<MtotheM>
If there is a path like this `bootloader.grub.enable` and `bootloader.systemd.enable` (i get that you don't want two boot loaders, but i'm asking about the syntax) can you group these somehow? like `bootloader.{grub, systemd}.enable`?
proofofkeags has quit [Remote host closed the connection]
<euank>
MtotheM: ^ there you go. I recommend not doing that though :)
<MtotheM>
does look really messy indeed.
<MtotheM>
I was mostly curious. cause I like the idea of opening a scope and then defining variables of a member inside of there.
<euank>
It might be handy syntax, but I don't _think_ there's any super clean way to do it in the current nix expression language
<euank>
there might be something cleaner than that mess above though
<MtotheM>
But if I have several options that are related like `desktop.xmonad, desktop.xfce` I would be nice not having to make newlines for every single enable flag
moet has joined #nixos
<MtotheM>
a function would make more sense I suppose `enableDesktops [ "xmonad", "xfce" ]`
<euank>
you can sorta visually group em with `desktop = { xmonad.enable = true; xfce.enable = true; }
<euank>
which at least gives you a scope and grouping within it
<MtotheM>
yeah, fair enough
<euank>
And even in cases where you need to preserve other parts of the attr, you can add in `// desktop` to merge it back in
<euank>
except the other way, desktop // {, I still sometimes mix that one up
<{^_^}>
#72949 (by Infinisil, 31 weeks ago, closed): Prevent `mkIf`'s from being `//`'d, and co.
<infinisil>
(just as a sidenote)
<euank>
indeed, with modules you don't need to worry about merging behavior as much
<euank>
thanks for pointing that out :)
<suzanne>
Hi :) Recently nix-shell started changing my $HOME to /run/user/1000.
<suzanne>
Probably after I followed some command's suggestion to add a nixbld group (that broke my permissions, so I rolled that back).
<suzanne>
Any idea on what keywords I should google to disable that HOME=/run/user/1000 behaviour?
<MtotheM>
I just have OCD for making things as compact as possible, as long as it's still readable of course. relying on loops and things will make it more difficult to understand.
ris has quit [Ping timeout: 246 seconds]
<euank>
suzanne: weird, can you quickly double check that `getent passwd $USER` shows your correct home directory?
Guest17609 has quit [Read error: Connection reset by peer]
Guest17609 has joined #nixos
<suzanne>
euank yep that gives the correct home dir.
azdle has joined #nixos
<euank>
neat. Lemme poke at how nix-shell could end up in the wrong place. I'm curious too now
<{^_^}>
[nixpkgs] @jonringer pushed commit from @r-ryantm to master « bdf2sfd: 1.1.1 -> 1.1.2 »: https://git.io/Jf5v6
<euank>
suzanne: I dunno, nix-shell seems pretty straightforward. The main places it seems like such a thing could happen is: 1) if HOME=$XDG_RUNTIME_DIR is happening somewhere in ~/.bashrc 2) if it's happening in a shell-hook somewhere in your defexpr..
<euank>
Yeah, those seem the only really likely possibilities I think
<suzanne>
euank: funny, I have XDG_RUNTIME_DIR=/run/user/1000 in my env… if I unset it nix-shell sets HOME=/tmp, and if I set it to $HOME then I get the correct behaviour
lord| has quit [Read error: Connection reset by peer]
upupbb-user3 is now known as ecfortes
lord| has joined #nixos
orivej has quit [Quit: No Ping reply in 180 seconds.]
metasyntactic has joined #nixos
<suzanne>
euank: hm, no relevant mention of HOME= or XDG_RUNTIME_DIR or NIX_BUILD_TOP neither in /etc nor in my dotfiles…
<suzanne>
Oh well. I'll just do if $HOME = bad_one; then export HOME=good_one; fi for now and file this as a mystery of life.
<metasyntactic>
could someone help with an issue I'm having trying to setup encrypted luks (non boot) drives? one disk is getting stuck waiting 10 seconds for udev during stage one and then times out and fails. Oddly though if I don't try to mount it automatically, it works fine manually. Any help would be very welcome
<euank>
I'm still curious. I want to say that this probably means a shell hook somewhere
orivej has joined #nixos
<metasyntactic>
8 other drives work fine, just the ninth one isn't working
<euank>
suzanne: if you `nix-shell --pure -p busybox` or such, does that one work correctly?
<energizer>
metasyntactic: what mount point is failing?
<metasyntactic>
it's the last disk, the one with the uuid starting with 7075
<suzanne>
euank: oh that one gives the correct $HOME. How do I get it to run busybox while still using my ~/default.nix ?
<metasyntactic>
if I comment out that drive in both places, the system boots and decrypts fine
<euank>
Wait, right, ofc you're doing nix-shell in a thing
<metasyntactic>
but I'm now missing a drive
<euank>
Can we see if it's `--pure` or not using the `default.nix` that you have which is causing it?
<metasyntactic>
but if I manually open it with luks and mount it, it works fine
<euank>
I was assuming you meant all nix-shell's were wrong, not just a specific default.nix's nix-shell
h0m1 has joined #nixos
ecfortes has quit [Remote host closed the connection]
<euank>
suzanne: is it all nix-shells (except --pure ones), or only ones using a specific default.nix?
<metasyntactic>
it applys the passphrase to all of the other drives, then when it gets to this drive it says waiting on udev for 10 seconds, then fails
<suzanne>
euank: -p busybox gives the correct HOME (regardless of --pure), and default.nix gives the wrong one (also regardless of --pure)
<euank>
Ah! So `default.nix` has a shell-hook that's doing the wrong thing!
<euank>
Can you share the default.nix?
<suzanne>
I already commented out all the shellHook, no difference… let me check with an empty default.nix
<euank>
suzanne: the other idea I have for a way to debug this would be adding `set -x` into `~/.bashrc` temporarily, and then seeing if the -x output happens to include HOME being said (and if we can sorta derive what code did it based on other stuff around it)
<metasyntactic>
energizer: any ideas?
<energizer>
metasyntactic: double check the uuid
<metasyntactic>
it's copy pasted
markus1199 has quit [Ping timeout: 256 seconds]
<metasyntactic>
I just reverified that they are identical
<suzanne>
euank: nearly there: it's one of the packages I installed via default.nix… a window manager apparently. Narrowing down on it :)
<euank>
Once you find the actual package, I'm kinda curious to see its setup hook / what it's doing
<suzanne>
welp. cue facepalm ^^. Thanks a lot for the help! Should I do anything about it or is it an "if the package says so nothing to fix" situation?
dongcarl has joined #nixos
<euank>
I'm not actually sure. Maybe that package just isn't meant to be used in nix-shell at all.
orivej_ has joined #nixos
<metasyntactic>
energizer: any other ideas?
<euank>
It's definitely fixable, i.e. if we don't export HOME= but instead repalce the build step with `HOME=$TMPDIR make` (or whatever build tool it's using) to only set home for that one step
<suzanne>
euank: yeah I just threw it in with all my packages but I think it's just a set of libs,
<euank>
But I don't know if it's "wrong" for a derivation to do that.
sheeldotme has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<suzanne>
but not quite the kind of thing you'd normally want in your $PATH, I suspect.
<euank>
does mkDerivation have a way to set environment variables other than setupHooks?
<gchristensen>
every attribute you set in mkDerivation is an env var
<euank>
if there's a cleaner way to add an environment variable only for the scope of that derivation's build, that would be much cleaner
orivej has quit [Ping timeout: 256 seconds]
<euank>
... Yeah, right, ofc. It could maybe just have HOME="$TMPDIR" as an attribute and not pollute nix-shell with a setupHook
<metasyntactic>
if I wanted to run a shell script at boot, what's the best way to do that?
<metasyntactic>
create a systemd thing?
<{^_^}>
[nixpkgs] @marsam opened pull request #90560 → rage: install manpages and completions → https://git.io/Jf5fW
<cransom>
yes.
<euank>
I guess if it put it in, say, `preBuild`, rather than `setupHooks`, that wouldn't hit `nix-shell` but would probably do the right thing?
<metasyntactic>
I'm thinking I should not be trying to decrypt these drives at init time, but I want to do so before services like docker containers run
<hodapp>
ugh. anyone familiar with this "javax.net.ssl.SSLKeyException: RSA premaster secret error"? I'm reading at ibm.com that due to export policy on crypto you have to put some US_export_policy.jar and local_policy.jar into the JRE directory, or something. This is the first time I've ever dealt with anything like this though
<hodapp>
of course this is also the only place that I use Oracle JDK...
<ldlework>
energizer: when i do what you did, then i add poetry2nix to home.packages it says that: The option value `home.packages.[definition 33-entry 1]' in `/nixcfg/modules/home/linux/poetry2nix/default.nix' is not of type `package'.
<ldlework>
is that what actually works for you?
<energizer>
ldlework: i have that in my overlays, yes
<ldlework>
I'm not exactly sure what I'm supposed to do
aljce has joined #nixos
Zer000 has joined #nixos
<aljce>
I would like to add my gpg keys to my nixos iso image im building but if I do that within nix my keys will end up in the nix store. how do people normally get their gpg keys to a new machine?
<Zer000>
error: packages '/nix/store/izgbi2dig2kpc7h447npa5wrvimz6yz1-dwarf-fortress-0.44.12/bin/dwarf-fortress' and '/nix/store/6wf1lbma8wsj4h0rrdyzp839pzpabp92-dwarf-fortress-full/bin/dwarf-fortress' have the same priority 5; use 'nix-env --set-flag priority NUMBER INSTALLED_PKGNAME' to change the priority of one of the conflicting packages (0 being the highest priority)
<ldlework>
But when I build it with nix-build, I get: error: attribute 'func_timeout' missing, at /nix/store/60x4lidqkfykym33lzi3xxs5svrbfk4n-source/default.nix:214:21
sheeldotme has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
suzanne_ has quit [Ping timeout: 240 seconds]
sheeldotme has joined #nixos
orivej has joined #nixos
orivej_ has quit [Ping timeout: 246 seconds]
ddellacosta has quit [Ping timeout: 256 seconds]
orivej has quit [Ping timeout: 240 seconds]
waleee-cl has quit [Quit: Connection closed for inactivity]
xensky_ has quit [Ping timeout: 246 seconds]
<MtotheM>
setting up development enviornments seems a lot more involved than packages
<simpson>
To hack on a local source directory for things already in nixpkgs, it's possible to just open up a nix-shell directly onto the source directory, but using the nixpkgs derivation.
user_0x58 has joined #nixos
<simpson>
Similarly, it's pretty common for leaf applications which build with Nix but *don't* have nixpkgs expressions to instead ship a default.nix for building. Some repos may even have a shell.nix for hacking.
xensky_ has joined #nixos
<MtotheM>
yeah.. I don't quite understand how this is supposed to work yet.
<pjt_014>
every time I un-sleep/hibernate I get assigned a new IPv6 ULA. Is this a thing that configuring/failing to configure in my config file would cause?
<pjt_014>
what's weirder is that it lists them as 'global' but fded:.. is not in the globally routable range
cript0nauta has quit [Ping timeout: 265 seconds]
Rusty1 has quit [Quit: WeeChat 2.3]
zupo_ has joined #nixos
endformationage has quit [Quit: WeeChat 2.6]
hoverbear has joined #nixos
<hoverbear>
Any tips for getting firefox to not hard crash every time it tries to open a save dialog? Persisted between multiple installes on unstable nixpkgs for at least the last month. :( Seems to be related to gsettings schemas?
zupo has quit [Ping timeout: 260 seconds]
<hoverbear>
Oh -- what... `firefox` works but devedition doesn't. Ok, that's good enough.
hoverbear has quit [Client Quit]
justanotheruser has quit [Ping timeout: 256 seconds]
knupfer has joined #nixos
nschoe has joined #nixos
knupfer has quit [Ping timeout: 260 seconds]
strickinato has joined #nixos
nschoe has quit [Ping timeout: 272 seconds]
<strickinato>
Question. Right now, I have hledger installed on my machine through nix. Running `hledger --version`, I see that it's `1.17.1.1`. I know version 1.18 has been released, and I'd like to upgrade. I started poking around the nixpkgs repo, and I actually don't understand at all how it got to be version `1.17.1.1`. There doesn't seem to be a version set anywhere.
<strickinato>
I'm curious how it works
sarcasticadmin has quit [Ping timeout: 260 seconds]
palo1 has joined #nixos
<lassulus>
there is this giant autogenerated file pkgs/development/haskell-modules/hackage-packages.nix
<cole-h>
strickinato: Haskell packages are kinda special. ~every week, peti updates the Haskell package set from the latest Stackage LTS. If you're on 20.03, you won't see 1.18 until 20.09 is released sometime after September.
<cole-h>
(And after you update to 20.09 of course)
palo has quit [Ping timeout: 264 seconds]
palo1 is now known as palo
<{^_^}>
[nixpkgs] @sheeldotme opened pull request #90579 → coc-nvim: add patch to hardcode node dependency → https://git.io/Jf5L3
cjpbirkbeck has quit [Quit: cjpbirkbeck]
<cole-h>
strickinato: So, it appears that it should be available after Friday (when Stackage LTS 16.0 will likely be merged, which is the Stackage snapshot with hledger 1.18) on nixpkgs master
<{^_^}>
#90032 (by peti, 5 days ago, open): Update Haskell package set to LTS 16.0 (plus other fixes)
<strickinato>
Interesting... Indeed it's in there! It seems that the contents of `pkgs/development/haskell-modules/hackage-packages.nix` are not indexed (perhaps because it's too big?) it didn't come up in the search results, but I do see the version bump in that diff!
<infinisil>
Note that with Nix you don't want to install compilers, but rather just have a nix-shell environment with the compiler available
<n000oob>
@infinisil: thanks a lot!
NeoCron has quit [Quit: Leaving]
orivej has quit [Ping timeout: 264 seconds]
NeoCron has joined #nixos
orivej has joined #nixos
inf has joined #nixos
eoli3n has quit [Remote host closed the connection]
n000oob has quit [Remote host closed the connection]
dkjii has quit [Quit: Lost terminal]
<{^_^}>
[nixpkgs] @FRidh merged pull request #89624 → doc: `pip install` flags are specified with `pipInstallFlags` → https://git.io/Jf1dX
<{^_^}>
[nixpkgs] @FRidh pushed commit from @danieldk to master « doc: `pip install` flags are specified with `pipInstallFlags` »: https://git.io/Jf5O0
<wedens[m]>
how can I make a derivation from some directory so that permissions and ownership are preserved? It'll be used for deployment as part of larger closure. is tar'ing this directory my best bet?
<jtojnar>
does anyone have `runPythonCommand` trivial builder?
<dminuoso>
wedens[m]: How do you preserve ownership... ?
pingiun has joined #nixos
<dminuoso>
wedens[m]: Recall, nix doesn't even know about users.
<dminuoso>
And the user/userid might not even exist on the remote machine
<wedens[m]>
I known that it will exist
<wedens[m]>
ownerhip is the least problem. I can always do `chown -R` permissions and dates are more problemactic
_rvl has quit [Ping timeout: 244 seconds]
<{^_^}>
[patchelf] @domenkozar pushed 2 commits to ci-musl: https://git.io/Jf538
<dminuoso>
wedens[m]: Well, ownership is a problem you can address in a nixos module.
<dminuoso>
Permissions should be possible, but I cant find it
<wedens[m]>
from the doc: After the build, Nix sets the last-modified timestamp on all files in the build result to 1 (00:00:01 1/1/1970 UTC), sets the group to the default group, and sets the mode of the file to 0444 or 0555 (i.e., read-only, with execute permission enabled if the file was originally executable). Note that possible setuid and setgid bits are cleared. Setuid and setgid programs are not currently supported by Nix. This
<wedens[m]>
is because the Nix archives used in deployment have no concept of ownership information, and because it makes the build result dependent on the user performing the build.
Chiliparrot has joined #nixos
nostrooo has quit [Read error: Connection reset by peer]
<dminuoso>
wedens[m]: Well you can definitely run chmod in the installPhase - you need this at the very least to make things executable.
<dminuoso>
Otherwise you couldn't have binaries in the nix store
karetsu has quit [Quit: WeeChat 2.6]
<dminuoso>
(Well strictly speaking you could.. but that's not whats happening)
neheist2 has quit [Read error: Connection reset by peer]
thc202 has joined #nixos
orivej has quit [Quit: No Ping reply in 180 seconds.]
neheist2 has joined #nixos
<raboof>
yay the gnutls update reached nixos-unstable-small, feeling safer again :D
m0rphism has joined #nixos
<NobbZ[m]>
<dminuoso "wedens: Well you can definitely "> He cited that the executable bit will be preserved but and otherwise read is always set, while write is always removed
<dminuoso>
Oh.
<dminuoso>
I missed that, sorry.
orivej has joined #nixos
<dminuoso>
wedens[m]: I suppose nixos is what you are after then. :)
<wedens[m]>
I don't see how nixos modules are related to that. It's just me using Nix as a fancy way to do rsync-like things :D
<dminuoso>
wedens[m]: nixos lets you talk about users and permissions.
<dminuoso>
nix is limited to whats happening inside the store, nixos lets you talk about the entire system.
EatThem has joined #nixos
EatThem has quit [Client Quit]
<dminuoso>
(So one common trick is to have some systemd unit with a preStart that sets up permissions and ownership)
<clever>
environment.etc can also automate that
<srk>
or systemd.tmpfiles
<wedens[m]>
the point is that I need to preserve original permissions. not manually re-setting them after
<dminuoso>
wedens[m]: The only hack I can see, is stuffing it inside a tar archive, and have an nixos module unpack that tar.. but you're pretty much breaking the determinism guarantees.
drewc has quit [Ping timeout: 246 seconds]
kreyren has quit [Remote host closed the connection]
<dminuoso>
Re-setting the permissions in a nixos module seems like the saner way
<wedens[m]>
yeah, that's what I was about to do, but I thought maybe there is semi-legal some option to tell Nix to preserve permissions
kreyren has joined #nixos
mananamenos has joined #nixos
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos
kleisli has quit [Ping timeout: 265 seconds]
<wedens[m]>
how do I make a derivation from private gitlab repo?
neheist2 has quit [Read error: Connection reset by peer]
<wedens[m]>
ideally, including .git directory
lopsided98 has quit [Quit: No Ping reply in 180 seconds.]
<NobbZ[m]>
But you can't untar into the store, as permissions will be adjusted then...
<clever>
wedens[m]: you cant keep the .git with private repos
<clever>
energizer: try setting startAt on the service instead, and it will auto-generate a timer for you
cosimone has joined #nixos
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
<wedens[m]>
clever: is there an example of creating luks-encrypted usb? as we don't have declarative partitioning (yet?)
orivej has joined #nixos
<clever>
wedens[m]: if you want legacy+gpt (best compatability), make gpt with 3 partitions, 1mb bios boot partition, 512mb fat32 efi system, rest for root
<clever>
wedens[m]: use cryptsetup to format and open the root, then mkfs.ext4 the block device cryptsetup created (check lsblk), and then mount everything under /mnt/ as normal
<clever>
wedens[m]: nixos-generate-config --root /mnt will auto-detect the luks
<wedens[m]>
clever: so, I just partition usb directly like a regular disk, without making iso
<clever>
yep
<clever>
the iso will be more readonly, so you cant easily add secrets
<dminuoso>
wedens[m]: Hold on a second. You are *that* paranoid, but don't care that the secrets end up in the *world* *readable* nix store?
<dminuoso>
Are you sure you spend a thought on this?
<energizer>
clever: https://bpa.st/DONA startAt works, OnCalendar doesnt. startAt docs say "This is equivalent to adding a corresponding timer unit with OnCalendar set to the value given here."
<wedens[m]>
dminuoso: I just don't trust 3rd party git hostings :D
<clever>
energizer: yeah, thats the bit that was missing, but startAt feels cleaner
<energizer>
dminuoso: secrets dont have to end up in the store
cosimone has quit [Quit: Quit.]
__monty__ has joined #nixos
<dminuoso>
energizer: Ah story of my life. With nixos/nixops its very hard and annoying to avoid that. :(
<dminuoso>
I decided that it's not a big problem since I can just spawn dedicated machines with no human access.
<energizer>
yes, "world readable" is an exaggeration in most circumstances
<wedens[m]>
dminuoso: I use private repo for dotfiles because I don't want to share email addresses, names, some private infrastructure details, etc. and I don't want to bother encrypting everything. but some (small) amount of data I don't feel like sharing with git hosting provider (like vpn configs), so I encrypt them.
* dminuoso
has a synology storage at home
<dminuoso>
Im not sure why people use git for everything
ArdaXi has quit [Ping timeout: 256 seconds]
<{^_^}>
[nixpkgs] @danbst closed pull request #75813 → [RFC] mkDerivation: support version suffix, which is part of derivation nam… → https://git.io/Je7zq
<mananamenos>
hi, I've installed direnv from nixos packages, created .envrc file in my project's root directory, but it doesn't work. What am i missing?
<eyJhb>
Nope, etu just gpg :p - Thanks betaboon :D Just did the cherry-pick etc.. But I could have done the overlay as well. Only need it till friday :p
<etu>
mananamenos: Have you enabled it in your shell config?
<SandwichChef>
Anyways, the reason that I'm here is because I've been trying to get a package imgbrd-grabber to build with help from a few others. It seems the issue is that we can't get "GL/gl.h" to be used from mesa
<SandwichChef>
When we use libglvnd it causes segfaults later even though the build succeeds
<SandwichChef>
Does anyone have any advice or insight into why this might be?
<mananamenos>
etu, should i just add `eval "$(direnv hook bash)"` to programs.tmux.extraConfig = builtins.readFile ./tmux.conf;
<etu>
mananamenos: Shouldn't be needed, if you shell evals it it will eval it in tmux as well.
orivej has quit [Ping timeout: 256 seconds]
<karetsu>
something is very wrong with my dbus (e.g. playerctl -l doesn't list spotify) but dbus is active so its a user issue, where would I start looking for problems with this on nixos?
orivej has joined #nixos
<mananamenos>
it doesn't. In xfce4-terminal it does the job, however in tmux it doesn't
<mananamenos>
etu, just tried putting the line expression in my tmux.conf but didn't help either (or maybe i need a restart for this..)
<mananamenos>
etu, yes, i've put `eval "$(direnv hook bash)"` in my .bashrc, which made xfce4-terminal/bash work. Dropping into tmux however stops direvn working
zupo_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
bennofs[m] has joined #nixos
<bennofs[m]>
i am a bit confused, there's both networking.supplicant and networking.wireless. Both of them seem to configure wpa_supplicant, which one should I use?
<gmr>
(i don't actually know who moderates this room, emily Ross Schulman
<{^_^}>
[nixpkgs] @peti pushed 493 commits to haskell-updates: https://git.io/Jf541
<rschulman>
Sure as hell isn't me.
<gchristensen>
gmr: what for?
<rschulman>
spam in the backscroll
<gchristensen>
where? I don't see it
dingenskirchen has quit [Quit: dingenskirchen]
dingenskirchen has joined #nixos
<simpson>
Is it on the other side of some bridge, perhaps?
<simpson>
(I cannot see it either, but I notice that gmr's connected through matrix.org)
<gmr>
lol sorry i'll stop i don't actually care about a little spam as long as it gets cleaned up before my daily speedread of what i'd missed. this room moves very fast especially with the bots. double edged sword :S (this is a gripe but i'm trying to be nice about it @mods)
<gchristensen>
on IRC, you can't remove messages already sent
<simpson>
You could filter notices from {^_^} if you don't want those updates. I think some folks already do that. Indeed you could filter notices entirely on this channel, since they are basically only used by {^_^}.
<jtojnar>
Yeah, it was linuxuser981:matrix.org, weird that it did not go through through the irc bridge
drdee has quit [Quit: Leaving]
munksgaard has joined #nixos
<clever>
jtojnar: not seeing any trace of that name on the irc end
<jtojnar>
it was a potty-mouth so perhaps the bridge filtered it out
<{^_^}>
[nix] @edolstra pushed 2 commits to store-path-cxx: https://git.io/Jf5BI
<munksgaard>
How can I overwrite the source and version for a particular package in my configuration.nix? nixpkgs currently has sway 1.4 (the newest release), but master has some fixes that I need for a multi-monitor setup. I have the feeling that I should be able to override the version and sha defined in https://github.com/NixOS/nixpkgs/blob/master/pkgs/applications/window-managers/sway/default.nix, but I don't know how.
<munksgaard>
Can anyone point me in the direction of some documentation on how to do that kind of thing?
<gmr>
simpson: the problem with filtering notices on this channel is that i don't want to filter non-bot notices. is there a reason why {^_^} uses notices?
<clever>
munksgaard: just use .overrideAttrs to change the version= and src=
<gmr>
Jan Tojnar: irc bridge is shadowbanning wut??? really??? who runs it?
<jtojnar>
gmr: Matrix.org runs it, not sure if it is shadowbanning or just a bug
<qyliss>
gmr: notices are mostly used by bots so that they're distinct from normal messages
<clever>
gmr: i dont like the bridge, it leads to confusion when a user is showing as jtojnar to some, and Jan Tojnar to others, and its not always obvious the 2 are the same person
<clever>
gmr: it also spams hard every time it goes down, lol
<gmr>
ah that makes sense ya - is there a good spot to get up to speed on bridges Jan Tojnar (or anyone else who knows)?
<clever>
munksgaard: anywhere within the main { ... } set
<munksgaard>
clever: All right, thank you.
<jtojnar>
gmr: what do you mean?
bennofs has joined #nixos
bennofs has quit [Client Quit]
<{^_^}>
[nix] @edolstra opened pull request #3702 → Rewrite StorePath class in C++ → https://git.io/Jf5BB
bennofs has joined #nixos
<jtojnar>
gmr: do you want to run your own irc <-> Matrix bridge?
<gmr>
yes
<eyJhb>
Is it possible to get all the requirements of a package to run? E.g. all the requirements to firefox, so I could build a minimal store from it?
<eyJhb>
gchristensen: hoping you have some good input from the Docker thing :D
<jtojnar>
gmr: see #irc:matrix.org
<{^_^}>
[nixpkgs] @maralorn opened pull request #90604 → nixos/systemd: Update warning for restarting oneshots → https://git.io/Jf5Bw
<gchristensen>
nix-env --query --requisites eyJhb
CMCDragonkai1 has joined #nixos
<{^_^}>
[nix] @gilligan opened pull request #3703 → Add tests for libfetchers/attrs.hh → https://git.io/Jf5By
leah2 has quit [Ping timeout: 258 seconds]
<eyJhb>
gchristensen: anything from within, so I can use it in a expression?
<eyJhb>
within -> function in nixpkgs
<gchristensen>
read how dockerToolsbuildImage works?
<eyJhb>
Will try again, so far I con't quite see how
orivej has quit [Read error: Connection reset by peer]
jeregrine_ is now known as jeregrine
orivej has joined #nixos
SandwichChef has quit [Ping timeout: 260 seconds]
Cache has joined #nixos
suzanne_ has joined #nixos
<munksgaard>
clever: So this is what I have now: http://paste.debian.net/1152344/ but when it doesn't seem like it's picking up the new version of the package? I've even pointed the overlay to my own fork where I've written stupid stuff in the `--version` output of sway, but it seems like I'm still using the old version of sway. Am I doing something wrong?
<tnks_>
gchristensen: I've been thinking about the argument that `nix` is still experimental. But later on, it occurred to me that it's officially how we get a repl (with `nix repl`), right? So this argument now feels inadequate. One benefit of documentation is that if it's hard to write, it's a good sign that the tool has a bad UX, so it seems like a mistake no to write documentation, even for new tools.
orivej_ has joined #nixos
orivej has quit [Read error: Connection reset by peer]
<gchristensen>
tnks_: definitely 1,000% agreed
<gchristensen>
tnks_: IMO we should write documentation as the is built, to litmus test "is it hard to document?"
drdee has joined #nixos
<tnks_>
gchristensen: I wouldn't mind trying to help write the documentation, but I do wonder who's taken ownership of the design of `nix` as a tool. Does it still make sense that `nix-env` functionality is separate? When I first saw `nix`, I thought it would be monolithic, with even a `nix install` and `nix uninstall`.
<gchristensen>
tnks_: ah. I think the part there is that nix has no nix-env equivalent, because we don't know what it should look like
<tnks_>
gchristensen: is it because we don't know what it should look like? Or because there's philosophical resistance to a monolith of that scope?
<gchristensen>
I don't think there is anything philisophical like that going on
<gchristensen>
tnks_: I'd love to get your help on this ... one sec
<gchristensen>
gilligan: you 'round?
<tnks_>
Either way, I think we can start with documenting what we have. And by that, I mean... make a man page... and update the official docs.
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<tnks_>
As for a next step after getting docs in place, I think it would be healthy for the Nix core contributors to think more about what it means for a feature to be experimental for too long. When is `nix` not experimental? I'm kind of thinking about Hackage packages where things are marked as "Unstable" left and right, but the whole community is numb to it and uses them all anyway.
suzanne_ has quit [Ping timeout: 246 seconds]
suzanne has joined #nixos
suzanne has quit [Client Quit]
<gchristensen>
tnks_: fwiw you're asking all the right questions, and I'm glad for it
<betaboon>
tnks_++
<{^_^}>
tnks_'s karma got increased to 3
<tnks_>
Well, I can write documentation, and then I'll have more skin in the game, because the next question will be whether to remove this "experimental" warning or not. How do we explain it in the documentation?
cfricke has joined #nixos
<tnks_>
As a point of reference, we have the experience of Debian, which took a /really/ long time to get out `apt` in its current state. And for a while, I new plenty of people who were frustrated by the weird array of tools you needed to know about... and there was even `aptitude` for a while, because the problem was unresolved.
<typetetris>
I try to override the `cookie` haskell packages using `callCabal2nix` but it yields an infinite recursion. I don't understand that, because I call `super.callCabal2nix` I thought it would then be a layering instead of a fix point search. Example nix expression: https://gist.github.com/typetetris/696a13f026d425d68af058d40755efc4
eoli3n has quit [Ping timeout: 246 seconds]
<{^_^}>
[nix] @Ericson2314 opened pull request #3704 → WIP: Add another missing #include → https://git.io/Jf5zD
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
<clever>
typetetris: i think that can happen when callCabal2nix is using pkgs.cabal2nix or something like that
<clever>
typetetris: you can sometimes work around it by doing `cabal2nix = super.cabal2nix.override { cookie = super.cookie; };` to force it to use the old one
rogue_koder has quit [Remote host closed the connection]
rogue_koder has joined #nixos
mog has joined #nixos
{^_^} has quit [Ping timeout: 256 seconds]
erasmas has joined #nixos
proofofkeags has joined #nixos
<gustavderdrache>
how do i resize the qcow image built by nixos-generators? i set system.build.qcow.diskSize = 21474836480; in configuration.nix, but i'm still seeing a / of 430MB
ym555 has quit [Quit: leaving...]
<clever>
gustavderdrache: did you delete the old qcow file before booting it?
FRidh has quit [Remote host closed the connection]
<gustavderdrache>
pretty sure i did, but let me try again to be sure
FRidh has joined #nixos
<gustavderdrache>
yeah, i did 'rm nixos.qcow2'
<clever>
if you read the script in result, what does it do?
drakonis_ has joined #nixos
<gustavderdrache>
hmm... i don't have a result link
<gustavderdrache>
here's a snip of my command line:
<__monty__>
I have a bunch of attributes and I want to reach into them `pkgs.nameX.components.exes.nameX`, I have the following list `with pkgs; [ nameX nameY nameZ ]`. Is there a function I can map over that list to get at those similarly named nested attributes? `map (attr: getAttrFromPath [ "components" "exes" "${name attr}" ])` is the closest I can come up with but I'm lacking the "name" function. Ideally I'd
<__monty__>
avoid having to pass in strings.
<clever>
gustavderdrache: and you cant use the existing `nixos-rebuild build-vm` stuff?
<gustavderdrache>
d'oh, i should've specified
<gustavderdrache>
i'm on an ubuntu box just using nixpkgs
<clever>
__monty__: i have something similar
<clever>
gustavderdrache: you can still access that, nix-build '<nixpkgs/nixos>' -A vm -I nixos-config=configuration.nix
<__monty__>
clever: That's very close but missing the final attribute selection, right? I'm not sure whether I care but I'd like to try with the slightly more restricted version first.
orivej has joined #nixos
dingenskirchen has quit [Remote host closed the connection]
dingenskirchen has joined #nixos
drakonis has joined #nixos
karetsu has quit [Quit: WeeChat 2.6]
orivej_ has quit [Ping timeout: 264 seconds]
<gustavderdrache>
clever: i did that and now i'm really confused... i'm seeing a / of 430M like i got with nixos-generate
<cole-h>
Isn't qcow sparsely allocated? Meaning it'll grow as needed? Or am I misremembering
<clever>
cole-h: the file on the host is sparse and tiny, but the size reported within the guest will remain the same
<clever>
gustavderdrache: the `-A vm` script uses virtualisation.diskImage, and you can override it with $NIX_DISK_IMAGE when being launched
<__monty__>
I keep getting this warning: `trace: WARNING: license "GPL-2.0-or-later AND BSD-3-Clause" not found`. First thought it was cabal but seems it's nix. Anyone know where this is coming from?
softinio[m] has quit [Quit: Idle for 30+ days]
vmallela[m] has quit [Quit: Idle for 30+ days]
GuilhemArvilMari has quit [Quit: Idle for 30+ days]
cejota[m] has quit [Quit: Idle for 30+ days]
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
exprez135[m] has quit [Quit: Idle for 30+ days]
cornu has quit [Quit: Idle for 30+ days]
cellofhuman[m] has quit [Quit: Idle for 30+ days]
kazzimazzi[m] has quit [Quit: Idle for 30+ days]
alfborge[m] has quit [Quit: Idle for 30+ days]
jjwatt[m] has quit [Quit: Idle for 30+ days]
bf0bdaf8-ad[m] has quit [Quit: Idle for 30+ days]
Arian[m] has quit [Quit: Idle for 30+ days]
ggpeti[m] has quit [Quit: Idle for 30+ days]
dot-product[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
ayrymmnwpnng[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
nbp[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
fiadliel[m] has quit [Quit: Idle for 30+ days]
M776f6c66[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
orivej has quit [Ping timeout: 246 seconds]
zupo has joined #nixos
atemu12[m] has quit [Quit: Idle for 30+ days]
loyon has quit [Quit: Idle for 30+ days]
icyhate5533[m] has quit [Quit: Idle for 30+ days]
jongeOld[m] has quit [Quit: Idle for 30+ days]
akymaky[m] has quit [Quit: Idle for 30+ days]
Jonas[m]2 has quit [Quit: Idle for 30+ days]
tamwile[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle for 30+ days"]
Chiliparrot has joined #nixos
daniels[m] has quit [Quit: Idle for 30+ days]
dtz has quit [Quit: Idle for 30+ days]
tom[m]2 has quit [Quit: Idle for 30+ days]
<Izorkin>
Mic92: aanderse: Who can merge this PR #87833 ?
rogue_koder has quit [Remote host closed the connection]
_ris has joined #nixos
muslimm has joined #nixos
rogue_koder has joined #nixos
* muslimm
If U believe in the only creator of heaven & earth(the one) that moses believed in him and Jesus(monothiesm) u will be safe * islam is your cave of wonders; jewel of THE SEAS. For he is: (Allah) the only God; our refuge; follow islam The true religion of God it's our way to saftiness, Time and life is in god,s Hand - Be muslims to be fine ****
muslimm has quit [Killed (Sigyn (Stay safe off irc))]
<eyJhb>
Thanks gchristensen :) ClosureInfo gives me all the store paths I need!
<eyJhb>
gchristensen++
<{^_^}>
gchristensen's karma got increased to 318
<gchristensen>
cool!
ris has quit [Ping timeout: 246 seconds]
fendor has quit [Remote host closed the connection]
pr06lefs has quit [Remote host closed the connection]
pr06lefs has joined #nixos
sigmundv_ has quit [Ping timeout: 260 seconds]
<Thra11>
When two unrelated things have the same name, do we generally take a first-come-first-served approach? i.e. the first one to be added to nixpkgs gets the name, and the second one has to have a prefix/suffix added?
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
asheshambasta has quit [Ping timeout: 272 seconds]
<simpson>
Hm. There are other possible rules, like "most popular", or "both get a prefix", but `helm` and `kubernetes-helm` don't follow either of those rules. So perhaps you're right. What packages are you working on?
<{^_^}>
[nixpkgs] @Ma27 pushed commit from @hlolli to release-20.03 « strace-graph: fix strace-graph shebang which points to perl »: https://git.io/Jf56m
<Thra11>
simpson: I'm just reviewing PR #85422. I assume the most popular will tend to be pacakged first, so first-come-first-served might tend to look like most-popuplar-takes-precendence.
<simpson>
Huh. Two different Gemini clients, both named "Castor"? Sounds like people read the same Greek stories. I'm not sure how to well-disambiguate them; a good example to look at might be `libtorrent`, which could refer to any of like three different libraries.
<drakonis1>
is there anything to make it easier to bump the flake lock?
<nh2>
drakonis1: if one of many parallely building derivations fails, and nix said which one fails at the end, run `nix log` on that `.drv` to get nice non-interleaved output
<drakonis1>
i'm tired of having to manually bump it every time i want to bump my system packages
seku has joined #nixos
nckx has joined #nixos
<jakobrs>
I have a suggestion: A symlink at /nix/var/nix/profiles/boot-default that points to the boot default configuration
<infinisil>
jakobrs: /run/booted-system?
<jakobrs>
As a motivating example, you could rewrite the boot loader entries after nix-collect-garbage -d using /nix/var/nix/profiles/boot-default/bin/switch-to-configuration boot
<jakobrs>
The boot default, not the booted system.
<jakobrs>
If you choose a different configuration at boot, or do nixos-rebuild boot at some point, it's going to be different from /run/booted-system
<adamse>
I want to package up a (generated) shell script and some auxiliary files, how do I best do this?
kalipso has joined #nixos
<adamse>
do I put a `cat ${generatedScript} > $out/bin/script-name` in the installphase?
<jakobrs>
You can just do `cp ${generatedScript} $out/bin/script-name`
<adamse>
if `generatedScript = '' ... '';` ?
<{^_^}>
[nixos-homepage] @garbas pushed 0 commits to fix-457: https://git.io/Jf5iG
<numkem>
I've made a custom module that define a user service, it used to work just fine but now even if I activate it or not it doesn't add the service to my user. What else could be there to prevent it? I'm on unstable
<clever>
adamse: if the string contains the raw script, then you probably want passAsFile
janneke_ is now known as janneke
<jakobrs>
adamse: If the only file your package generates is this script, `pkgs.writeScriptBin` might be better
<clever>
adamse: `foo = ''......''; passAsFile = [ "foo" ]; installPhase = "cp $fooPath $out/bin/foo";` if you want to have many files in $out
<clever>
or writeScriptBin if you only want the script
<adamse>
there will be at least 2 files so i'll take a look at passAsFile! thank you :)
<kalipso>
Hey. in https://github.com/NixOS/nixpkgs/pull/85254 a bug that caused wrong library paths when trying to find boost libraries with cmake that got fixed some weeks ago. now i did a channel update and nixos-rebuild switch. when i now do a nix-shell -p boost171 it still points to the old derivation, is there a way to "update" that? when i try to get a
<kalipso>
new boost package "nix-env -p boost172" it downloads a newer derivation that containes the fix, but not for the boost171 one.
<jakobrs>
"Bug" I just found: If you try to do `switch-to-configuration boot` on a configuration that is not a generation of one of the system profiles, the systemd bootloader config won't be written
eoli3n has joined #nixos
pjt_014 has joined #nixos
<jakobrs>
Quotes around "bug" because I can't see this being a problem
<jakobrs>
Assuming `nixos-rebuild build` builds a new generation that is not already in one of the system profiles.
<adamse>
clever, jakobrs: thank you, I successfully wrote the script using passAsFile!
<jakobrs>
w
<jakobrs>
umm, wrong window
lsix has quit [Ping timeout: 272 seconds]
<monokrome>
,locate mbsync
<{^_^}>
Found in packages: isync
kenji has quit [Ping timeout: 240 seconds]
<bigvalen>
So, I was trying to move my root from a ZFS filesystem on disks, to a new pool, on an nvme drive. Made up the config, got everything ready to go.../boot on a USB sdcard.
<bigvalen>
# nixos-install
<bigvalen>
[5 built]
<bigvalen>
building the configuration in /mnt/etc/nixos/configuration.nix...
<bigvalen>
copying channel...
<bigvalen>
installing the boot loader...
<bigvalen>
and it ... says nothing else.
kenran_ has joined #nixos
<bigvalen>
I think the last command the installer ran was "nixos-enter --root /mnt -- /run/current-system/bin/switch-to-configuration boot" - any idea why that might have hung ?
asheshambasta has quit [Ping timeout: 244 seconds]
<jakobrs>
$ ls /nix/var/nix/profiles
lsix has joined #nixos
kenran has quit [Ping timeout: 256 seconds]
<jakobrs>
boot-default per-container per-user system system-18-link system-profiles
<cransom>
the new /boot, do you mean /mnt/boot?
<simpson>
Thra11, fgaz: Ah, I misread. No idea, then. `castor-browser` doesn't sound bad. At the end of the day, I am a chicken, not a pig; I'm not in charge of these names.
drdee has quit [Remote host closed the connection]
<redcedar[m]>
does anyone have any advice for troubleshooting a upnp device not showing up (in file manager or vlc)? I have a usb drive on my router which is easily discoverable by both a ps4 and vlc on my phone, but neither of my nixos machines discover it.
<{^_^}>
[nixpkgs] @Ma27 pushed commit from @Luflosi to release-20.03 « youtube-dl: 2020.06.06 -> 2020.06.16.1 »: https://git.io/Jf5Pf
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
kleisli has joined #nixos
proofofkeags has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @veprbl pushed commit from @marius851000 to master « openmw: 0.45.0 -> 0.46.0 (#90589) »: https://git.io/Jf5Pt
<jakobrs>
If I implement my suggestion, where should I create the boot-default symlink?
<jakobrs>
It feels wrong to put it in /nix/var/nix/profiles because it's not actually a profile
proofofkeags has quit [Remote host closed the connection]
proofofkeags has joined #nixos
kenran has joined #nixos
lucc has joined #nixos
kenran_ has quit [Ping timeout: 256 seconds]
aveltras_ has quit [Quit: Connection closed for inactivity]
<Elorm[m]>
Anyone know why the nixos subreddit has been restricted please?
<jakobrs>
It seems there hasn't been a single post in a month
gentauro has quit [Read error: Connection reset by peer]
<jakobrs>
maybe it was spammed
gentauro has joined #nixos
drdee has joined #nixos
amerigo has joined #nixos
jakobrs has left #nixos ["WeeChat 2.7.1"]
<numkem>
I've tried pretty much everything and I can't figure out why a definition of `systemd.user.services."name" {}` isn't showing up in `systemctl --user status "name"`
ruuda has joined #nixos
<numkem>
it works on another machine and it's the exact same code
<bqv>
add that line into nix.conf and persist it by adding all three to your config
<bqv>
you won't be able to enable it just for one particular environment
<bqv>
but there's no harm in doing it globally
<ingenieroariel>
could I replace nix in the nix shell with the one in nixFlake?
<bqv>
sounds like a good start
<ingenieroariel>
I'll try something like that - I want end users of this to be able to git clone and nix-shell
<bqv>
flakes are still kinda WIP, it can't be that easy right now
muslimm has joined #nixos
sbetrium has joined #nixos
* muslimm
If U believe in the only creator of heaven & earth(the one) that moses believed in him and Jesus(monothiesm) u will be safe * islam is your cave of wonders; jewel of THE SEAS. For he is: (Allah) the only God; our refuge; follow islam The true religion of God it's our way to saftiness, Time and life is in god,s Hand - Be muslims to be fine ****
muslimm has quit [Killed (Sigyn (Stay safe off irc))]
pamplemousse has joined #nixos
<bqv>
what on earth
<seku>
well, at least it didnt last long.
<{^_^}>
[nix] @edolstra merged pull request #3702 → Rewrite StorePath class in C++ → https://git.io/Jf5BB
<{^_^}>
[nixpkgs] @Ekleog merged pull request #87790 → krop: move libsForQt5.poppler out of propagatedBuildInputs → https://git.io/JfBJ6
<{^_^}>
[nixpkgs] @Ekleog pushed commit from @doronbehar to master « krop: move libsForQt5.poppler out of propagatedBuildInputs (#87790) »: https://git.io/Jf512
<__monty__>
Yep, thanks. energizer++ cole-h++
<{^_^}>
cole-h's karma got increased to 70
<{^_^}>
energizer's karma got increased to 20
<eyJhb>
Never really got sc-im to work great
<infinisil>
> randomPackage
<{^_^}>
"pkgs.nms: A command line tool that recreates the famous data decryption\neffect seen in the 1992 movie Sneakers.\n < https://github.com/bartobri/no-more-secrets >"
<cole-h>
\n
<{^_^}>
[nixpkgs] @jonringer opened pull request #90636 → python2Packages: default to dontUsePythonRecompileBytecode = true; → https://git.io/Jf51K
* evils
adds nms to a mental list of stuff for `hollywood`
pjt_014 has quit [Read error: Connection reset by peer]
pjt_014 has joined #nixos
<bqv>
how have you managed to get random data into a pure system
<bqv>
ok, i can vaguely figure out how, but the real question is why >_>
<gchristensen>
it doesn't actually
<gchristensen>
> :v randomPackage
<{^_^}>
randomPackage = let inherit (r randomPackage') name value; in "pkgs.${name}: ${builtins.replaceStrings ["\n"] [" "] value.meta.description} < ${value.meta.homepage} >"
<gchristensen>
> :v random
<{^_^}>
random = fnv1a32 builtins.currentTime
<bqv>
ah, ghetto random
<infinisil>
Good enough!
<gchristensen>
> vom "good enough for this!"
<{^_^}>
"good enough for this!"
<gchristensen>
oh. heh
lucc has quit [Quit: WeeChat 2.8]
* infinisil
votes for colors in #nixos
<gchristensen>
nah it makes spam even worse
ebopp has joined #nixos
pamplemousse has quit [Ping timeout: 260 seconds]
<bqv>
technicolor spam
<ingenieroariel>
I ended up caving and using: nix.package = pkgs.nixFlakes;
* bqv
smirks
<{^_^}>
[nix] @edolstra pushed to master « Remove StorePath::clone() and related functions »: https://git.io/Jf5Mv
lilpea has joined #nixos
gxt has quit [Quit: WeeChat 2.8]
davidv7 has quit [Ping timeout: 264 seconds]
noam has joined #nixos
noam has quit [Client Quit]
MtotheM1 has joined #nixos
lilpea has quit [Read error: Connection reset by peer]
<energizer>
i have nixos on an M.2 nvme ssd and i'm trying to boot my laptop. it works when i use the M.2 usb adapter, but when i plug the card directly into the M.2 slot in the laptop, stage 1 boot fails to find the encrypted device /dev/disk/by-uuid/... with the root partition on it
<energizer>
i have boot.initrd.availableKernelModules = [ "xhci_pci" "ahci" "usb_storage" "sd_mod" "rtsx_pci_sdmmc" "uas" "nvme" ];
<energizer>
the stage 1 interactive shell doesn't have bin/blkid and /dev/disk doesnt exist so i can't see what uuids are available
<energizer>
i figure there's some kernel module missing
lsix has quit [Ping timeout: 256 seconds]
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
cosimone has quit [Quit: Quit.]
orivej has joined #nixos
lilpea has quit [Read error: Connection reset by peer]
<glittershark>
timokau[m]: do you have a quick elevator pitch on what this gets me besides the existing github pull request review statuses + workflow?
<{^_^}>
[nixpkgs] @benley opened pull request #90638 → melonDS: init at 0.8.3 → https://git.io/Jf5DJ
<{^_^}>
[nixpkgs] @eliaslfox opened pull request #90639 → nixos/nvidia: Only install nvidia-persistenced when enabled as a service → https://git.io/Jf5DU
<glittershark>
timokau[m]: cool, knew a comment like that had to exist somewhere :)
<timokau[m]>
Basically if you have merge permissions and are looking for already reviewed PRs, searching github for approvals is not very useful. People use approvals to mean all kinds of things.
<glittershark>
ah sure, so this is to help the workflow of people with a commit bit
<timokau[m]>
glittershark: For now yes, mostly. But also improvements for unprivileged reviewers (hopefully much more rewarding if your reviews actually lead to a merge) and PR authors. Many extensions possible down the line.
dingenskirchen1 is now known as dingenskirchen
<glittershark>
timokau[m]: cool!
<timokau[m]>
evils: Yes! And PRs ready for review.
NinjaTrappeur has quit [Ping timeout: 256 seconds]
jhuizy has quit [Ping timeout: 256 seconds]
<energizer>
i see some people encrypt their data with a key file and then encrypt the key file with a password. what's the benefit of that over just encrypting the data with a password without the middle step?
jhuizy has joined #nixos
reanimus has joined #nixos
NinjaTrappeur has joined #nixos
jean-lopes has joined #nixos
mekeor has quit [Ping timeout: 256 seconds]
mekeor has joined #nixos
szicari has joined #nixos
aleph- has joined #nixos
fendor_ has joined #nixos
ericnoan has quit [Ping timeout: 264 seconds]
zgrep has joined #nixos
ebopp has quit [Ping timeout: 256 seconds]
ArdaXi has quit [Ping timeout: 256 seconds]
blindidiotgod has quit [Ping timeout: 256 seconds]
philipp[m]2 has quit [Ping timeout: 256 seconds]
blindidiotgod1 has joined #nixos
Philipp[m]21 has joined #nixos
rardiol has quit [Ping timeout: 246 seconds]
vykook has quit [Ping timeout: 246 seconds]
<manveru>
keys usually have more entropy that passwords
ebopp has joined #nixos
ArdaXi has joined #nixos
<energizer>
but the password is sufficient to get the key, so how does that help?
fendor_ has quit [Client Quit]
agsdheidjd has quit [Ping timeout: 256 seconds]
<manveru>
well, you usually try to protect the key more than the file you're encrypting :)
vykook has joined #nixos
<manveru>
the password is only there in case you somehow leak your key
<manveru>
energizer: just ask gchristensen directly then :)
<infinisil>
energizer: Usually the reason is that it adds another factor: You need to have the file and know the password to decrypt the data. Having one of them isn't eneugh
logan12358[m] has joined #nixos
emmanuelrosa[m] has joined #nixos
Notkea[m] has joined #nixos
chvp has joined #nixos
koschitzky[m]1 has joined #nixos
Luigi[m] has joined #nixos
Nacho[m]1 has joined #nixos
MawKKe[m] has joined #nixos
roberth has joined #nixos
CRTified[m] has joined #nixos
trufas[m] has joined #nixos
dxtr[m]1 has joined #nixos
sm[m] has joined #nixos
alienpirate5 has joined #nixos
mt[m] has joined #nixos
iwvt[m] has joined #nixos
anarchomoh[m] has joined #nixos
sibachian[m] has joined #nixos
bobfett1 has joined #nixos
NickHu1 has joined #nixos
Fulgen has joined #nixos
adit[m] has joined #nixos
ibinzari[m] has joined #nixos
musicmatze[m] has joined #nixos
chreekat[m] has joined #nixos
ky0ko has joined #nixos
fsmnarmosta[m] has joined #nixos
davidak[m] has joined #nixos
wangoe[m] has joined #nixos
gsals[m] has joined #nixos
freeman42x[m] has joined #nixos
Ericson2314 has joined #nixos
wak-work has joined #nixos
<infinisil>
And somebody that steals the key file doesn't know the password, and somebody that steals the password doesn't have the key file
<energizer>
infinisil: they're on the same disk
<manveru>
in this case you have to steal the laptop
<infinisil>
Then I don't think it adds anything extra
<gchristensen>
tbh I think I did it so I'd only have to type the password in once, or rotate the user-typed in only one place
<gchristensen>
but I wouldn't read any special sauce
shah^ has quit [Ping timeout: 264 seconds]
<energizer>
otherwise you'd have to type the password for each partition?
asymptotically has quit [Quit: Leaving]
<gchristensen>
you don't actually need to, no, nixos will try the typed password on all partitions
shah^ has joined #nixos
shah^ has quit [Max SendQ exceeded]
<gchristensen>
it may have been misguided from the start
<gchristensen>
a better bet would be using the laptop's TPM as part of the decryption process
<gchristensen>
(better bet than using a key partition)
<manveru>
man... that amazon-init service module is really annoying the hell out of me :|
<manveru>
really wanted to avoid having to make my own AMI again
lunik1 has quit [Ping timeout: 256 seconds]
lunik15 is now known as lunik1
q3k has quit [Remote host closed the connection]
`slikts has joined #nixos
`slikts has quit [Changing host]
`slikts has joined #nixos
q3k has joined #nixos
<manveru>
i feed a nixos configuration into amazon-init via user_data, which in turn spawns another systemd service that does a second nixos-rebuild so i can actually build a flake with more than one file...