<fresheyeball>
and is needed for the current relase of torchvision
<clever>
ah, so it could be that pytorch is just old
<zeta_0>
vaibhavsagar: thanks for responding to the issue, i wasn't sure if the problem was ihaskell, ghc or emacs related, anyways i'll go to the emacs guys and show them the problem
o1lo01ol1o has quit [Remote host closed the connection]
ddellacosta has quit [Read error: Connection reset by peer]
<fresheyeball>
clever: I updated it to the latest release!
<fresheyeball>
how did they manage to release torchvision when it depends on pytorch unreleased
<fresheyeball>
fishy fishy
<clever>
fresheyeball: is the sha256 correct?
ddellacosta has joined #nixos
<fresheyeball>
yes
<fresheyeball>
I just tried downgrading torchvision 0.4.1
<fresheyeball>
and it still needs that file
<fresheyeball>
hmm actually
<fresheyeball>
if I mess with the hash for pytorch I am not seeing nix complain
<zeta_0>
i installed the ghc package called `haskellPackages.brick` in home.nix, and when i use it in a haskell program it throws the error: `ghcmod could not find module Brick`?
<clever>
zeta_0: haskell modules are only ever visible if you correctly use ghcWithPackages
<fresheyeball>
zeta_0: I would not advice you to manage your projects that way
<clever>
zeta_0: you must use the ghc that comes from `haskellPackages.ghcWithPackages (ps: [ ps.brick ])`
<clever>
zeta_0: and i would recomend using a shell.nix and default.nix to provide that, so you dont get ghc's mixed up between projects
nornagon has quit [Quit: Ping timeout (120 seconds)]
jgt2 has quit [Ping timeout: 250 seconds]
<fresheyeball>
clever: ok confirmed, the hash for pytorch is correct
nornagon has joined #nixos
<fresheyeball>
I am going to try pointing pytorch at HEAD
<Jonathan83>
I am trying to get a nix-shell environment to build EmptyEpsilon. I copied the EE nix expression from nixpkgs to default.nix. But nix-shell is complaining.
<Jonathan83>
I tried `nix-shell '<nixos>'` and it still is not working.
<Jonathan83>
I guess I am going about it wrong, what am I missing?
Synthetica has quit [Quit: Connection closed for inactivity]
<jared-w>
zeta_0: did you run something like: nix-shell -p "haskellPackages.ghcWithPackages (p: [p.brick])" --run "runhaskell HelloWorld.hs"
<clever>
zeta_0: if you install ghc with home-manager, you will run into problems when you need a different ghc next week
nornagon has quit [Ping timeout: 240 seconds]
<jared-w>
Yes, this is something that's a massive departure from the normal way of doing things. Install everything "from scratch" in every repo or project or folder as far down as possible
<clever>
i make a shell.nix for each project, that pulls in the right libraries for that project
<zeta_0>
i just use ghc in home.nix for my own projects, and i plan on using nix-shell for work, maybe this is not the best approach?
<zeta_0>
i like the declarative installation in home.nix
nornagon has joined #nixos
<clever>
zeta_0: 2 months later, when your on a different machine (or your home.nix has evolved), you wont be able to build that project anymore
<zeta_0>
i did run a nix-shell command, so that's why `brick` is not visible?
<clever>
zeta_0: what args did you give to nix-shell?
<clever>
zeta_0: what does `which ghc` return inside that shell?
<clever>
machotool also includes an example of how to override libraries
<clever>
in this case, the macho library was so ancient, it didnt compile with modern ghc, and was missing a couple features
<zeta_0>
the command threw an error: `user error (Terminal does not define required capability "cup")` , i am using eshell in emacs, so maybe eshell can't do things like this?
<clever>
zeta_0: try testing with a plain nix-shell first
<clever>
CMCDragonkai: only way to make it shorter, is to put that string into a .nix file, and run `nix-shell thatfile.nix`
<CMCDragonkai>
hmmm i see
<clever>
CMCDragonkai: if you name it shell.nix, then you can just `nix-shell`
<zeta_0>
in emacs i use haskell-mode integrated with haskell interactive, so i just type: C- c l and the file gets loaded in ghci with autocompletion etc
mla has joined #nixos
<CMCDragonkai>
I was wondering that given that we submit Python packages to nixpkgs, how does it know whether a package is python 3-only compatible
<clever>
zeta_0: about the only thing my workflow is missing (vim + ghcid) is context aware auto-completion
<fresheyeball>
I added `ninja` to nativeBuildInputs
<fresheyeball>
maybe that will do it
<clever>
zeta_0: but youcompleteme in vim will auto-complete any "word" present in any open buffer, and i usually have those open for cross-referencing, or they are already used nearby
<fresheyeball>
clever: have you tried HIE with Coc?
<clever>
fresheyeball: nope
mla has left #nixos [#nixos]
<fresheyeball>
clever: it's nice. Full LSP stuff in nvim
<fresheyeball>
autocomplete, error reporting, documentation, the works!
<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/Je9g3
fresheyeball has quit [Ping timeout: 265 seconds]
<jared-w>
maybe it's the laptop? I'm using a dell xps 9350 when I've used HIE. It could also be that I just find it super slow compared to the language server experience typescript has
<jared-w>
which, admittedly, isn't an entirely fair comparison
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/Je9g0
thc202 has quit [Ping timeout: 250 seconds]
<zeta_0>
jared-w: what text-editor/ide do you use?
Soo_Slow has quit [Quit: Soo_Slow]
mla has quit [Remote host closed the connection]
<zeta_0>
jared-w: i use emacs, after fixing a few problems, hie runs fast
<EdLin>
I'm a bit of a nixos newbie, so I might be doing something wrong?
<simpson>
EdLin: I think that you'll have to specify user passwords in the configuration.nix that you use when building the VM.
<EdLin>
oh.
<EdLin>
also, on a similar note, how do I delete old VMs?
<EdLin>
rm seems to tell me it's on a ro file system
lambda-11235 has joined #nixos
<simpson>
You could set users.users.myCoolUser.password, or set .hashedPassword. See https://nixos.org/nixos/manual/options.html for details on how to compute password hashes.
<clever>
EdLin: `rm result` not `rm result/`
<simpson>
I think that the VMs are in the Nix store, so you'd have to remove your `result` symlink as clever says, and then collect garbage from the Nix store (nix-collect-garbage, read the man page first!)
<jared-w>
zeta_0: I've done that before, but I like neovim better. It just works better for how I like to use a text editor
fusion809 has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @grahamc pushed 2 commits to release-19.09: https://git.io/Je9gD
fresheyeball has joined #nixos
<zeta_0>
jared-w: that's cool, everyone has their own preferences, i like emacs because you can use lisp to hack it making it very flexible, and its part of the gnu foundation, which is great
<fresheyeball>
clever: did that failure snippet look like just adding `ninja` would work to you?
<fresheyeball>
Failure cycle on this is like an hour
<fresheyeball>
:(
<zeta_0>
jared-w: there are haskell text-editors, but right now they are not very good
scassy has joined #nixos
erasmas has joined #nixos
waleee-cl has quit [Quit: Connection closed for inactivity]
<clever>
fresheyeball: the postUnpack was for fuzen
<fresheyeball>
oh
mwdev has joined #nixos
<fuzen>
yep, which doesn't seem like that fixed the problem
<fuzen>
>.<
<fuzen>
I guess mavenix doesn't work with that
<mwdev>
heyo- got a n00b question that can hopefully be resolved shortly: how would I go about utilizing a specific commit from the 19.09 branch of nixos-channels system-wide on my nixos laptop?
boxscape has quit [Remote host closed the connection]
<fuzen>
I'm thinking at this point, it would bbe simpler for me to just run cryptomator's gui through a docker container than to build it
lord| has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @jonringer pushed commit from @r-ryantm to master « python38Packages.pikepdf: 1.6.5 -> 1.8.0 »: https://git.io/Je92k
<CMCDragonkai>
Is there a way to add a python package to an overlay such that it becomes available to all intepreters, as if you had edited the top-level/python-packages.nix?
<buckley310>
mwdev, so theres a change in the repo that hasnt hit the mirrors yet?
<{^_^}>
[nixpkgs] @jonringer pushed commit from @r-ryantm to master « python38Packages.simpleeval: 0.9.8 -> 0.9.10 »: https://git.io/Je92I
sudoforge has joined #nixos
<mwdev>
buckley310, not quite- I have a system76 darp6 and I think they need to update some stuff in their firmware for kernel 5.4 to work well with their driver/firmware
lord| has joined #nixos
<mwdev>
to that end, I need to pin my system to kernel 5.3, which works
mla has quit [Remote host closed the connection]
<fuzen>
mwdev: did you check with fwupd (I think system76 uses this) for firmware updates?
<buckley310>
on unstable you can just add `boot.kernelPackages = pkgs.linuxPackages_5_3;` but im not sure on 1909, let me check what options they have
<{^_^}>
[nixpkgs] @jonringer pushed commit from @r-ryantm to master « python38Packages.limnoria: 2019.09.08 -> 2019.11.09 »: https://git.io/Je92m
<mwdev>
fuzen, not yet- I've been using stites' repo which uses fetchFromGithub <- I updated the shas to be recent as well
<mwdev>
buckley310, fair point. I'm trying to stick to `stable` as much as I can- haven't switched to unstable yet, but I'm willing to if no other options exist
<fuzen>
looks like they are using LVFS and system76-firmware, guess fwupd wouldnt update everything
<buckley310>
download the linux-5.3.nix file from unstable and in your config do something like `boot.kernelPackages = (callPackage /path/to/linux-5.3.nix {})
<CMCDragonkai>
in other words, can we use `isPy3k` to disable a package that doesn't work in Python 2? Like ` disabled = !isPy3k;`
<simpson>
CMCDragonkai: I'm not sure whether that's really a rewording of your first question, but yes to your second question. You'll find that sort of pattern already in nixpkgs, I think.
<{^_^}>
[nixpkgs] @grahamc pushed 2 commits to release-19.03: https://git.io/Je926
<CMCDragonkai>
but i still haven't figured out how to overlay the python packages such that it becomes available for all interpreters
<clever>
hexa-: if you want remote unlock, then your /boot and /boot/efi must be plaintext
<hexa->
yep
<clever>
CMCDragonkai: i'm not sure there is a simple option for that
<hexa->
thanks for being so helpful … and concise :)
jluttine has joined #nixos
<clever>
hexa-: and one thing i always worry about, somebody can just modify the initrd to phone-home with the pw, and then wait for you to ssh in and unlock it
<zeta_0>
so what does ./. mean in nix?
<clever>
zeta_0: its a path, pointing to whatever directory the nix file is in
<hexa->
clever: true :3
<zeta_0>
clever: alright, but that is a strange symbol
<o1lo01ol1o>
How can I understand what paths are available during a given phase for `mkderivation`? I'm trying to add a custom "installPhase" command but I don't know what the default "buildPhase" does with the output.
<clever>
zeta_0: paths start with either ./ (relative path) or / (absolute)
<clever>
zeta_0: and must end with something not a /
<zeta_0>
clever: what does relative mean in this case(does it depend on what directory you are currently in), the word relative has always confused me, especially in physics
<clever>
zeta_0: relative to the file that contains the path
<mwdev>
buckley310, so I tried that, but no dice. Seems like it might work, but I'm getting `value is a set while a list was expected, at /nix/store/wrjvn1wysg1y98pr53priyp79di20h7s-nixos-19.09.1548.3a1861fcabc/nixos/pkgs/os-specific/linux/kernel/generic.nix:91:9`
<clever>
o1lo01ol1o: you can always start with `pwd ; ls -lh` in the installPhase, to figure out whats available and where you are
<clever>
mwdev: which nixos option did you set to what value?
<clever>
hexa-: ive not tried mirrored efi yet, so i dont know how its setup properly
<clever>
mwdev: then line 4 is a single derivation with linux, and line 8 is expecting a set of all linux packages, and thats where it fails
<clever>
mwdev: you must put a linuxPackages set into kernelPackages, such as pkgs.linuxPackages_5_3
<hexa->
clever: ok, it obviously isn't since it resides on the root fs which is brfs
<clever>
hexa-: youll need to read the install-grub.pl file, and see what exactly its doing when mirrored boots is set, what efiSysMountpoint does, and what it runs grub-install with
<hexa->
ok, thanks for the filename :)
<clever>
hexa-: that filename is on line 37 of your pastebin, and its also in nixpkgs
<clever>
fresheyeball: pytorch is building without cuda support
<fresheyeball>
clever: oh good point!
<fresheyeball>
I am using `pytorchWithCuda`
<fresheyeball>
but have been building `pytorch` in my testing this whole time *facepalm*
<mwdev>
clever: is line 4 not what buckley310 had suggested with `boot.kernelPackages = (callPackage /path/to/linux-5.3.nix {})` (admittedly split across lines 4 and 8)?
<clever>
zeta_0: cabal2nix only works if you have a cabal file
<notgne2>
is there any known way of running LoL on NixOS
<clever>
mwdev: i would not expect that line to work, you have to pass the linux packakage thru pkgs.linuxPackagesFor first
<mwdev>
clever: sweet- makes sense.. looking into how to set that up now
<clever>
mwdev: examples are in all-packages.nix
<zeta_0>
clever: so i need to run `cabal init` first, i am just following those instructions from that link?
<clever>
zeta_0: you need a my_script.cabal file, and i think `cabal init` will make one
<clever>
fresheyeball: dang is this a huge build, lol
<clever>
i think an `exit 1` and `--keep-failed` in the installPhase would help more then a big computer
<zeta_0>
[zeta@nixos:/my_script]$ ls
<zeta_0>
CHANGELOG.md Setup.hs my-script.cabal
<zeta_0>
Main.hs default.nix my_script.hs
<gchristensen>
clever, fresheyeball: how about I get y'all a big machine to borrow?
ddellacosta has joined #nixos
<clever>
i just noticed, i'm barely 1/6th into the build, lol
<clever>
and the count isnt going up
<buckley310>
clever, mwdev, maybe this? :) im just sorta free-handing here `boot.kernelPackages = recurseIntoAttrs (linuxPackagesFor (callPackage ../os-specific/linux/kernel/linux-5.4.nix {}));`
<clever>
buckley310: recurseIntoAttrs is only needed to make nix-env -i search inside it
<zeta_0>
well i don't understand why some of the instructions are left out, i am guessing i am going to have to use cabal2nix to translate the my_script.cabal to my_script.nix , and do i need to be super user, it keeps telling me permission denied?
<buckley310>
ah
<clever>
zeta_0: callCabal2nix will handle cabal2nix for you
<clever>
buckley310: the rest looks like it should work
<clever>
zeta_0: what is inside `default.nix` ?
<zeta_0>
clever: do i need to place callCabal2nix in default.nix?
<gchristensen>
clever: I'll get y'all a bigger machine to experiment on :)
<zeta_0>
gchristensen: thank you (and the others) for setting up those nix-pills they really help, i am currently on chapter 11, garbage collection
<o1lo01ol1o>
clever: bug in the expr overwriting the ssl option.
<clever>
zeta_0: developPackage is just a giant fancy wrapper around callCabal2nix
<clever>
zeta_0: so your already using callCabal2nix
<clever>
zeta_0: found your problem, the directory is my_script, but the cabal file is my-script.cabal
<clever>
zeta_0: they must match (the _ vs -) or it wont work
levdub has joined #nixos
<clever>
zeta_0: you can also pass `name = "my-script";` to `developPackage`, then the directory name wont matter (the name= must match the cabal file then)
<zeta_0>
clever: oh, when i ran cabal init , that's what it named the .cabal file
<zeta_0>
clever: i can just rename the .cabal file right?
<clever>
zeta_0: you have 3 choices
<clever>
1: rename the file to my_script.cabal
<clever>
2: rename the directory to my-script
<clever>
3: set name = "my-script"; in developPackage
<clever>
4: rename both the cabal file and set name=!
<clever>
5: rename both the directory and cabal file!
<clever>
5 choices*
<zeta_0>
i guess i will just rename the directory since cabal init generates the .cabal with a - by default
<zeta_0>
clever: choice overload
<mwdev>
clever: you may have saved my bacon- I updated the gist to reflect the changes, but now I think I misunderstand the overlay setup. The current build complains, "attribute 'system76-dkms' missing, at /etc/nixos/system76-nixos/default.nix:23:32" <- I understand the error, but not the path to resolution =/
<{^_^}>
[nixpkgs] @jonringer pushed commit from @0mp to master « Fix a typo »: https://git.io/Je9a9
<clever>
mwdev: linuxPackagesFor (callPackage ...)` will ignore the overlay
<clever>
mwdev: so you may want to do (linuxPackagesFor (callPackage ...)).extend, to apply an overlay to it
<clever>
mwdev: and then it gets simpler to use nixpkgs overlays, to do it in steps, one to add the linux_5_4 and linuxPackages_5_4, then a 2nd to modify linuxPackages_5_4
<{^_^}>
[nixpkgs] @jonringer pushed commit from @dywedir to master « oil: 0.7.pre6 -> 0.7.pre9 »: https://git.io/Je9ad
<zeta_0>
it throws an error when i try to rename the directory, how strange
scassy has quit [Ping timeout: 276 seconds]
<clever>
zeta_0: what error does it throw?
andreas303 has quit [Remote host closed the connection]
<clever>
[1851/2983] Building CXX object caffe2/CMakeFiles/torch_cpu.dir/__/aten/src/ATen/native/quantized/cpu/kernels/QuantizedOpKernels.cpp.AVX.cpp.o
<mwdev>
clever: ho- lee- ba nay nays- it built XD I changed the last line to `boot.extraModulePackages = [ (pkgs.linuxPackages_5_3).system76-dkms ];` and all was copacetic. That said, I'll be looking into your suggestion as well
<clever>
mwdev: that might be mixing a 5.3 module with a 5.4 kernel, which will fail to load
<clever>
mwdev: you should only ever use config.boot.kernelPackages within extraModulePackages
<zeta_0>
clever: a weird emacs error, anyways i changed: `my-script.cabal` to `my_script.cabal`
andreas303 has joined #nixos
<zeta_0>
clever: and it works, thanks
<zeta_0>
clever++
<{^_^}>
clever's karma got increased to 275
<mwdev>
clever: I see what you mean- sagely advice
<fresheyeball>
gchristensen: do it
<fresheyeball>
I have to stay up until this builds basically
<clever>
fresheyeball: the race is on, can gchristensen's machine beat mine, even with a late start!
<gchristensen>
I'm not going to actually try it, I'm just going to give you access
<fresheyeball>
you can have access to mine personal box
<fresheyeball>
but I don't think it would help much
szicari has joined #nixos
<gchristensen>
fresheyeball, clever: root@147.75.39.61 (caveat: this machine can disappear at any moment, but probably wont. make sure to build with --cores 56, as the default system config probably has cores = 1.)
spacefrogg has quit [Quit: Gone.]
aw has quit [Quit: Quitting.]
<clever>
fresheyeball: i'm in a screen session, `screen -x`
<fresheyeball>
clever: watching you go, this is really cool
<clever>
fresheyeball: `top` is in screen 1, i see 48 cores!
<clever>
the core list doesnt even fit on a split screen, lol
<gchristensen>
oops right this one is 48 cores to 56 (they'are nice cores though, epyc 7401p)
<fresheyeball>
clever: how do I switch screens?
<clever>
fresheyeball: ctrl+a 1, to go to screen 1
<DigitalKiwi>
is this a race for the cash
<fuzen>
clever: how did you extract appimages? The manual says I should be using "./<crpytomator image> --appimage-mount" and the application launches instead and the mount steps for Type1 fail.
<clever>
fuzen: there is also an appimage-run in nixpkgs
<clever>
fresheyeball: this will re-build it, with cuda support enabled
<fresheyeball>
I think pytorchWithCuda does that too
LLED2_2 has joined #nixos
<{^_^}>
[nixpkgs] @andersk opened pull request #75477 → python: testresources: Move pbr to propagatedBuildInputs → https://git.io/Je9Va
<DigitalKiwi>
oh fresheyeball I just have to cut those trans nixos stickers which turned out fine (some other ones I did are not good) and paint the haskell ones still...
<DigitalKiwi>
I had adhesive residue ruin a bunch of the nixos ones I did :<
zebrag has quit [Read error: Connection reset by peer]
<fuzen>
I'm starting to wonder why cryptomator was made in java, if it cant be packaged in a portable way like .deb anymore
zebrag has joined #nixos
LLED2_3 has quit [Ping timeout: 265 seconds]
<fresheyeball>
watching this machine work is just nuts
<gchristensen>
when I handed it over I was disappointed: "aww, it only has 64G of ram..." and that was the moment I realized I am deeply, *deeply* spoiled
<clever>
gchristensen: lol, i only have 32gig of ram each in my laptop&desktop
<clever>
samueldr: ive gotten it to work impurely, using rustup i think it was, in a nix-shell
<samueldr>
and replacing the import of nixpkgs with pkgsCross for aarch64 results in a failure to build the carnix'd thing
<clever>
samueldr: i basically used nix-shell to get a linux->windows cross-compiler, then used the standard rust tools to get a windows crate, and off it went
<samueldr>
note: I'm not knowledgeable in rust so it doesn't help in tracking down issues
<samueldr>
error[E0463]: can't find crate for `std`
<clever>
samueldr: yeah, std was it, i used some impure cmd to download a build of std for windows and it just landed in $HOME
<clever>
ive no clue how to do it purely
<samueldr>
wondering if it's carnix that doesn't do something or what
<samueldr>
since the compiler itself built just fine
<fuzen>
I think carnix targets a no-std environment
<fuzen>
Can't seem to be able to build cryptomator due to missing hashes on bintray
mexisme_ has quit [Ping timeout: 268 seconds]
<clever>
fresheyeball: turning on cuda support seems to slow it down heavily at the testing phase
<clever>
fresheyeball: i can confirm that pytorch is not a split output derivation
<clever>
fresheyeball: and thats what causes it to install the missing header, so we must read that cmake file
<clever>
nothing there looks optional, but the file itself may be optionally pulled in elsewhere
<clever>
fresheyeball: and thats where things get complicated, i dont see any obvious thing loading c10/cuda/CMakeLists.txt
<clever>
fresheyeball: it almost looks like your meant to build c10 seperately, in its own package
<fresheyeball>
clever: this is nuts
<clever>
fresheyeball: Caffe2: Cannot find cuDNN library. Turning the option off
<clever>
fresheyeball: do you know which package thats in?
<srid>
zeta_0: oh hey, good to know that my tutorial helped you to some extent. :) sorry it is not as comprehensive as you'd have liked. i wrote it assuming that someone already had a valid cabal project.
<{^_^}>
[nixpkgs] @worldofpeace merged pull request #74295 → Remove lots of pygtk using software → https://git.io/JeXIP
erasmas has joined #nixos
<{^_^}>
[nixpkgs] @worldofpeace pushed 22 commits to master: https://git.io/Je9VN
<samueldr>
(sorry to be so memetic)
<samueldr>
referring to the box about adding compilers to your systemPackages
<clever>
fresheyeball: pkgs.cudnn duh!
<clever>
CMake Error at CMakeLists.txt:41 (target_compile_options): Cannot specify compile options for target "c10_cuda" which is not built by
<clever>
fresheyeball: aha, c10/CMakeLists.txt has an `add_subdirectory(cuda)` which loads the file in question
drakonis has joined #nixos
<rotaerk>
hmm, I'm running xmonad in nixos with a high res monitor; I'd like to increase the size of my mouse cursor. is there some standard nix configuration file way of doing this, or do I just need to run xsetroot somewhere
<clever>
fresheyeball: which is being loaded by the root...
<rotaerk>
and if xsetroot, I'm not sure where I should get the cursor file(s) from
<clever>
fresheyeball: despite building with cuda enabled, the logs say cuda is disabled...
<rotaerk>
also, I need to get the mouse moving farther across the screen ... it takes several swipes of the touchpad to get it all the way across the screen
<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/Je9wL
<clever>
fresheyeball, gchristensen: kind of stuck now, not sure what to check next
<rotaerk>
though the documentation is unclear on what the format of the speed should be ... guess I'll dig into nixpkgs to figure out where that's going
zeta_0 has quit [Quit: rcirc on GNU Emacs 26.3]
<clever>
gchristensen: shal we kill it or leave it up?
<clever>
fresheyeball: my current idea, is to edit it to fail after the configure script, skip the huge build, and then iterate until the configure phase says cuda is actually enabled
drakonis has quit [Read error: Connection reset by peer]
<gchristensen>
fresheyeball, clever will yall try tomorrow?
<clever>
gchristensen: the plan i have now doesnt require an insane core count
<clever>
gchristensen: the problem appears to be before the computation heavy phase
<clever>
gchristensen: and ive put all changes of importance into the above gist, so the machine doesnt matter much
<gchristensen>
I'm inclined to leave it up
<gchristensen>
let me know tomorrow if you don't need it
<samueldr>
rotaerk: humour me, and if you can, use 4.19 and see if the issue happens there too
knupfer has joined #nixos
<rotaerk>
using 5.3.8. and I can't use a much earlier version because it doesn't support my hardware
<samueldr>
aw
<samueldr>
there might be something weird (and not nixos specific) with 5.3 (maybe 5.4 too?) that makes the computer unresponsive for small amount of time
<samueldr>
at first I thought it was the mouse, but it's not, and it happened to a friend of mine on archlinux, and happened to another nixos user that confirmed that 4.19 didn't have the issue for them either :/
<samueldr>
rotaerk: right, that's why "if you can", as I saw you said it was a laptop and I assumed a recent one :)
<rotaerk>
it's not just the mouse; the touchpad and touchscreen sometimes stop responding and *sometimes* recover, but I also sometimes see the entire system freeze, unresponsive to the keyboard as well
<samueldr>
yeah, it's not just the mouse either for me
<samueldr>
it just looked like that
cfricke has joined #nixos
<samueldr>
though your issue sound different as you say "sometimes recover"
<rotaerk>
I'm just not sure how to begin to diagnose this
<rotaerk>
well I think the mouse issue and the entire system freezing are two different problems
<samueldr>
here it always recovers, and not for long, just infuriatingly annoying hiccups
cfricke has quit [Client Quit]
<samueldr>
oh, great, compound issues :(
<rotaerk>
though yeah, the mouse issue alone, it didn't always recover
scassy has joined #nixos
felixfoertsch has quit [Quit: ZNC 1.7.3 - https://znc.in]
felixfoertsch has joined #nixos
<keithy[m]>
I am getting similar issues come to mention it, imac 2008, running elementary Hera (kernal 5.1)
<samueldr>
I haven't tried anything in-between 4.19 and 5.3 so it's possible the issue is not strictly new
<samueldr>
(and it's not exactly reproducible, so bisecting would end up during days!)
<rotaerk>
I also get an issue where occasionally when I poweroff or reboot, the screen goes black but the system is still running (and in the case of reboot, it never reaches system startup)
scassy is now known as scasc
<rotaerk>
have to hold the power button
szicari has quit [Remote host closed the connection]
szicari has joined #nixos
knupfer has quit [Ping timeout: 245 seconds]
<scasc>
@keithy[m]: elementarz *Hera*'s out?
mwdev has quit [Ping timeout: 260 seconds]
<scasc>
I checked like last week or the one before and it was still 5.0
brightflame has joined #nixos
<keithy[m]>
yes, and 5.0 updates!
<keithy[m]>
(but tidy up /boot before it accidentally fills up)
<brightflame>
i'm slowly getting used to this strange nixos-y world. why is it that when i install something via nix-env i have to name it like `nixos.lxappearance-gtk3`, but when i run it via nix run i name it like `nixpkgs.lxappearance-gtk3`?
<keithy[m]>
ok... one solution I found (bit random) The solution I found was to disable power management of the USB controller that the mouse is connected to.
<rotaerk>
how do you manage power management settings
<scasc>
@keithy[m], does it elementaryOS work nicely with nix?
<keithy[m]>
no idea - I jst put it on an old imac in the guest bedroom for visitors to use
<keithy[m]>
it gives them a browser, youtube, and now spotify via sideloaded flatpak
<scasc>
Yeah, I was also looking into it for an older machine, not my workhorse, for my children; the target is very different (and I don't have a necessity for nix there).
<keithy[m]>
rot sorry that was just clutching at staws
<rotaerk>
I'm trying to figure out .Xresources ... the Arch documentation for it gives the general format but...
<rotaerk>
how do you know what specific name.class.resource things are valid ... I don't see documentation of the actual values you can use
<rotaerk>
except stray references to one or two here and there
erasmas has joined #nixos
CMCDragonkai has joined #nixos
<rotaerk>
for instance, you're supposed to be able to set Xcursor.theme and Xcursor.size, but they aren't working for me ... so where do I find out more details
<rotaerk>
such as what names I can use for the theme ... if I'm referencing xcursor theme folders, where it is looking for them
<rotaerk>
or what sizes are valid for a given theme
<CMCDragonkai>
one point of confusion when using overlay to override, is that if you orride python3, it doesn't override python36, python37, python38... etc etc
erasmas has quit [Ping timeout: 276 seconds]
lambda-11235 has quit [Read error: Connection reset by peer]
ar has joined #nixos
lambda-11235 has joined #nixos
bahamas has joined #nixos
jgt2 has joined #nixos
palo1 has joined #nixos
palo has quit [Ping timeout: 246 seconds]
palo1 is now known as palo
peanutbutter144 has quit [Quit: No Ping reply in 180 seconds.]
peanutbutter144 has joined #nixos
<{^_^}>
[nixpkgs] @tbenst opened pull request #75482 → pythonPackages.caiman: init at 1.6.4 → https://git.io/Je9rr
<kolbycrouch>
Does anyone know how to get toolchain binaries without a tuple prefix in a pkgsCross stdenv? A lot of packages are failing to respect $AR, $LD, $CC etc..
<yorick>
kolbycrouch: best to patch the packages I think
<yorick>
kolbycrouch: it's not like the configure script will work like that
<kolbycrouch>
yorick: I'm just wondering were at. I tried using ln in pre/postFixup. Didn't work. Not sure if it's an upstream default, a binutils or a cc-wrapper issue.
<yorick>
kolbycrouch: which packages?
<yorick>
kolbycrouch: packages generally have to explicitly support cross-compilation, since they often use build outputs as part of the compilation process
<kolbycrouch>
yorick: argp-standalone is one. it tries to call ar from path without a cross-compile prefix. I think having a symlink from i686-unknown-linux-musl-ar should just work.
<yorick>
kolbycrouch: argp-standalone is an autoconf package, that should support AR and such
<kolbycrouch>
yorick:I thought I would be able to avoid pkgsCross with i686-musl on my x86_64 seeing as it can run natively, but when using an expression that combines pkgsMusl and pkgsi686Linux I get an error about unsupported stdenv.
<yorick>
hmm
<kolbycrouch>
yorick: I know it is uses autoconf. Still it fails with command not found for ar. it's respecting CC obviously because the command line output shows the full path for it during compilation, but ar is getting called as just ar.
smatting has quit [Ping timeout: 240 seconds]
<ar>
yup. almost noone calls me by my name ;)
<kolbycrouch>
ar: Well this makefile is :D
mmercier has joined #nixos
<kolbycrouch>
I recently built nix and its dependencies with pkgsCross.rpi and everything went smoothly. Don't think argp-standalone was in the closure, but still.
<{^_^}>
[nixpkgs] @michaeladler opened pull request #75487 → nixos/systemd: use global proxy config in systemd-importd → https://git.io/Je9ik
cfricke has quit [Quit: WeeChat 2.6]
Chiliparrot has joined #nixos
<Unode>
hi all, I've upgraded to 19.09 yesterday. All going well but one service (systemd-timesyncd) is failing right at start. Looking at the service file the ExecStart command has two exclamation marks as a prefix (ExecStart=!!/path/to/timesyncd). I've never encountered this syntax. Does anyone know what it means or if this is the reason why the service is failing?
<immae>
Unode: man systemd.service :)
<immae>
(to partially answer your question, !! at the beginning is a valid syntax for systemd services, but it has some consequences that are described in the manpage)
veske has joined #nixos
grumble has quit [Quit: SOMEBODY ONCE TOLD ME]
veske has quit [Quit: This computer has gone to sleep]
<{^_^}>
#31540 (by MrSorcus, 2 years ago, open): Failed to start Network Time Synchronization.
<DigitalKiwi>
that's a lot older than I thought it would be hmm
<DigitalKiwi>
I feel like there were release notes about the timesyncd problems
veske has joined #nixos
<makefu>
DigitalKiwi: with every system i update from 19.03 to 19.09 i trip over this
sudoforge has quit [Ping timeout: 268 seconds]
magthe has joined #nixos
<{^_^}>
[nixpkgs] @mat8913 opened pull request #75488 → internetarchive: Provide setuptools at runtime → https://git.io/Je9iF
<DigitalKiwi>
Because of the systemd upgrade, systemd-timesyncd will no longer work if system.stateVersion is not set correctly. When upgrading from NixOS 19.03, please make sure that system.stateVersion is set to "19.03", or lower if the installation dates back to an earlier version of NixOS.
<{^_^}>
[nixpkgs] @c0bw3b pushed commit from @wamserma to master « samba: 4.10.10 -> 4.10.11 (#75456) »: https://git.io/Je9Pz
__monty__ has joined #nixos
magneticduck has quit [Quit: Leaving]
veske has quit [Quit: This computer has gone to sleep]
erasmas has joined #nixos
ng0 has joined #nixos
alexherbo28 has joined #nixos
Diagon has quit [Quit: Leaving]
erasmas has quit [Ping timeout: 268 seconds]
sb0 has joined #nixos
<pistache>
rotaerk: for your mouse issue, check for USB resets in dmesg
iyzsong has joined #nixos
fusion809 has joined #nixos
brightflame has quit [Ping timeout: 245 seconds]
civodul has joined #nixos
<{^_^}>
[nixpkgs] @hedning opened pull request #75489 → nixos/gdm: only set session on auto login → https://git.io/Je9Xz
logzet has quit [Ping timeout: 250 seconds]
silver has joined #nixos
<Taneb>
My Hydra server's logs are annoyingly full of "Hydra::Plugin::GitlabSatus=HASH(0x3a27f48)->buildFinihsed: Can't call method \"value\" on an underfined value at hydra/libexec/hydra/lib/Hydra/Plugin/GitlabStatus.pm line 57" :(
<jtojnar>
hedning I propose using initialSession for the first time session
<cfricke>
Hei, question regarding recent unstable changes. My window manager (xmonad) doesn't seem to honor my user path anymore. I assume prior to the change my login shell (using .zshenv) was honored. Can't find anything in nixpkgs/issues. Anyone got a clue?
domogled has joined #nixos
domogled has quit [Ping timeout: 246 seconds]
levdub has joined #nixos
<{^_^}>
[nixpkgs] @filalex77 opened pull request #75493 → db-backup: init at 1.0 → https://git.io/Je91G
Chiliparrot has quit [Ping timeout: 245 seconds]
hmpffff has joined #nixos
ckauhaus has joined #nixos
levdub has quit [Ping timeout: 240 seconds]
<infinisil>
cfricke: Not sure what you mean
Chiliparrot has joined #nixos
Chiliparrot has quit [Client Quit]
sb0 has quit [Quit: Leaving]
<{^_^}>
[nixpkgs] @Elyhaka opened pull request #75494 → pop-gtk-theme: init at 3.1.2 → https://git.io/Je91o
<cfricke>
Well, I have executables in my user (home) PATH, which I define define in .zshenv, which is read by the default login shell. So, I have in my /etc/nixos/configuration.nix a sessionCommand: "${pkgs.systemd}/bin/systemctl --user import-environment PATH &". Before I updated today, my ~/bin was included. Now it's only the default nix defined path.
<evanjs>
In this case, specifically, doom-modeline seems to get enabled and all the icons work fine, but when connecting via emacsclient, the icons are not rendered
<angerman>
yorick: the flamegraph.pl thing keeps failing me ... eats too much memory.
<yorick>
angerman: yeah, I run it on a 64GB server
<angerman>
yorick: but the flamegraph would kinda be waht you see on the left in the first screenshot. I'd like to generate that in vscode directly. Without flamegraph.pl. However nice that is.
<yorick>
angerman: so this thing is O(n^2) right? :D
<angerman>
the % annotations are not correctly calculated for anything besides toplevel, so they are garbage here. However the fequency counters give the number of times that line showed up in the trace. That is slightly misleading as we end up aggregating the same line multiple times if there were multiple traces for the same line but different functions (I think).
marek_ has joined #nixos
bvdw has quit [Quit: bvdw]
turion has quit [Quit: Leaving.]
bvdw has joined #nixos
<Unode>
immae: thanks, I couldn't search for it on google. Always annoying when syntax isn't searchable
<immae>
Unode: yes I understand. As far as I saw so far, systemd manpages are rather complete, so I always start there (the only problem I have is finding which systemd manpage applies :p )
<makefu>
hey all, what would be the correct way to override the cacerts on nixos for jdk11? i managed to get it working by overriding cacerts for the package but how does it work in version11 which does not provide the cacerts overridable parameter?
<duairc>
Or you could just use terraform to set up the machines and then nixops to do your nixos configuration
dansho has quit [Remote host closed the connection]
ng0 has quit [Ping timeout: 260 seconds]
dansho has joined #nixos
<angerman>
gchristensen: yea, I wish there was some flame graphy thing for vscode. but i haven't found it yet. wiring the sqlite database up to it shouldn't be that hard.
cosimone has joined #nixos
<angerman>
gchristensen: I'm still conflicted if I'd rather have nix write the sqlite db directly instead of dumping that info to stdout.
<stammon>
I'm still looking for some help with npm and nix derivations. I have a npm run script, that I want to run when building my derivation. The run script needs several node modules available and I fail to get it working with node2nix. I also can't find too help on nix and npm run scripts online.
EdLin has joined #nixos
<EdLin>
I keep getting errors during nixos-rebuild switch with collisions associated with gstreamer files I think.
<EdLin>
is this something I should be concerned about?
jakobrs has quit [Ping timeout: 245 seconds]
<infinisil>
EdLin: Does it error because of those?
<ivan>
fresheyeball: are you interested in the very latest torchvision and compatible pytorch, or pytorch 1.3.1 and whatever torchvision works with that?
<EdLin>
anyone have ideas on how to fix these errors?
<gchristensen>
does `perf record` automatically start watching all the threads of a process?
polman has quit [Excess Flood]
polman has joined #nixos
<zeta_0>
i plan using haskell for some projects(right now just web development using the yesod framework and warp web server, as well as sql for the backend(i'm going to use the selda package to write haskell to generate sql code)), anyways, i want to correctly set up nix-shell to install and configure all the packages that i am going to be using, i would also like do the nix code in a very generic and abstract way so i can
<zeta_0>
reuse it for other projects, as well as to automate everything as much as possible, could someone help in guiding me through the steps?
<fresheyeball>
the goal is to have a nix-shell where I can use this project
<fresheyeball>
with cuda
teehemkay has joined #nixos
<noonien>
is there a way to get the current value of a configration i've setup in configuration.nix?
<zeta_0>
and is there a good link for `yesod development with nix`?
<infinisil>
,dnw alexherbo2
<{^_^}>
alexherbo2: Provide as much info as possible on problems. What specifically doesn't work? What command did you run? What's the error? This makes debugging easier and helps getting faster responses
<noonien>
i want to execute a command and tell me, for example, what value 'hardware.opengl.driSupport32Bit' has
<zeta_0>
maybe its buried somewhere in documentation and i can't find it
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<fresheyeball>
clever: let me know where we left off, it appears the big box is still around for us
<gchristensen>
I'm looking in to why hydra-queue-runner is reporting 400 Active jobs when (copying inputs + rnuning + copying outputs) is only about 100. I did a flame graph of hydra-queue-runner but it doesn't seem very interesting. any tips
<drakonis>
there's a strong intersection between haskell and nix
<infinisil>
alexherbo2: Now open the surf expression, can be done with `nix edit nixpkgs.surf -I nixpkgs=/path/to/your/nixpkgs`
sigmundv has quit [Ping timeout: 246 seconds]
<alexherbo2>
done
<infinisil>
alexherbo2: The current expression fetches it from a release url, we need to change that to use a git revision
<drakonis>
dont ping me on haskell development stuff though, i dont write haskell
orivej has joined #nixos
<infinisil>
alexherbo2: Namely you need to replace fetchurl with fetchgit
<infinisil>
alexherbo2: And fetchgit takes three necessary arguments: `url` for the git repository url, `rev` for the git revision to fetch, and `sha256` for the hash
<infinisil>
Can you fill those out?
polman has quit [Ping timeout: 250 seconds]
Lears has joined #nixos
<alexherbo2>
hm
<alexherbo2>
not sure how to get the sha
<infinisil>
,tofu is useful for this
<{^_^}>
is useful for this: 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. See: tofu-vim
<drakonis>
,tofu-vim
<{^_^}>
<esc>52i0<esc>
<drakonis>
hmm
<drakonis>
what.
<infinisil>
That's a vim package name I presume, not a command here :)
<drakonis>
is that a emoji code?
<drakonis>
i see
[Leary] has quit [Read error: No route to host]
<infinisil>
Oh!
<infinisil>
No it is a command here, didn't see that
<infinisil>
That's the keys you need to press in vim to insert 52 0's
<drakonis>
oh i see
boxscape has joined #nixos
<drakonis>
neat.
zupo has joined #nixos
<alexherbo2>
infinisil how can I test my local nixpkgs?
<zeta_0>
is there a way to use nix to automatically upload your projects to a repository(gitlab, github) using nix?
m4ts has quit [Ping timeout: 250 seconds]
<kcalvinalvnn>
zeta_0 a bash script might better suited for that purpose
<zeta_0>
to an external repository that is
<immae>
infinisil: would a patch for nix to make it accept "0" as a syntaxically-valid hash be accespted? That’s much easier to write than "0000000000000000000000000000000000000000000000000000" when we just want a fake hash :p
<zeta_0>
kcalvinalvnn: i was just wondering, i normally use git to manually do this
[Leary] has joined #nixos
levdub has joined #nixos
<infinisil>
immae: Hm, we could just extend fetch* functions to use 52 0's by default if no hash is specified
Lears has quit [Read error: No route to host]
<kcalvinalvnn>
zeta_0 unless you want to push it after a successful nix build, I don't see the benefits gained using nix
<immae>
ah that’s an even better idea
cosimone has quit [Quit: Quit.]
[Leary] has quit [Remote host closed the connection]
Lears has joined #nixos
Jackneill has quit [Remote host closed the connection]
levdub has quit [Client Quit]
Thra11 has joined #nixos
polman has joined #nixos
judson has joined #nixos
<Thra11>
What's the easiest way to get a shell in which I can't
<zeta_0>
kcalvinalvnn: that's fine, i was just curious
<Thra11>
*can run make menuconfig in linux sources
polman has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @FRidh pushed commit from @florianjacob to release-19.09 « matomo: 3.11 -> 3.13 »: https://git.io/Je9bi
polman has quit [Excess Flood]
fendor__ has joined #nixos
polman has joined #nixos
mexisme_ has joined #nixos
polman has quit [Read error: Connection reset by peer]
polman has joined #nixos
fendor has quit [Ping timeout: 245 seconds]
chloekek has joined #nixos
willghatch has joined #nixos
evils has quit [Quit: Lost terminal]
fendor has joined #nixos
fendor__ has quit [Ping timeout: 240 seconds]
<willghatch>
Hello. Usually on nixos I use systemPackages to define which packages I have. Recently I installed just the nix package manager on a non-nixos system. Is there a way to specify a list of packages declaratively when just using the package manager like systemPackages instead of using nix-env?
smatting has quit [Ping timeout: 250 seconds]
evils has joined #nixos
cosimone has joined #nixos
polman has quit [Read error: Connection reset by peer]
polman has quit [Read error: Connection reset by peer]
polman has joined #nixos
t58 has quit [Quit: Leaving]
mexisme_ has quit [Ping timeout: 268 seconds]
jgt2 has joined #nixos
<tilpner>
,declarative willghatch
<{^_^}>
willghatch: There are multiple ways of managing declarative profiles. 1) Attrset, compatible with imperative use of nix-env https://git.io/fAQHW ; 2) buildEnv, providing more control over the paths that are linked into the profile https://git.io/fp0aU ; 3) home-manager, providing nixos-like config for your ~ https://github.com/rycee/home-manager
Laarlf has joined #nixos
<{^_^}>
[nixpkgs] @jollheef opened pull request #75524 → codeql: init at 2.0.0 → https://git.io/Je9NW
polman has quit [Read error: Connection reset by peer]
ThatDocsLady_ has quit [Ping timeout: 252 seconds]
polman has joined #nixos
jmarin217 has joined #nixos
jmarin217 has left #nixos [#nixos]
marek has quit [Ping timeout: 252 seconds]
inkbottle has joined #nixos
marek has joined #nixos
werner291 has joined #nixos
<worldofpeace>
We have surprisingly little information about software licensing in the nixpkgs manual https://nixos.org/nixpkgs/manual/#sec-meta-license, especially when packaging something. I fear expressions get merged that are violative, we could also be requiring reviewers to nudge upstream to make some informed changes.
<worldofpeace>
Other distros have a better wealth of this information referenced or to refer to. And if you even look at their review process it holds a greater importance.
<Laarlf>
anyone runs pagekit on nixos? i cant get the site to show up properly. i use nginx and phpfpm but all it shows is "Pagekit" after throwing me onto site.com/installer. the relevant pieces of my config are here: https://paste.debian.net/1120698/ some help would be appreciated
ng0 has quit [Ping timeout: 260 seconds]
<qyliss>
worldofpeace: yeah I somewhat often find things not marked as unfree that should be
mexisme_ has joined #nixos
polman has quit [Ping timeout: 268 seconds]
zupo has joined #nixos
polman has joined #nixos
mexisme has joined #nixos
inkbottle has quit [Ping timeout: 268 seconds]
mexisme_ has quit [Ping timeout: 268 seconds]
<Laarlf>
if i can somehow provide helpful logs, please tell me which logs and where i can find them, thanks
polman has quit [Read error: Connection reset by peer]
polman has joined #nixos
ng0 has joined #nixos
cosimone has quit [Quit: Quit.]
cfricke has joined #nixos
<acowley>
Can I get the number of build threads I should use within a nix derivation?
<{^_^}>
[nixpkgs] @jonringer opened pull request #75525 → [Draft] teams: init at 1.2.00.32451 → https://git.io/Je9Am
<acowley>
gchristensen: I should have found that. I had expected it to be a nix value like `config`.
siiky has quit [Quit: leaving]
andreas303 has joined #nixos
<{^_^}>
[nixpkgs] @robertodr opened pull request #75526 → Add GNU threading and Fortran libraries → https://git.io/Je9AZ
polman has quit [Ping timeout: 268 seconds]
siiky has joined #nixos
polman has joined #nixos
<acowley>
Are there any docs to speak of regarding patchShebangs? I have this package that has dozens of bash scripts involved in its build and at runtime. I currently call patchShebangs on every file in the source tree just to be sure, but it takes quite a while and feels crude.
<{^_^}>
[nixpkgs] @reset closed pull request #68709 → dotnet-sdk: add darwin support, set myself as maintainer → https://git.io/Je9AW
lovesegfault has joined #nixos
goolord has joined #nixos
wildtrees has joined #nixos
polman has quit [Read error: Connection reset by peer]
polman has joined #nixos
wrl has quit [Read error: Connection reset by peer]
<goolord>
for some reason i can't open uris with custom protocols with xdg-open even though there's a default application for the associated mime type https://pastebin.com/xjhUViLQ
<clever>
let foo = import ./. {}; foo // { using = stdenv.mkDerivation { name = "foo"; buildInputs = [ foo ]; }
<clever>
then nix-shell will let you build gridsync, and `nix-shell -A using` lets you use a copy nix built purely
<clever>
but that depends on if you want to develop it more
<exarkun>
cool, I'll mess around with that
polman has quit [Ping timeout: 268 seconds]
<{^_^}>
[nixpkgs] @offlinehacker opened pull request #75527 → hasura: init at 1.0.0-rc.1 → https://git.io/Je9x6
mzabani has joined #nixos
nobbo has joined #nixos
murmeli is now known as rural_fox
<mzabani>
Good afternoon, everyone. I'd like to delete profiles in the future (not in the past) and collect garbage to save space, but I don't know how to do that.. can anyone help me out on this one?
<clever>
mzabani: which generation do you want to delete?
<mzabani>
all but the current one. It seems the current one is number 44 according to `nix-env --list-generations`, but the others don't appear in the output of that
<clever>
mzabani: that would imply that they have already been deleted from the profile
<clever>
mzabani: do you have a large thing in /nix/store you know the path of, that you want deleted?
rural_fox is now known as urban_mumeli
<mzabani>
hmmm that's strange. Why do they still appear as boot options?
urban_mumeli is now known as murmeli
<clever>
mzabani: the bootloader config is not updated until you re-run switch or boot
<clever>
mzabani: `/run/current-system/bin/switch-to-configuration boot` will make whatever your currently running the default boot, and update the list to reflect what --list-generations prints
murmeli is now known as urban_murmeli
<mzabani>
ok, that's very new to me..
urban_murmeli is now known as murmeli
<clever>
mzabani: nixos-rebuild switch would also update the bootloader, but that also builds the current configuration.nix
<clever>
which may be something you didnt want done
<mzabani>
yes.. 19.09 is still broken for me, but I already updated the channel.. I could just go back to 18.09 - which works - and run switch again, I guess
kolaente_ has joined #nixos
<mzabani>
but I was wondering how to do this without running switch again
<clever>
mzabani: there is also `nix-channel --rollback`, since channels have their own generations
<mzabani>
ah, really? That's good to know.
jgeerds has quit [Ping timeout: 265 seconds]
<mzabani>
thanks a lot, @clever, I'll try to reboot and see how that goes. Otherwise I'll try running switch again with a working channel.
murmeli is now known as rural_fox
mzabani has quit [Remote host closed the connection]
<infinisil>
philipp[m]2: Because buildInputs /= runtime
<infinisil>
,runtimeDeps philipp[m]2
<{^_^}>
philipp[m]2: In order of preference: Patch source OR ((if it uses PATH -> wrap with new $PATH) AND (if it uses dlopen, (patchelf --set-rpath in postFixup OR wrap with new LD_LIBRARY_PATH)))
<infinisil>
I guess this isn't even complete ^^, but "runtime" can mean a lot of things. It really depends on how the package uses it
sondr3 has joined #nixos
cosimone has quit [Quit: Quit.]
<philipp[m]2>
It's python, so it uses pythonpath. I thought there was some magic in buildPythonApplication, that did it.
<clever>
i think the magic, is that all inputs wind up in PYTHONPATH at build time
<clever>
and something will create a shell-script wrapper, that uses the build-time $PYTHONPATH
<philipp[m]2>
clever: Yes, that's what I thought, but it somehow doesn't here.
<clever>
so the PYTHONPATH persists at runtime
<clever>
does such a shell script exist?
<philipp[m]2>
Yes, there is a wrap script, but it doesn't set PYTHONPATH at all, so I ran out of ideas and came here.
<{^_^}>
[nixpkgs] @risicle opened pull request #75529 → [r19.09] glibc: add patch for CVE-2019-19126 → https://git.io/Je9pw
<kolaente_>
Hi there, I added nixos unstable to my channels to get some packages from unstable instead of 19.09. The entry in my configuration.nix looks pretty much like the question on https://stackoverflow.com/q/48831392/10924593 and has worked well in the past. However, I'm now trying to install go 1.13 from unstable, by adding unstable.go to my user
<philipp[m]2>
Ah! I need propagatedBuildInput, not buildInput.
<kolaente_>
's packages, but I still get 12.9, which is the version on stable
<clever>
kolaente_: which user did you add the channel to? are you running `nix-channel --update` as that user?
<kolaente_>
clever: I added the config to users.users.<user>.packages in my configuration.nix
<kolaente_>
I updated the channel as root
<kolaente_>
Since I'm also building my system as root
<clever>
kolaente_: what does `which --all go` return?
<kolaente_>
which: bad option: -l
<clever>
kolaente_: `which which` ?
<kolaente_>
clever: `shell built-in command`
<clever>
ah, thats why
<clever>
we want the which in $PATH
<kolaente_>
`which go` gets me `/home/user/.nix-profile/bin/go`
<clever>
try `command which --all go` ?
<clever>
yep, theres your problem
<clever>
you installed go with `nix-env -i`, so its ignoring what configuration.nix did
<fresheyeball>
o1lo01ol1o: Its a great thing, but doesn't fully do what I need
<clever>
you must uninstall that copy with `nix-env -e go`
<kolaente_>
clever: As my user?
<clever>
kolaente_: yeah
<kolaente_>
interesting though, I cannot remember doing that (which does ofc not mean I didn't do it -.-)
<clever>
kolaente_: ls /nix/var/nix/profiles/per-user/clever/profile*/bin/go
<clever>
kolaente_: do you see which generation it was added at?
<kolaente_>
clever: That gives me ~20 lines of /nix/var/nix/profiles/per-user/user/profile-<n>-link/bin/go
<kolaente_>
clever: How do I find the generation?
ckauhaus has quit [Quit: WeeChat 2.6]
murmeli is now known as urban_mumeli
drakonis has quit [Ping timeout: 265 seconds]
<clever>
kolaente_: what is the lowest <n> in the list?
<kolaente_>
clever: 2
<kolaente_>
Does that mean the second generation?
<clever>
yep
<clever>
kolaente_: and then `nix-env --profile /nix/var/nix/profiles/per-user/user/profile --list-generations`
<clever>
that will show dates for each
urban_mumeli is now known as murmeli
murmeli is now known as murmeli_
murmeli_ has quit [Quit: WeeChat 2.6]
<kolaente_>
clever: Ah ok, I see it is an old one, back from june
<clever>
thats likely when you installed go to the user's profile
<kolaente_>
And the newer versions came through updates? Even though I never did a `nixos-rebuild switch` as my user?
cfricke has quit [Quit: WeeChat 2.6]
dbmikus__ has quit [Ping timeout: 276 seconds]
dbmikus has joined #nixos
<kolaente_>
clever: I guess I need to login again for `nix-env -e go` to have an effect? Since I did that and `go version` still gives me 12.9
<clever>
nope, it should take effect immediately
<clever>
you can use `which go` and `nix-env -q` to confirm
<clever>
you may need to use the exact name `nix-env -q` shows
<kolaente_>
clever: `which go` gives me /home/user/.nix-profile/bin/go and `nix-env -q` only `home-manager-path`... I'm not sure if I understood that correctly
<clever>
kolaente_: oh!, so its been installed by home-manager
<clever>
kolaente_: but `users.users.<user>.packages in my configuration.nix` isnt home-manager, as far as i know? ...
<clever>
kolaente_: double-check your home-manager config?
<kolaente_>
clever: Yes, I just checked it and indeed, there it is... damn...
<kolaente_>
The thing is, I already had go 1.12 pinned in my configuration.nix, and because of that I just assumed it was installed from there and never looked further
<kolaente_>
clever: half?
<fresheyeball>
clever: where did we leave off on pytorch?
<clever>
kolaente_: we have the how, but not the why
<fresheyeball>
what should I investigate?
<fresheyeball>
I feel like no one wants ma money
<clever>
fresheyeball: the configure phase, use postConfigure = "exit 1"; to speed up iteration, and then try and figure out why cmake claims cuda is disabled
<fresheyeball>
clever: got it
<kolaente_>
clever: I threw the go config out of the home-manager config and after a `home-manager switch` it works like it should
<kolaente_>
I also took a look at my git, it seems I added that go statement in my home manager config back when I first tried nix - when I really had no clue about it all
Jonathan83 has quit [Remote host closed the connection]
<zeta_0>
if i understand correctly, i can use ssh-agent to automatically use my ssh password, but what about gpg, i don't want to have to type the password over and over again when using git?
<lassulus>
gpg-agent?
<clever>
zeta_0: gpg-agent can remember the gpg passphrase for a set period of time, before forgetting and asking again
<zeta_0>
clever: so gpg remembers the passwords for both ssh and gpg?
<zeta_0>
i thought i needed do setup ssh-agent and gpg-agent?
<zeta_0>
it remembers it automatically?
ddellacosta has joined #nixos
<clever>
zeta_0: when you run `ssh-add` with the gpg ssh agent, it will ask for 2 passwords, first one to decrypt id_rsa, then a 2nd one, to re-encrypt it, within the gpg keyring
<clever>
zeta_0: then it will be saved into the gpg keyring forever, and any time an app wants to use it, gpg will ask you for the 2nd pw
<clever>
and this one makes it remember the unlock forever (until shutdown)
<ajs124>
Are there any firefox (packaging) experts around? I'm trying to add policies support to firefox, but it searches in the unwrapped path, instead of the wrapped one.
<zeta_0>
so do i need to enable a nix option or put some commands in sessionCommands, i am confused?
MmeQuignon has quit [Quit: WeeChat 2.6]
pikajude has joined #nixos
bvdw has quit [Read error: Connection reset by peer]
<clever>
zeta_0: is gpg agent running by default already?
bvdw has joined #nixos
ddellacosta has quit [Read error: Connection reset by peer]
ddellaco1 has joined #nixos
<zeta_0>
clever: i just setup gpg and ssh keys setup for both gitlab and github'
<zeta_0>
clever: i don't know if gpg is running by default?
<evanjs>
Any general guidelines on what might be happening for GTK(2) apps that encounter 'The name org.freedesktop.secrets was not provided by any .service files' ?
<ornxka>
yeah the whole reason i got the ci20 was to run bare metal code on it :p
<ornxka>
i had no idea you could do it so easily with nix
<clever>
ornxka: currently, it only works on the rpi 1-3, and its even more baremetal then the examples, because the arm isnt online, neither is the ram!
<clever>
zeta_0: 1481 is the new pid
<ornxka>
getting cross compilers set up on any other system is a gigantic pain in the butt
<ornxka>
>neither is the ram
<ornxka>
ram bootstrapping code to me has always seemed like the most arcane black magic ive ever heard of
<ornxka>
i dont know where they get those timing numbers, i think they just make them up
<elvishjerricco>
So I want to install NixOS on a new system, using a USB thumb drive for `/boot`. But I awkwardly only have one USB drive lying around at the moment, which I use to boot the installer. Is there a way to reuse the USB drive from within the installer?
<clever>
ornxka: my fork has nix support, and if you just `nix-build -A helper`, it will generate all of the proper cross-compilers (line 3, it doesnt care which channel your on), then cross-compile the entire firmware, and generate a bash script that updates the vfat partition in your SD reader
<zeta_0>
clever: no that command returned nothing
<qyliss>
elvishjerricco: boot installer with copytoram, then you can do whatever you want with the drive
<clever>
zeta_0: replace 2179 with the new pid
<ornxka>
man that is so cool
<elvishjerricco>
qyliss: ah, didn't know there was such a thing.
<clever>
zeta_0: 2019-12-11 19:39:28 < clever> zeta_0: 1481 is the new pid
<elvishjerricco>
thanks
<qyliss>
it's in the menu when you boot the installer :)
<ornxka>
i hope some day to be half as good at nix as you are
<clever>
ornxka: yep
<elvishjerricco>
now to find a monitor for this blasted thing so I can see the boot menu :P
<elvishjerricco>
or is there a way to make an iso that does copytoram by default?
<qyliss>
probably :P
<clever>
elvishjerricco: the multi-boot helper i have may also work for that
<clever>
elvishjerricco: if you run nix-build on this, it will generate a directory with 3 files, nixos-kernel, nixos-initrd, and grub-fragment.cfg
<clever>
elvishjerricco: you then add that grub-fragment to a grub.cfg, and grub-install as usual
<clever>
elvishjerricco: the rootfs is fully contained within nixos-initrd, so its always in a copytoram mode
<elvishjerricco>
interesting...
<clever>
elvishjerricco: it was originally made for somebody that wanted a usb stick that can boot all the installers (multiple distro's in the grub menu)
<zeta_0>
clever: so that means that gpg-agent is working and will automatically save the passwords for both ssh and gpg?
<clever>
elvishjerricco: if another linux distro is currently working on that machine, you can just copy this magic tarball over, run 2 commands, and it will instantly be running nixos from ram
<elvishjerricco>
clever: Ah, no there is currently not another linux distro on it
<clever>
and that is the rabbit hole that led me to learning the M$ bootloader, lol
<elvishjerricco>
:P
<elvishjerricco>
Can windows be kexec'd into?
<clever>
i wanted to add linux to the M$ bootloader, so you could transition over like that, without a usb stick, and without dd'ing over the live disk
<clever>
elvishjerricco: to windows, probably not, from windows, maybe
<clever>
elvishjerricco: also, colinux was a thing