<simpson>
The behavior is driven by typical git commands, so it's certainly not expected to get a failure from that. How does it fail to work?
<FireFly>
it complains that "fatal: reference is not a tree: 1cebc0895d236bfc5cd6797d03e02c55c773ddf1" when trying to `git checkout <hash>` after it runs `git fetch origin`
<FireFly>
I tried changing the git fetch in checkout_hash in nix-prefetch-git to specifically check out "$hash", and in that case it works as expected... but also, that line has seemingly been like that forever, and the function is documented as "fetch everything and checkout the right sha1", so I'm not sure if changing to 'git fetch origin "$hash"' might break something else..
<FireFly>
the repo is https://github.com/jmesmon/illum @ tag v0.5 if you want to test/experiment yourself (came across as I was looking if I could update that package)
bennofs_ has joined #nixos
cosimone has quit [Ping timeout: 264 seconds]
thelounge52 has joined #nixos
dandart has joined #nixos
mbrgm_ has joined #nixos
<fuzzypixelz>
thanks colemickens, mgdm!
mbrgm has quit [Ping timeout: 272 seconds]
mbrgm_ is now known as mbrgm
bennofs__ has quit [Ping timeout: 256 seconds]
dandart has quit [Remote host closed the connection]
gustavderdrache has joined #nixos
Cale has quit [Ping timeout: 256 seconds]
dansho has quit [Ping timeout: 272 seconds]
thelounge52 has quit [Ping timeout: 272 seconds]
work_ has quit [Quit: Connection closed for inactivity]
<iqubic>
I keep getting the error: "LSP :: Error from the Language Server: Rust Analyzer is still loading... (Unknown error)" printed in the emacs minibuffer again and again.
<bbigras>
rust-analyzer is in the PATH? you can also try to restart lsp
<iqubic>
Also, I'm literally using a version of Rust Analyzer from Feb. 8th of this year.
thelounge52 has joined #nixos
rb2k has quit [Ping timeout: 264 seconds]
growpotkin has joined #nixos
<ece3>
Just submitted my first package init pr: https://github.com/NixOS/nixpkgs/pull/112603 I've bumped packages and added a tmux plugin, but this is my first package, kinda proud
<danderson>
Ohai. Did kernel 5.8 get removed from nixos 20.09?
<danderson>
I ask because virtualbox currently refuses to function on >=5.9 kernels in 20.09, so afaict I can't both upgrade and keep virtualbox any more.
<danderson>
if this builds, I guess I should send a PR to release-20.09 to unbreak the module. Weirdly that change wasn't a backport, it was just a random peppering of "mark a bunch of things broken"
<danderson>
with no explanation as to why.
kaivai has joined #nixos
<cole-h>
Maybe part of ZHF?
<cole-h>
e.g. everything prior to release that has yet to be fixed is marked broken.
<danderson>
this was in release-20.09, well after the release though
<danderson>
as far as I can tell, for a reason I haven't found yet, linux 5.8 got removed from 20.09 after release, and it seems like during that process, all version-dependent kernel modules were marked as broken on >5.8
<gchristensen>
ah
<gchristensen>
well I can answer why 5.8 was removed -- it isn't a supported kernel upstream anymore
<danderson>
I feel like I'm missing a well-understood release management process that led to this, I'm not familiar enough with how nixos releases get maintained after release
<danderson>
okay, so that makes sense. 5.8 got dropped, the vbox shipped in 20.09 didn't support 5.9+, so it got marked broken. Later on a fixed vbox was backported, but the kernel module didn't get un-marked broken.
<gchristensen>
yeah, that sounds likely
<danderson>
assuming it does in fact compile *stares at build box*, I'll send a PR to correct
<gchristensen>
wonderful :)
<cole-h>
danderson++ gchristensen++
<{^_^}>
danderson's karma got increased to 0xb
<{^_^}>
gchristensen's karma got increased to 423
<gchristensen>
there was a lot of discussion on if stable should even get non-"longterm" kernels since they do disappear
<gchristensen>
ultimately decided to do it to go for more choice, at the expense of people, well, having them disappear sometimes
CMCDragonkai1 has joined #nixos
dansho has quit [Remote host closed the connection]
<danderson>
gchristensen: I for one am grateful for that option, since my laptop's GPU doesn't boot at all on the longterm kernel :)
<danderson>
even if it leads to some wonkiness now and again
ArdaXi has quit [Ping timeout: 258 seconds]
marcusr has quit [Remote host closed the connection]
dansho has joined #nixos
ArdaXi has joined #nixos
marcusr has joined #nixos
lockeanarchist has quit [Ping timeout: 256 seconds]
<gchristensen>
yeah, that was essentially the deciding factor iirc :)
srl295 has quit [Quit: Connection closed for inactivity]
<gchristensen>
opa
thelounge52 has joined #nixos
<aleph->
Hmm can I specify a package to grab from cachix with `nix copy` or do I need to specify the exact revision, build, etc? Wanting to stand up a local small mirror of cachix at home.
jonringer has quit [Remote host closed the connection]
patagonicus6 has joined #nixos
thelounge52 has quit [Ping timeout: 240 seconds]
patagonicus has quit [Ping timeout: 240 seconds]
patagonicus6 is now known as patagonicus
Diogo9 has joined #nixos
Diogo has quit [Read error: Connection reset by peer]
<clever>
looks less like a netsplit, and more like the matrix gateway died
<clever>
netsplit is more of a problem on the freenode end of things
<t3hyoshi>
Ah, I guess that's what the [m] means.
nhs has joined #nixos
zakame has quit [Ping timeout: 240 seconds]
sarcasticadmin has joined #nixos
iv_nn[m] has joined #nixos
Hirmes[m] has joined #nixos
us3r[m] has joined #nixos
kazimazi[m] has joined #nixos
tsmeets[m] has joined #nixos
ncm[m] has joined #nixos
awwayaya24hours[ has joined #nixos
linus0[m] has joined #nixos
jlv has joined #nixos
cdknight[m] has joined #nixos
nikivi[m] has joined #nixos
marijan[m] has joined #nixos
ccao001[m] has joined #nixos
macerbi[m] has joined #nixos
mica[m] has joined #nixos
alekhine[m] has joined #nixos
hsiktas[m] has joined #nixos
felschr[m] has joined #nixos
travelion[m] has joined #nixos
phirsch has joined #nixos
cwslimy[m] has joined #nixos
kueckieben[m] has joined #nixos
viq[m]1 has joined #nixos
a7p[m] has joined #nixos
qasaur[m] has joined #nixos
<danderson>
Would anyone be up for reviewing a small patch series in a single PR? I'm updating the nmigen ecosystem, which involves boilerplate touches to 3 packages, and a minor bugfix on a 4th.
MrGoose[m] has joined #nixos
dnlkrgr[m] has joined #nixos
chrismatheson has joined #nixos
littlebenlittle[ has joined #nixos
caf3d00d[m] has joined #nixos
simara[m] has joined #nixos
<ashkitten>
how do i enable a kernel config option?
davegallant[m] has joined #nixos
fast1[m] has joined #nixos
<clever>
packageOverrides = pkgs: rec {
<clever>
XXXlinux_4_9 = pkgs.linux_4_9.override {
<clever>
extraConfig = ''
<clever>
NF_CT_PROTO_DCCP n
<clever>
ashkitten: an old override i have since disabled
Ollie[m]1 has joined #nixos
Supersonic112 has joined #nixos
<ashkitten>
so if i want to enable FUTEX2
Supersonic has quit [Ping timeout: 258 seconds]
Supersonic112 is now known as Supersonic
e8zhf[m] has joined #nixos
<ashkitten>
what would i put in extraConfig?
<clever>
ashkitten: change the extraConfig to `FUTEX2 y` i believe
<JasonO>
I am receiving the following error after trying to install python. I'd like to preferably not ignore broken packages:
simba1 has quit [Ping timeout: 272 seconds]
<JasonO>
building Nix...
<JasonO>
building the system configuration...
<JasonO>
error: Package ‘python2.7-pyside-1.2.4’ in /nix/var/nix/profiles/per-user/root/channels/nixos/pkgs/development/python-modules/pyside/default.nix:24 is marked as broken, refusing to evaluate.
<JasonO>
a) For `nixos-rebuild` you can set
<JasonO>
{ nixpkgs.config.allowBroken = true; }
<JasonO>
in configuration.nix to override this.
<JasonO>
b) For `nix-env`, `nix-build`, `nix-shell` or any other Nix command you can add
<JasonO>
{ allowBroken = true; }
<JasonO>
to ~/.config/nixpkgs/config.nix.
<JasonO>
(use '--show-trace' to show detailed location information)
<JasonO>
Any help would be appreciated
thelounge52 has joined #nixos
nhs has joined #nixos
thc202 has joined #nixos
nhs has quit [Ping timeout: 272 seconds]
danderson has quit [Remote host closed the connection]
<valya>
hello! I'm trying to get a haskell thing up & running, and don't know much about nix. I get the following error: "error: anonymous function at /nix/store/kj67mb95xdlvmyl7b4kr1q1yi4knmrdj-cabal2nix-factorio-total-compiler/default.nix:1:1 called without required argument 'base64', at
<valya>
i think that cabal2nix generates a file that mentions base64, my dependency, but nix interprets base64 as a name of its function or something? how do i make sure it understands it's a package?
Jd007 has quit [Quit: Jd007]
palo1 has joined #nixos
cole-h has joined #nixos
palo has quit [Ping timeout: 264 seconds]
palo1 is now known as palo
<clever>
valya: does haskellPackages.base64 exist?
<valya>
I'm trying to get a haskell thing up & running, and don't know much about nix. I get the following error: "error: anonymous function at /nix/store/kj67mb95xdlvmyl7b4kr1q1yi4knmrdj-cabal2nix-factorio-total-compiler/default.nix:1:1 called without required argument 'base64', at
<valya>
i think that cabal2nix generates a file that mentions base64, my dependency, but nix interprets base64 as a name of its function or something? how do i make sure it understands it's a package?
<dmj`>
There’s base64-bytestring is rn
<dmj`>
This* what you refer to!
<dmj`>
?*
<valya>
no, just base64... i guess i picked the wrong package
n-osborne has quit [Remote host closed the connection]
mallox has quit [Quit: WeeChat 3.0]
<dmj`>
We need a Haskell, that’s all I can say
<dmj`>
new Haskell *
n-osborne has joined #nixos
nhs has quit [Ping timeout: 246 seconds]
<attila_lendvai_>
could it be that my laptop HDD doesn't spin down with nixos? i'm using it as a server. is there a standard way to configure it using nixos? i don't see anything hdparm related in the docs.
thelounge52 has joined #nixos
ahmedelgabri has quit [Remote host closed the connection]
<jeeger>
Hey! I'm trying to build a nix shell for a java application, and I've succesfully created a shell.nix to run the app (yay!). However, I seem to have a problem with the gsettings schemas being put into the wrong directory (boo!). My XDG_DATA_DIRS contains a gsettings-desktop-schemas entry. The gsettings.compiled file is missing though, leading to crashes. How can I go about debugging this.
<cole-h>
So I guess I was wrong: it's *15* builders at the moment
<cole-h>
Mic92: btw you're probably right -- it's likely that the x86_64 queue was bogged down and once it got to mumble, the tag had been force-pushed
<jeeger>
eyJhb: Sorry, that was misleading. The problem is the missing gsettings schema, which lead to the crash.
<cole-h>
but aarch64 had already fetched the source before the tag was force-pushed
<jeeger>
The gsettings.compiled file is in a subdirectory ($out/gsettings-schemas/gsettings-deskotp-schemas-3.38.0/glib-2.0/schemas) that's not in XDG_DATA_DIRS, and I'm wondering how to get the gsettings-desktop-schemas derivations to put it into the correct directory ($out/share/glib-2.0).
aidalgol has left #nixos ["Using Circe, the loveliest of all IRC clients"]
Robin[m]2 has left #nixos ["User left"]
ArdaXi has quit [Quit: WeeChat 3.0]
<Mic92>
jeeger: I have seen this issue before in gsettings
<Mic92>
I mean the gsettings commandline app
sangoma has quit [Read error: Connection reset by peer]
mgdelacroix has left #nixos ["ERC (IRC client for Emacs 26.3)"]
<jeeger>
Mic92: Thanks, I think I get it. How would I apply this in a nix shell? I'm wondering where $GSETTINGS_SCHEMA_PATH is defined, so I can get it into the nix shell.
haphaphap has left #nixos [#nixos]
<jeeger>
Ah, I got it. It's in the environment when I pull in glib.
<endocrimes>
Is there a convenient way to package python packages that depend on an older version of another package than we have in nixpkgs? - Or is it a case of duping the dependencies, reverting the version and just having them in my personal pkgs repo?
attila_lendvai_ has quit [Ping timeout: 265 seconds]
lord| has quit [Ping timeout: 246 seconds]
lord| has joined #nixos
attila_lendvai_ has joined #nixos
nhs has quit [Ping timeout: 256 seconds]
<jeeger>
Mic92: Alright, that did it. Thanks for your help! Just a quick aside, is there a way to get GSETTINGS_SCHEMAS_PATH without pulling in glib as a dependency?
civodul has joined #nixos
<patagonicus>
endocrimes: If anything you can copy the old python derivation from nixpkgs and use that one for the package, but you might end up needing to copy some of the dependencies as well. You could also use a pinned nixpkgs, but that means not getting any security updates for anything that the package/that python depends on, so I wouldn't do that long
<patagonicus>
term.
<Mic92>
jeeger: maybe by implementing its setup-hook: nix-shell -p glib --run 'cat $buildInputs/nix-support/setup-hook'
<adisbladis>
endocrimes: How do you want to use these packages?
<endocrimes>
It's a CLI tool that I have to use for work, annoyingly packaged as a wheel so poetry probably wouldn't work?
malook has joined #nixos
<endocrimes>
I haven't really interacted with python beyond the bare minimum in a decade tho so whenever I need to package non-trivial python stuff I usually just end up building a container for it instead :|
<energizer>
do you have access to the source code of the package you're talking about?
<endocrimes>
Technically yes, but not in a way I can reference from open source afaik
<energizer>
does the package include C extensions (the package itself, not including its deps)
nhs has joined #nixos
malook has quit [Client Quit]
<energizer>
how many deps does it have, are they all in nixpkgs
<endocrimes>
(it's https://pypi.org/project/dcli/, and afaik no C extensions, it installed pretty cleanly in venv and stuff with c deps rarely do)
<endocrimes>
It's a small handful, everything I ran into was in nixpkgs, but something was too new (3.x vs 2.x) - I last tried like 2 weeks ago tho so don't remember what it was off the top of my head
<energizer>
`poetry init` will make a pyproject.toml and poetry.lock, then make a default.nix containing {poetry2nix}: poetry2nix.mkPoetryApplication {projectDir = ./.;}
<energizer>
if you're lucky that'll be it. sometimes you'll have to add some overrides (see the poetry2nix docs for that)
MasseR has quit [Ping timeout: 264 seconds]
nhs has quit [Ping timeout: 264 seconds]
<energizer>
(pure python wheels are just zip files)
* energizer
out
mog has quit [Ping timeout: 258 seconds]
<energizer>
...actually, i'm not sure if what i said will work. i haven't tried directly packaging a wheel with poetry2nix. adisbladis what would you do?
mog has joined #nixos
<endocrimes>
could probably actually hack around this - add the wheel as a dependency, then (hopefully) poetry has some way to resolve those deps and add them to the project?
zupo_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nhs has joined #nixos
lsix has joined #nixos
kaletaa has joined #nixos
nhs has quit [Ping timeout: 240 seconds]
<kaletaa>
hello - how am i supposed to change /etc/issue? I haven't found any config options for it
waleee-cl has joined #nixos
j2t has quit [Quit: Leaving]
<lassulus>
environment.etc.issue.text = "bla";
<kaletaa>
thanks
j2t has joined #nixos
CMCDragonkai1 has joined #nixos
<kaletaa>
is there a reason for it not being on the option index?
<kaletaa>
nevermind, i understand
kaletaa has quit [Client Quit]
<immae>
kaletaa: it is
<AmandaC>
Environment.etc.<name> is -- it's a attrset of files to put under /etc
<lassulus>
hes gone
MasseR has joined #nixos
<AmandaC>
So I see, that's what I get for tryingto help from my phone. :P
j2t has quit [Ping timeout: 246 seconds]
PraiseuD has joined #nixos
PraiseuD has quit [Connection closed]
nhs has joined #nixos
elux has joined #nixos
<elux>
hi all
<elux>
im having trying to home-manager --switch, but getting the error: The option `services.redshift.brightness' does not exist. Definition values: { day = "0.95"; }
<gchristensen>
big <3 energy to sigyn and freenode staff
<jess>
:>
orivej has joined #nixos
<matthewcroughan>
gchristensen: how would you deploy changes in a configuration to machines?
<gchristensen>
<3 jess
<{^_^}>
jess's karma got increased to 1
<matthewcroughan>
I could imagine git pull, but what would you do?
kaletaa has joined #nixos
<gchristensen>
matthewcroughan: ask me in a few hours? time for breakfast
<matthewcroughan>
I just had mine :D
<matthewcroughan>
enjoy yours
<matthewcroughan>
I could also imagine etcd, so many ways
<kaletaa>
sadly setting `environment.etc.issue.text` didn't work, my /etc/issue is still the nixos default one, i also tried setting /etc/static/issue this way, same result
Darkmatter66 has quit [Quit: ZNC 1.8.2 - https://znc.in]
oldandwise has joined #nixos
<{^_^}>
[nix] @edolstra merged pull request #4535 → Revert "narinfo: Change NAR URLs to be addressed on the NAR hash instead of the compressed hash" → https://github.com/NixOS/nix/pull/4535
<{^_^}>
[nixpkgs] @peti pushed 2 commits to haskell-updates: https://git.io/JtozK
benedikt93 has quit [Remote host closed the connection]
nhs has joined #nixos
rb2k has quit [Ping timeout: 264 seconds]
lsix has quit [Ping timeout: 260 seconds]
saschagrunert has quit [Remote host closed the connection]
<exarkun>
Anyone used services.gitlab and want to make a recommendation or warning?
vidbina has quit [Ping timeout: 272 seconds]
<meh`>
exarkun, use gitea instead if you can
<exarkun>
I have 500MB of GitLab exports that I gotta do something with :/
<shapr>
exarkun: oh hey, did you get it fixed? Was it nix-channel done as user instead of root?
<exarkun>
shapr: I fixed it but I won't claim to understand what was happening
<shapr>
heh, fair enough
<exarkun>
My user has NIX_PATH set to include nixpkgs=/nix/var/nix/profiles/per-user/root/channels/nixos
<exarkun>
but I don't know what set that
<exarkun>
And then I guess `nix-channel` and `nix repl` disagree on how they find nixpkgs in some way I don't understand?
<shapr>
hm, my NIX_PATH is already set that way
rb2k has joined #nixos
lesby has quit [Quit: Connection closed for inactivity]
<exarkun>
And then my user has ~/.nix-channels but /nix/var/nix/profiles/per-user/exarkun/ also exists
<exarkun>
So there's ... at least three different sources of truth about what channels exist that could be used
<gchristensen>
~/.nix-channels is just a list to tell `nix-channel` what to do
<gchristensen>
`nix-channel` creates a new channel profile in /nix/var/nix/profiles/per-user/exarkun/ when run
<exarkun>
gchristensen: Actually when run? Or a little bit later - maybe when nixos-rebuild is run?
<exarkun>
Or `nix-channel --update`?
respawn_ has joined #nixos
cyraxjoe has quit [Read error: Connection reset by peer]
<gchristensen>
right, so add, remove, update (maybe list) touches that file
Qwerky has quit [Remote host closed the connection]
thelounge52 has joined #nixos
disasm has quit [Quit: WeeChat 2.0]
<exarkun>
Ah okay. So /etc/profile sources a `/nix/store/...-set-environment` and that sets NIX_PATH to include nixpkgs pointing at root's channels
<gchristensen>
everything but nix-env uses NIX_PATH to find nixpkgs, nix-env is a bit weird and uses ~/.nix-defexpr, which is a bit of a vestigial and historical thing. Probably one reason people don't like nix-env :)
disasm has joined #nixos
<exarkun>
Fortunately there's plenty of reasons not to like nix-env so I never use it
<shapr>
same here
<exarkun>
Unfortunately I do like `nix repl` but colemickens had some negative things to say about it yesterday
fuzzypixelz has quit [Quit: Leaving.]
<gchristensen>
oh?
<simpson>
I dislike the ergonomics of typing '<nixpkgs>' since I use US QWERTY, but that is a problem with my keyboard-pokers.
<exarkun>
something about behavior of 1.0 vs 2.0 commands
<redcedar[m]>
Hi! I'm trying out i3wm (which is great!) but I seem to have lost a way to control power settings on my laptop. I've tried using mate, but it doesn't seem to actually change settings. Is there a best practice for this, or something I'm missing?
n-osborne has quit [Remote host closed the connection]
attila_lendvai_ has quit [Ping timeout: 256 seconds]
simonpe^^ has quit [Remote host closed the connection]
n-osborne has joined #nixos
n-osborne has quit [Remote host closed the connection]
berberman_ has joined #nixos
<exarkun>
So ... looks like services.gitlab is pretty rudamentary, really. For example, it doesn't even make gitlab start listening for http/s connections?
<{^_^}>
[nixpkgs] @petabyteboy opened pull request #112671 → matrix-synapse-plugins.matrix-synapse-shared-secret-auth: init at 1.0.2 → https://github.com/NixOS/nixpkgs/pull/112671
berberman has quit [Ping timeout: 260 seconds]
<qyliss>
I don't know anything about GitLab in particular, but that's often a policy decision
<qyliss>
because if you're already running a web server, like Nginx, you're going to want it to front GitLab
<exarkun>
But it doesn't listen on _anything_... so nothing for nginx to reverse-proxy to
<qyliss>
oh
<exarkun>
it does start a redis server listening on :: and 0.0.0.0, though
n-osborne has joined #nixos
<patagonicus>
Weird. On my system https://github.com/NixOS/nixpkgs/commit/9c213398b312e0f0bb9cdf05090fd20223a82ad0 seems to lose the extraConfig specified in the hostPlatform. From what I can tell the change to the assignment to intermediateNixConfig looks good. I'm currently building the kernel with manually specifing the stuff from the hostPlatform extraConfig
<patagonicus>
in my configuration.nix to see if that fixes it. This is armv7l-hf-multiplatform, but I don't think the behavior is platform specific. If anyone has an idea what's going on, that would help. :)
n-osborne has quit [Remote host closed the connection]
<exarkun>
Oh. That's not even the random unix socket I stumbled over. Heh.
aleph- has quit [Client Quit]
steell has quit [Remote host closed the connection]
<exarkun>
I was hoping that services.gitlab would be as easy to get started with as the bitwarden gitlab ec2 ami, but maintainable after setup instead of refuse slowly decomposing
<exarkun>
but this is way harder
<exarkun>
probably time to give up
respawn_ has joined #nixos
Paul84 has joined #nixos
aleph- has joined #nixos
ahmedelgabri has joined #nixos
simba1 has joined #nixos
<Paul84>
Hello everyone, I have a quick beginner question, which I hoped someone could me point towards the right direction about: I am missing the readline manual, although readline seems to be in my current environment via bash. I set `documentation.dev.enable = true`. Appearently the man pages where not even in my nix store because after `nix-shell -p
<Paul84>
readline70.man` it was fetched. In the resulting shell `man readline` still fails. What am I doing wrong? I had hoped I would get the man pages of all the stuff I'm using, even transiently (the readline manpage has instructions for .inputrc for example which are relevant for bash and others). Related: Is there a way to install manpages for standard
<Paul84>
stuff, e.g. `man 7 execve`? Thanks
aleph- has quit [Read error: Connection reset by peer]
<aleph->
Is there a way I can specify a target release in nixops? Keeps upgrading me to 21.03/unstable
<aleph->
Which breaks a few things
<Paul84>
aleph- I have no clue, but isn't that driven by the channel you are using. What's your nix-channel --list saying?
yjftsjthsd has joined #nixos
<Paul84>
oh, nixops
<Paul84>
no idea
<Paul84>
sorry
<aleph->
Ah good to know that's set like that.
<Paul84>
so for me (NixOS), its `nixos https://nixos.org/channels/nixos-unstable`, so when I do e.g. `nixos-rebuild --upgrade switch` it will fetch unstable head first and upgrade.
<cole-h>
aleph-: I think this is limited to nixops 2.0, but setting `network.nixpkgs` in your deployment.nix lets you set the nixpkgs to use for the network
<Paul84>
So I put "readline70.man" into systemPackages now, this gets me the manpages. It's odd, that "nix-shell -p readline70.man" didn't get me the same result. Feels like a hack though as I have no business with readline70 otherwise... Happy for suggestions ;)
tarruda2 has joined #nixos
<Paul84>
Maybe it's an issue with the "bash", "gdb", etc. packages though as this aspect of their configuration is in that manpage. So they should pull them in?
<Siyo>
Does anyone use nix-direnv with nix flakes? I noticed it tends to wreck havoc with a lot of tools as it creates a symlink loop where /foo/.direnv/flake links to /foo.
<Siyo>
not great for applications that don't ignore .direnv
<colemickens>
I use nix-direnv with flakes and have never noticed what you're talking about
<colemickens>
well, actually, I guess it's in my gitignore though, but I don't have a .direnv dir actually here...
ahmedelgabri has quit [Ping timeout: 264 seconds]
<Siyo>
I wonder if there's something weird on my end then. I do see some symlinking logic for gc roots, but that's it
Qwerky has quit [Remote host closed the connection]
<Siyo>
I deleted .direnv and let it re-create it.. now it's fine lol
<Siyo>
no idea how that happened
vermelhas has joined #nixos
<Siyo>
ok actually I think I have some idea. It's not a symlink loop, but it looks like each time the environment is updated, it creates a symlink to a new /nix/store/sources-* path, with the new one referencing the old symlink
<Siyo>
so after 10 updates you will have 10 layers of your source code in the directory tree
xelxebar has quit [Remote host closed the connection]
xelxebar has joined #nixos
thelounge52 has joined #nixos
simba1 has quit [Ping timeout: 260 seconds]
thelounge52 has quit [Ping timeout: 246 seconds]
simba1 has joined #nixos
hiro98 has quit [Remote host closed the connection]
Qwerky has joined #nixos
simba1 has quit [Ping timeout: 265 seconds]
<ronthecookie>
woo just ported my first thing to nixpkgs :D
jeeger has quit [Remote host closed the connection]
<shapr>
yay!
hiro98 has joined #nixos
simba1 has joined #nixos
kenran has joined #nixos
hiro98 has quit [Remote host closed the connection]
kenran has quit [Client Quit]
rb2k has quit [Ping timeout: 260 seconds]
<disasm>
I just tried out starship for a prompt and in bash it looks great but in zsh even with no config aside from the hook, it has my $PWD output on the right side of the prompt. I also notice that I see that as well if I don't have starship enabled. Anyone else using zsh and starship with nixos that knows how to fix this?
zebrag has quit [Quit: Konversation terminated!]
zebrag has joined #nixos
neminis has quit [Quit: Ping timeout (120 seconds)]
<bootstrap-prime>
most relevant part being: /nix/store/wcz8lsibkw2nikdb0bjl22p75lfwahzr-binutils-2.35.1/bin/ld: /nix/store/m0xa5bz7vw7p43wi0jppvvi3c9vgqvp7-glibc-2.32-25/lib/crt1.o: in function `_start':
<bootstrap-prime>
It builds fine in nix-shell, however. g++, make, bash, etc are all the same version, so is there any conceivable reason for why this isn't working?
<bootstrap-prime>
any assistance with this would be appreciated; I'm really out of my depth here. that it's not building is fucking weird and I'm half tempted to just use nix-shell inside the nix build.
<disasm>
answering my own question, starship hook belongs in promptInit, not shellInit
<bootstrap-prime>
you can verify that the downloaders and all that work by inserting breakpointHook into the nativeBuildInputs, and then cloning the source directory to your local, turning network access off for a shell, and going into nix-shell
<bootstrap-prime>
and then running the same commands in buildPhase
<pisquo[m]>
Hi everyone I'd like to package spacevim as a package or a nixos module, thing is the installation process is an sh that replaces cone config files in the home with the spacevim ones. Is there a way to execute this shell script within the configuration while keeping information on the home of the installing user?
<ronthecookie>
pisquo[m]: sounds like it could be better as a home-manager module
<ronthecookie>
then you can do stuff like `home.file.".config/whatever".source = ./whatever.cfg`
<tpw_rules>
bootstrap-prime: well for one it's doing -I/usr/local/include
<tpw_rules>
and stuff
<bootstrap-prime>
what does that mean? why would that cause this failure?
<tpw_rules>
because that's a directory you're not allowed to access at build time
<pisquo[m]>
<ronthecookie "pisquo: sounds like it could be "> That's what I thought, is there any way to automate this process or I have to translate the script from bash to nix?
<ronthecookie>
pisquo[m]: i mean you can do that for the whole folder
<bootstrap-prime>
tpw_rules: ah. should I try and substituteInFile and get rid of it?
<bootstrap-prime>
tpw_rules: specifying that gives me this error
<bootstrap-prime>
"/nix/store/c72jf56ans028wax054mcrbpl0qxkpyc-cmake-3.19.2/nix-support/setup-hook: line 10: ./tensorflow/lite/tools/make/downloads/neon_2_sse/cmake/Config.cmake.in.tmp: Permission denied"
simba1 has joined #nixos
<tpw_rules>
all the paths in the store are read only and that attribute got copied when you cp'd the dependencies over. maybe do chmod +w ${downloaddir}
<tpw_rules>
chmod -R +w *
hugolgst has quit [Ping timeout: 265 seconds]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<bootstrap-prime>
tpw_rules: now it's complaining about the source directory not existing, and I'm not sure where this even came from
<bootstrap-prime>
CMake Error: The source directory "/build/source/build/tensorflow/lite" does not exist.
<bootstrap-prime>
since I fed it cmakeDir = "tensorflow/lite";
<bootstrap-prime>
oh ok nvm
<bootstrap-prime>
I just had to specify from root
<bootstrap-prime>
tpw_rules it still hasn't eliminated the original cmake error of abseil cpp not existing though
srl295 has quit [Quit: Connection closed for inactivity]
<tpw_rules>
look into how it determines whether or not it exists
pushqrdx has quit [Remote host closed the connection]
<tpw_rules>
i don't have any more time to help you, sorry. but now you should be in a better place for probing around in nix-shell to help you
<pisquo[m]>
<ronthecookie "pisquo: i mean you can do that f"> Unfortunately is not that simple afaiu since the replaced file are in the user's home and should be editable (although the module can provide an option to extend them thus removing this problem).
pushqrdx has joined #nixos
zupo has joined #nixos
<pisquo[m]>
But I think that converting the script or storing the link from the store is indeed the "nixier" way to do it
<pisquo[m]>
* But I think that converting the script and storing the link from the store is indeed the "nixier" way to do it
<{^_^}>
[nixpkgs] @SuperSandro2000 merged pull request #112667 → androidStudioPackages: set JAVA_HOME in wrapper to bundled version → https://github.com/NixOS/nixpkgs/pull/112667
sigmundv__ has quit [Read error: Connection reset by peer]
Martin[m]8 has joined #nixos
kaliumxyz has joined #nixos
sigmundv__ has joined #nixos
<JasonO>
is anyone having issue with Pantheon where they cannot type in their admin password b/c the window continues to shake?
danielzi[m] has joined #nixos
teto has joined #nixos
<teto>
when I run 'nix run 'github:teto/neovim/flake-fix?dir=contrib#nvim-debug' --no-write-lock-file' '.../neovim is not allowed to refer to the following paths' /nix/store/h1yd31nnd68pyb2d3dh6n713k0g0m9xp-clang-wrapper-11.0.1 . any idea ?
remexre has quit [Read error: Connection reset by peer]
remexre has joined #nixos
attila_lendvai_ has quit [Ping timeout: 240 seconds]
<colemickens>
uh, where does Nix store cached host keys?
nhs has quit [Ping timeout: 240 seconds]
marcusr has quit [Remote host closed the connection]
marcusr has joined #nixos
meh` has quit [Ping timeout: 246 seconds]
Martin[m]8 has left #nixos ["User left"]
jonringer has quit [Remote host closed the connection]