfsymmetric-mu>
Does anyone know of a such a scoped-logs tracking ticket? If one doesn't exist I think I should make one. Which repository should that be submitted to? My guess is https://github.com/NixOS/nix. Thoughts?
alex`` has quit [Ping timeout: 265 seconds]
alex`` has joined #nixos
<selfsymmetric-mu>
I'm not claiming that enabling scoped logs is _easy_, mind you, just that I think there should be a tracking issue.
<qyliss>
selfsymmetric-mu: it'll be Nixpkgs if you're talking journald logs
rsoeldner_ has quit [Ping timeout: 276 seconds]
<selfsymmetric-mu>
qyliss: What are the logs that occur during `nixos-rebuild switch`?
<qyliss>
I'm not sure how useful a tracking issue would be if nobody intends to work on it though.
<qyliss>
selfsymmetric-mu: I'm not sure I understand what you mean there
<qyliss>
example?
<infinisil>
qyliss: Pretty sure they mean nix-build logs in general
<selfsymmetric-mu>
infinisil: That's right. The stdout on the build.
<infinisil>
How multiple jobs get overlapped stdout
<qyliss>
oh, right
<infinisil>
So that would be in the Nix repo, not nixpkgs
<qyliss>
yeah
<selfsymmetric-mu>
Great, thank you infinisil. I'll look there and create an issue if I don't see one.
[Leary] has joined #nixos
<bendlas>
@Ericson2314:matrix.org: next step towards ... what exactly?
<bendlas>
I feel that anything finer-grained than per - `.o` dependencies would rather be a liability. And that's already provides by `-MMD` (or whatnot).
<bendlas>
Also, gcc support.
<selfsymmetric-mu>
I think there is value in documenting feature requests even if nothing is immediately happening. That is because other curious people can determine whether the feature is 1) being worked on already 2) available for contribution 3) rejected because it is ill-defined for some fundamental reason, etc.
<infinisil>
Yeah agreed selfsymmetric-mu
<selfsymmetric-mu>
:)
Lears has quit [Ping timeout: 276 seconds]
<Ericson2314>
bendlas: well what exactly did you get caught on before?
<arcnmx>
sounds like it could be relevant to the nixpkgs repo though, because basically you could fix it by using nix build -L instead of nix-build?
<arcnmx>
(if I'm understanding this correctly)
<bendlas>
<Ericson2314 "bendlas: well what exactly did y"> Chromium builds eating up hydra and everybody working on them. What did you think, you were responding to?
<Ericson2314>
bendlas: so if the build changes too much in between releases, or the rebuilds are due to e.g. stdenv changing or something, I am not sure how much incrementality will help you
<Ericson2314>
(My clang thing was yes, me responding more to the thread than your original issue)
<freeman42x>
is it possible to split a function call on multiple lines? like replace "bla" "abc" "blaqweqwe". Get each argument on a different line
mexisme has quit [Ping timeout: 265 seconds]
<freeman42x>
sorry, wrong channel
<{^_^}>
[nixops] @edolstra pushed to flake-support « Support deploying from a flake »: https://git.io/Jecjt
<selfsymmetric-mu>
Hrm. Why don't I see `-L` in the `nix-build` manpage for 2.3?
<michaelpj>
yeah, I tried to swap in the nix 2 commands into nixos-rebuild but it turns out to be Not That Easy
boxscape has quit [Remote host closed the connection]
<niksnut>
selfsymmetric-mu: the -L option is only supported by the 'nix' command
<selfsymmetric-mu>
niksnut: Ah, thank you.
drakonis__ has joined #nixos
<bendlas>
<Ericson2314 "bendlas: so if the build change"> It had feared that. Still incrementalizing will help with:
<slabity>
I'm having trouble getting a derivation to build correctly. It keeps failing to find a library, but one of the `buildInputs` is a package that includes that library. Is there something else that needs to be added?
drakonis_ has quit [Ping timeout: 240 seconds]
<arcnmx>
selfsymmetric-mu: Clearly instead of nixos-rebuild we should all be using something like nix run -Lf '<nixpkgs/nixos>' system -c switch-to-configuration switch
<infinisil>
slabity: Gotta needa whole lot more details than just that to help
<Ericson2314>
bendlas: OK then then you can just ignore the depfile stuff ih ninja
<exarkun>
Hm, I think I'm using the default compiler... but then, ghc 8.6.5 is also being built ... :/
<bendlas>
<Ericson2314 "bendlas: OK then then you can ju"> Yeah, I mean, there is shared interest, because incrementalizing needs the depfiles as well, but since ninja doesn't (yet?) provide pregenerating them, we'll have to rewrite cc invokations in a pregeneration pass, right?
<slabity>
Infinisil: I would have thought it would invclude all NIX_LDFLAGS
<infinisil>
exarkun: What are you building at what version?
ent has quit [Ping timeout: 240 seconds]
<exarkun>
building some of my own software at the last version of nixpkgs before openssl went to 1.1.1d
datakurre has quit [Ping timeout: 252 seconds]
fendor has joined #nixos
<{^_^}>
[nixops] @edolstra opened pull request #1202 → [WIP] Flake support → https://git.io/Jecjd
liff has quit [Ping timeout: 250 seconds]
ent has joined #nixos
liff has joined #nixos
datakurre has joined #nixos
rsoeldner_ has quit [Ping timeout: 276 seconds]
<bsima>
does the guile package omit the info manual, or am I missing something? I don't see it in my INFOPATH
rsoeldner_ has joined #nixos
<slabity>
infinisil: I don't suppose there's a way I can modify my derivation to produce a pkgconfig file?
<infinisil>
slabity: I'd try first removing pkgconfig files from something in nixpkgs to see if it doesn't work without them
<bendlas>
@Ericson2314:matrix.org: So .. your drv trampoline at rfc#40 .. how would you recommend to compose this into a cc wrapper? Is the code in a state where somebody might try?
<infinisil>
slabity: Such a pkgconfig file looks to be pretty simple though, you could probably just hand-write it
wfranzini has joined #nixos
<exarkun>
infinisil: I have an openssl override that makes the default 1.1.1 instead of 1.0.whatever, I *guess* that's the cause but I can't seem to prove it by looking at stuff
fusion809 has quit [Remote host closed the connection]
<fuzen>
Does dockerTools build the packages for the target image or is it based on the host machine?
<infinisil>
exarkun: Oh that's definitely the cause
<infinisil>
Almost everything depends on openssl
<{^_^}>
[nixpkgs] @esclear opened pull request #70416 → fileshare: init at 0.2.4 → https://git.io/JeCef
<infinisil>
exarkun: You could use a more recent version of nixpkgs though, 1.1.1 is now the default
<exarkun>
infinisil: 1.1.1d is too new, unfortunately, I need exactly 1.1.1c
<exarkun>
infinisil: so I picked the revision before the revision where it went from 1.1.1c to 1.1.1d
captn3m0 has quit [Read error: Connection reset by peer]
janneke has quit [Quit: janneke quits Mes'sing]
<exarkun>
infinisil: my override was put in place before 1.1.1 was the default so maybe I don't need it now ... but if I don't need it then isn't it also a no-op?
<exarkun>
I am building it with `nix-build --keep-going -I nixpkgs=https://github.com/NixOS/nixpkgs-channels/archive/5d5cd70516001e79516d2ade8bcf31df208a4ef3.tar.gz nixos/system-tests.nix` but the fishy changes seem to pin nixpkgs to something else
<exarkun>
so now I wonder if it's actually using the nixpkgs I thought it was ...
troydm has quit [Ping timeout: 246 seconds]
rsoeldner_ has quit [Ping timeout: 252 seconds]
<infinisil>
Hm there's no default.nix file though?
<infinisil>
Oh
<infinisil>
I missed the nixos/system-tests.nix
drakonis__ has quit [Ping timeout: 240 seconds]
<infinisil>
Lol in the comments -> # Test the machines with a Perl program (sobbing).
<Ericson2314>
bendlas: because ninja only uses the depfiles for rebuilds, it should be ok to ignore them first time. the downside is the entire source repo + all built headers are availible for preprocessor step, which is hugely non incremental
<Ericson2314>
bendlas: there is no WIP C++ Nix code for RC 40
rsoeldner_ has joined #nixos
<esclear>
For some reason OfBorg did not act on a pull request I created. (#70416) Is there a possibility to trigger it manually somehow? I'm quite sure that the format is correct.
<exarkun>
infinisil: yea. so ... does that have an effect or not? if openssl is already 1.1.1 by default in nixpkgs, then isn't it a no-op?
fenedor has joined #nixos
<exarkun>
I guess I should read the definition of openssl and openssl_1_1 in the new nixpkgs I picked ...
<l33>
another love hate day with nixos
leah2 has joined #nixos
<infinisil>
It should be a no-op yeah
<exarkun>
I guess I have the bad habit of picking a new nixpkgs and not looking at what has changed
<exarkun>
but ... it's so huge ... I can't really look at all the changes
<infinisil>
That is a valid point :)
Okinan has joined #nixos
drakonis has joined #nixos
<l33>
change.....fear it or embrace it... (or both)
<infinisil>
exarkun: Would be great if you could isolate the thing that causes the ghc rebuild
<infinisil>
Try disabling overlays one-by-one
<exarkun>
I asked about visualization and someone pointed out https://github.com/craigmbooth/nix-visualize but I didn't try it yet. not sure why I think visualization would help, exactly.
<infinisil>
exarkun: I think `nix-store -q --tree` is good enough for this, just shows the tree in a text view and you can search for stuff
<exarkun>
I tried `nix show-derivation` on some stuff to try to trace it back to openssl but didn't find anything obvious
<infinisil>
nix-store -q --tree on the .drv file specifically
<exarkun>
some (all?) `nix-store` commands have the shortcoming that you have to have already built the thing?
<exarkun>
maybe I haven't tried it on .drv though
fendor has quit [Ping timeout: 245 seconds]
<infinisil>
Yeah this will work without it having been built
<exarkun>
now the path to the ghc drv has scrolled out of my terminal history
<infinisil>
exarkun: Replace nix-build with nix-instantiate
jd823592 has joined #nixos
<infinisil>
This will instantiate the derivation and print its path
<exarkun>
cool
<jd823592>
anyone successfully playing youtube videos in the `next` browser? I am getting "your browser does not currently recognize any of the video formats available"
<l33>
'next' browser?
<l33>
not used that i'm affraid.
<jd823592>
it's in nixpkgs master
<exarkun>
infinisil: oh yea that's pretty useful
<jd823592>
i've read that playback may depend on gst-plugins and also that gst-plugin discovery experienced difficulties on nixos
rsoeldner_ has quit [Ping timeout: 268 seconds]
<l33>
where did you read this?
<jd823592>
l33: the dependency on gst (https://aur.archlinux.org/packages/next-browser-git/), the difficulties with gst (i noticed some mentions somewhere in here https://nixos.org/nix-dev, I really could not tell if that was related, bause I have no detailed insight in what's going on in webkit regarding video playback)
<l33>
i've tried firefox and brave..both seems to work fine for video.. not much help i know......
<l33>
ah! gstreamer.... I've always had problems with that beast on other distro's...
<l33>
i used to see 'gst' and shudder with fear.
abathur has joined #nixos
<jd823592>
l33: i am also using ffox with no problems, gstreamer is my nemesis, everytime i have to interact with it I know I've lost
<{^_^}>
[nixpkgs] @Lassulus pushed commit from @danieldk to master « maturin: 0.7.2 -> 0.7.6 »: https://git.io/JeCvq
boxscape has joined #nixos
<hlolli>
hi, I'm in a firefighting mode on my server, a node-gyp build is failing on a systemctl service, where the c-compiler is always resolving to unknown (node-v64-linux-x64-unknown) while it should resolve to glibc. Is there any env-flag I can use to mask the name? Even if nixos uses unknown internally.
<clever>
hlolli: is the build happening under nix-build/nix-shell?
jb55 has quit [Quit: jb55]
<hlolli>
it's happening under system-wide nix-rebuild
<clever>
hlolli: can you pastebin the entire output when it fails?
<clever>
hlolli: are you running `yarn build` from a systemd service?
<hlolli>
since there's a prebuild binary, which runs when I run it as user and resolves correctly, I wouldn't mind node-gyp just downloading it
selfsymmetric-mu has joined #nixos
<{^_^}>
[nixpkgs] @Lassulus merged pull request #69214 → iso-image: add loopback.cfg support to boot iso from grub → https://git.io/JesJo
<{^_^}>
[nixpkgs] @Lassulus pushed to master « iso-image: add loopback.cfg support to boot iso from grub »: https://git.io/JeCvs
<hlolli>
yes, because I thought node-gyp could just rebuild it.
<hlolli>
but it doesn't make sense to go that path
<clever>
hlolli: dont do that, its not pure
<clever>
hlolli: do all building in nix, and just use the built result in the service
<selfsymmetric-mu>
What is the correct way to specify font hinting in my NixOS? Specifically I want to set antialiasing to true, hinting to none, and sub-pixel order to RGB.
<hlolli>
yes, I learn it the hard way
<clever>
hlolli: you probably want yarn2nix, to build the package
<selfsymmetric-mu>
I thought fontconfig would be the right place to set that but now I see that the penultimate rules are now apparently the default? It's a big confusing.
<clever>
selfsymmetric-mu: there is an entire fonts tree of options in configuration.nix
<selfsymmetric-mu>
Brilliant, thank you. So `fonts.fontconfig.hinting.enable = false` would be the right way to disable hinting, yes?
<hlolli>
yes, I've used that and node2nix, was hoping I didn't need to go down that path. But a non-important express endpoint suddenly came crucial over time :) typical lazyness that hunts a programmer in the back
<selfsymmetric-mu>
Hopefully XFCE will read that correctly in its settings.
<clever>
hlolli: builds will only function if you do them under nix-build or nix-shell
<clever>
hlolli: your likely not using nix-shell, so things fail in weird ways
<hlolli>
I was hoping that I could start the systemd as a normal user, in the normal user's environment
<hlolli>
I was able to get ssh keys like that
<clever>
hlolli: nixos also sets PATH differently, to only include what the service needs, so it doesnt happen to work due to what youve nix-env -i'd
<clever>
hlolli: to help reduce the "but it worked for me" situations
<hlolli>
I added the missing deps :P, dug my grave deeper
jb55 has joined #nixos
<bendlas>
<Ericson2314 "bendlas: because ninja only uses"> Unfortunately, we need them to generate the correct derivation tree. Ignoring would mean putting fewer than the necessary files into the `.drv` of a `.o` file, or all of them, neither of which does any good.
<hlolli>
Im going node2nix now.. thanks clever!
<bendlas>
<Ericson2314 "bendlas: there is no WIP C++ Nix"> Too bad. So IFD it is, for now.
<selfsymmetric-mu>
I've never been able to get PragmataPro embedded bitmaps to work. Even with `fonts.fontconfig.useEmbeddedBitmaps = true`. Instead they get automatically stuck to pixels in an ugly way. It's a shame since the author spent countless hours tuning them at many sizes.
<selfsymmetric-mu>
Of course we can wait for monitors to become effectively continuous, but it would be nice to use them.
<{^_^}>
[nixpkgs] @Lassulus pushed commit from @r-ryantm to master « sysvinit: 2.95 -> 2.96 »: https://git.io/JeCvo
<bendlas>
Ericson2314: Oh, you already said that .. so yeah, I'm thinking more along the lines of: generate tailored include directories, per `.o` file.
<clever>
bendlas: snack does similar with haskell, there is a dedicated binary, that will list the modules a given module depends on, and its ran via IFD, for each module
<clever>
bendlas: the major performance cost, is that IFD must happen serially, so it can take 20+ mins to do the first pass
<clever>
it would likely be far faster, if you batched it into a single IFD step, but then it costs more when nothing changes
<clever>
so id want it to be a non-IFD step, that you just generate the dep tree once ahead of time, and then import the file
selfsymmetric-mu has joined #nixos
hlolli_ has joined #nixos
<wrl>
hey all, I'm having considerable trouble getting the nixos installer to boot
<wrl>
i'm flashing the ISO onto a sdcard in a usb/sdcard reader using pv (same as dd, basically)
<clever>
wrl: are you writing to sd? or sd?1 ?
<wrl>
it boots fine into UEFI mode – /sys/firmware/efi/efivars/ exists (it's empty though)
<wrl>
clever: sd
<bendlas>
clever: for chromium, the initial build would need to be as fast as possible: 4G of source directory; 20K+ .o files
<bendlas>
`importJSON` did really good for melpa, so I think I'll go for that.
<wrl>
it stalls at "waiting for device /dev/root to appear" and then refuses to go further
<wrl>
i can launch the interactive shell and look around
<clever>
bendlas: id want to pre-generate the dep tree info, and ship that with the rev+sha256
<wrl>
despite the usb_storage module being loaded, i don't have /dev/sd* anything
rsoeldner_ has joined #nixos
<clever>
bendlas: but yeah, with 20k derivations, nix will still suffer massively
<wrl>
nor do i have the device nodes for the NVME SSD that would function as the installation drive
hlolli has quit [Ping timeout: 276 seconds]
<wrl>
this is arguably a bleeding-edge setup – amd 3400G APU in an asrock x570 motherboard (both released less than 6 months ago)
Neo-- has quit [Ping timeout: 276 seconds]
<bendlas>
<clever "bendlas: but yeah, with 20k deri"> Hopefully still less, that it does now 😅
<ajs124>
So I'm trying to use disabledModules with a module right now that has "(mkRemovedOptionModule ["services" "tt-rss" "checkForUpdates"] ''...'')" in its imports. On a scale of 1-10, how screwed am I?
<freeman42x>
and I think I need to use cabal v1 because I am using HIE
<freeman42x>
so I can't use cabal Nix builds
<HKei>
hello, this is a bit awkward but does anyone know how/where user display settings are stored for kde? I've played around with xrandr because it was refusing to output in 1440p, but now I'm not getting any output from KDE at all after I log in (everything's still working fine in the login manager)
<HKei>
I can't really change them using xrandr from a terminal session either because obviously xrandr only affects the current session...
<HKei>
(by no output I mean both of my monitors complain they're literally getting no output send to them and go to sleep)
<lassulus>
does it work as another user
<lassulus>
?
drakonis__ has quit [Quit: Leaving]
<HKei>
maybe? But if that's the case then settings are per-user, which means I *still* need to find out how to change the settings for *that* user unless I want to abandon that user account
<HKei>
also, is the 1440p thing a known issue? Googling it I could only find various people complain that they don't get 1440p offered as a mode; but I do, it just doesn't work
<lassulus>
ah sure, but at least you would know if its a per user problem and then you could start searching in .config/.local or whereever it could be
freeman42x has quit [Remote host closed the connection]
l33 has quit [Remote host closed the connection]
HKei has quit [Remote host closed the connection]
HKei has joined #nixos
<HKei>
yeah, seems the settings are per user
Ariakenom has quit [Read error: Connection reset by peer]
<{^_^}>
import-from-derivation (IFD) is when you evaluate nix from a derivation result, for example `import (pkgs.writeText "n" "1 + 1")` will evaluate to 2. This is sometimes problematic because it requires evaluating some, building some, and then evaluating the build result. It has been described as "such a nice footgun."
<clever>
freeman42x: but your probably better off just doing `nix-shell -p zlib` and then `cabal build`
<freeman42x>
clever, I really want a solution like, forever, so I want to avoid using nix-shell
<freeman42x>
clever, `/run/current-system/sw/bin/pkg-config: Python script, ASCII text executable`
<clever>
freeman42x: then that tool isnt actually doing what it claims to do
<freeman42x>
but... why would science lie to me?!
<clever>
freeman42x: did you read the readme?
<clever>
nix-pkgconfig relies on a database of mappings between pkg-config .pc files and the nixpkgs attribute they are provided by. A minimal example database (default-database.json) is included which can be installed via:
<freeman42x>
yeah, I will try the other install methods
shibboleth has joined #nixos
<clever>
freeman42x: in addition to installing the binary, you must also copy a db file to the right dir