<sterni>
gchristensen: inb4 something important for nix uses #!/usr/bin/env perl for some reason
ghuntley has quit [Remote host closed the connection]
taktoa[c] has quit [Remote host closed the connection]
srhb has quit [Remote host closed the connection]
manveru has quit [Remote host closed the connection]
jackdk has quit [Remote host closed the connection]
manveru has joined #nixos-chat
taktoa[c] has joined #nixos-chat
srhb has joined #nixos-chat
ghuntley has joined #nixos-chat
jackdk has joined #nixos-chat
<srhb>
Is there a tool like agenix that permits the use of an agent for encryption? I really like agenix for the ease of specifying who/what should be able to decrypt a secret, but not being able to use my own key for decryption is a bit of a show-stopper.
steveeJ has quit [Remote host closed the connection]
raboof has quit [Remote host closed the connection]
steveeJ has joined #nixos-chat
raboof has joined #nixos-chat
<gchristensen>
sterni: oh well :)
pie_86 has quit [Quit: Connection closed]
pie_38 has joined #nixos-chat
<srhb>
etu: The nice people at AMD whipped up a kernel patch to fix my monitor with 120Hz+ on kernel>5.4
<srhb>
The level of open source support compared to enterprise/consumer never ceases to amaze me.
<srhb>
It's like... If you pay for support, you know you're on your own in advance.
<sterni>
gchristensen: okay according to grepping /run/current-system/sw, there isn't
<etu>
srhb: Woo :)
<etu>
srhb: I haven't tried to run it in more than 120Hz, that's plenty fast for me :D
<srhb>
Ditto, but 120Hz was broken for me too, and 60Hz is just _too_ plebian! ;-)
<etu>
My work laptop can't do more than 60Hz anyways on this resolution :(
<etu>
And that's the one with thunderbolt
<etu>
It just works worse in every single way with my thunderbolt dock than my T495 that doesn't have thunderbolt
<etu>
(work laptop is T480s)
<srhb>
I've given up getting anything nice from my laptop on this monitor. It's just tearing galore. :P
<srhb>
But it's *fantastic* with my workstation :)
* etu
have spent hours to research a bug today, result hours later: Added a single space.
<andi->
gchristensen: are you sure it is worth the time investment?
<SuperSandro2000>
> Phel is a functional programming language that compiles to PHP
<{^_^}>
undefined variable 'Phel' at (string):477:1
<SuperSandro2000>
well
<andi->
Since this is for nixpkgs it should be written in bash. Bash is our main language. Surely we are good in that.
<leah2>
Shel
<andi->
Unfortunately we still have a bunch of Nix in the repo that we must replace at the same time.
veleiro has quit [Ping timeout: 244 seconds]
emily has quit [Ping timeout: 244 seconds]
colemickens has quit [Ping timeout: 244 seconds]
<cransom>
i was led to believe that people write bad bash, so instead we should be writin bad yaml thats digested by some also questionable python.
Dotz0cat has joined #nixos-chat
<cole-h>
gchristensen++
<{^_^}>
gchristensen's karma got increased to 429
<sterni>
keep it in php if the alternative is rewriting in bash
<sterni>
the world doesn't need more shell scripts
<sterni>
as shell programmers we have a collective record of …… deliberately not quoting variables because “it never contains a space anyways”
<andi->
Maybe we just haven't invented the right language for this problem just yet...
<andi->
github webhooks to rabbitmq is a very special problem.
<SuperSandro2000>
sterni: out does not need quoting
<SuperSandro2000>
pure shell scripts should listen to shellcheck
<sterni>
SuperSandro2000: you are contradicting yourself, shellcheck does precisely tell you to quote your variables
<sterni>
SuperSandro2000: and de facto not quoting $out leads to the impossibility to build nixpkgs with a nix store that has a space in its path
<sterni>
(which I guess doesn't matter that much)
<sterni>
writing portable shell scripts is also non-trivial
<sterni>
we just escape this issue because /bin/sh is always bash
<srk>
Phel, nice!
Dotz0cat has quit [Ping timeout: 256 seconds]
veleiro has joined #nixos-chat
<srk>
andi-: I've altered the problem to cron, pull, git post-receive and rabbitmq
<andi->
sterni: not always :D
<SuperSandro2000>
'/bin/sh being bash is a far bigger issue than making /nix non portable
emily has joined #nixos-chat
<sterni>
andi-: oh god oh no
colemickens has joined #nixos-chat
<andi->
sterni: on nixos: sandbox-paths = /bin/sh=/nix/store/6hzp1f155v38fl7j5zj1xms8w1c4vqz2-busybox-1.32.1-aarch64-unknown-linux-musl/bin/busybox
<sterni>
SuperSandro2000: well it is the same issue in the end, nixpkgs being full of non-optimal shell scripts which make too many assumptions about their environment
<andi->
that is inside of a build
Dotz0cat has joined #nixos-chat
<sterni>
andi-: oh crazy
<andi->
outside of the build env it is bash
<andi->
nice impurity (on a different scale) there
rajivr has quit [Quit: Connection closed for inactivity]
Dotz0cat_ has joined #nixos-chat
Dotz0cat has quit [Ping timeout: 265 seconds]
hodapp has quit [Ping timeout: 260 seconds]
hodapp has joined #nixos-chat
energizer has quit [*.net *.split]
spudly has quit [*.net *.split]
kraem has quit [*.net *.split]
ornxka_ has quit [*.net *.split]
jD91mZM2 has quit [*.net *.split]
matthewcroughan has quit [*.net *.split]
judson has quit [*.net *.split]
samueldr has quit [*.net *.split]
energizer has joined #nixos-chat
judson has joined #nixos-chat
jD91mZM2 has joined #nixos-chat
matthewcroughan has joined #nixos-chat
samueldr has joined #nixos-chat
ornxka_ has joined #nixos-chat
spudly has joined #nixos-chat
kraem has joined #nixos-chat
samueldr has quit [Max SendQ exceeded]
samueldr has joined #nixos-chat
<ekleog>
andi-: ohai :) ISTR you tweeting about some OSS web photo library thing, would you remember the name of it so I could have a look? (currently looking for something, ideally it'd also replace digikam, but if it can at least handle selective album sharing with the family, photo geoloc / date / people's names & stuff like that, it'd already be great :))
rj has quit [Quit: rj]
<hexa->
ekleog: photoprism
slack1256 has joined #nixos-chat
<philipp[m]1>
Has anybody here worked with easer on android before?
<philipp[m]1>
I'm wondering whether I can use it to toggle my wireguard vpns depending on what WiFi I'm connected to.
<hexa->
philipp[m]1: use tasker
<hexa->
yes, you have to buy it, but it works pretty well
<philipp[m]1>
I don't have gapps.
<philipp[m]1>
So no google play.
<hexa->
awh rip
<hexa->
i haven't found any other app to do this
<hexa->
it requires to enable external control in the wireguard app
BaughnLogBot has quit [Ping timeout: 240 seconds]
<hexa->
disable power saving for both wireguard and tasker
<hexa->
nd then create a rule
<hexa->
match the wlan network, if it moves away, vpn on, else off
<philipp[m]1>
I think easer can do it in theory with broadcasts/intents.
<hexa->
ah, ok
BaughnLogBot has joined #nixos-chat
<ekleog>
hexa-: neat, thank you :)
<__monty__>
Is bonding the interfaces an option? Or is always on wg not an option?
<philipp[m]1>
__monty__: All I want to do is toggle an additional wg network when I'm home so I don't have to do a roundtrip via my server when I want to stream something in my lan...
<ekleog>
hmm looks like it doesn't have face recognition yet (according to reddit), guess I'll stay on digikam a bit longer for now
cosimone has quit [Quit: cosimone]
Dotz0cat_ has quit [Ping timeout: 256 seconds]
Dotz0cat has joined #nixos-chat
Dotz0cat has quit [Ping timeout: 256 seconds]
Dotz0cat_ has joined #nixos-chat
<f0x>
ok the time has finally come #f0xnixos2021
* makefu
is disappointed, channel is empty
omnd has quit [Ping timeout: 240 seconds]
<f0x>
hm, yeah, hashtags on irc :P
<samueldr>
I'm angry at whoever coded this thing I'm looking at
<samueldr>
and I will not git blame it
<samueldr>
the risk is high that I am to blame
<sterni>
now you're making it exciting
<samueldr>
ugh!
<aleph->
hexa-: I had a really crummy C prog to do that on my Laptop
<samueldr>
why is it even legal to try running ioctls on a closed fd?
<samueldr>
or, why is the error message so bad
<samueldr>
>> Inappropriate ioctl for device
<aleph->
Reasonably if I'm not on my lan, then shell out and run wg.
<samueldr>
thanke -ENOTTY
<samueldr>
thanks*
<aleph->
samueldr: Oh boy, that's scratching at my brain...
<joepie91>
<samueldr> the risk is high that I am to blame
<aleph->
What's the issue exactly?
<joepie91>
well well well, if it isn't the consequences of my own actions...
<joepie91>
:P
<aleph->
Haha
<samueldr>
the issue is things that should have been working didn't!
<aleph->
joepie91: This JS codebase is hell
<samueldr>
because I was trying to write into a closed fd!
<aleph->
I am so glad this project is almost done
<aleph->
Maybe I can find time to refactor it into sanity then.
<samueldr>
weird how that once I fixed that small thing everything just works
<aleph->
Like using actual templates and not string buffers full of html with string templating...
<joepie91>
oh... oh no...
<joepie91>
why
<aleph->
joepie91: Because I inherited a god tier C-programmers quick node codebase
<aleph->
Because we needed a internal log in management service for work for our customer instances.
<aleph->
joepie91: The actual backend part is a mixture of promises, await, and callbacks. :/
<joepie91>
uuuuugh
<joepie91>
C JS is the worst
<joepie91>
with some distance
<joepie91>
it's actually worse than academic JS
<aleph->
It is. Most of it I can rewrite to just use async await fairly easily thankfully. The templating will be... interesting however.
<joepie91>
async/await would not be my choice, but promises, yes
<aleph->
joepie91: So we have a tenant_cache object/hashmap. We pass that to a renderHost() func that then generates a column, full of divs, tooltips, buttons, and a drop down.
<aleph->
And then we just append that to the string buffer.
<aleph->
:/
<aleph->
joepie91: Yeah I'm just used to C stuff too much. So async/await makes it easier for me to order stuff and quickly grok it. Otherwise I'd use promises. They're fine.
<joepie91>
aleph-: unfortunately async/await in JS runs the risk of race conditions, *especially* when working with code that mutates shared state, which this code will almost certainly do
<joepie91>
it looks synchronous, but isn't, and you still need to understand 100% of the semantics of promises to use it safely
<aleph->
Eyeppp.
<f0x>
hmm what would be the best way to try automatically boot into an older system config if the new one failed?
<aleph->
Thankfully there's only one func that mutates shared state, runs on a setInterval timer.
<f0x>
(as this is a totally headless system)
<joepie91>
aleph-: I somehow have difficulty believing that in C JS :)
<joepie91>
it almost always has shared state squirreled away somewhere...
<aleph->
joepie91: Nope, very small codebase thankfully. Everything else just reads from that single shared state.
<__monty__>
f0x: I think maybe infinisil's nixus can do that.