<infinisil>
Had to make sure the bot doesn't double-spam other channels
<samueldr>
now I see what you were up to
pie__ has joined #nixos-chat
pie___ has quit [Ping timeout: 250 seconds]
<ottidmes>
gchristensen: now I remember, I wanted to reply to your message "love downloading ~300 texlive NARs :D" at the time, but could not remember what the context, now I just did, when I do a `du` of the current system's closure, it feels like 70% is texlive...
<gchristensen>
haha
<ottidmes>
gchristensen: ah, it is actually 65% ... no wonder XD (grep -c texlive temp ~> 3200, grep -c /nix/store temp ~> 4905)
sir_guy_carleton has joined #nixos-chat
drakonis has joined #nixos-chat
drakonis_ has quit [Ping timeout: 252 seconds]
jasongrossman has quit [Ping timeout: 252 seconds]
lassulus_ has joined #nixos-chat
<infinisil>
,fancy-uninstall
lassulus has quit [Ping timeout: 250 seconds]
lassulus_ is now known as lassulus
<infinisil>
,fancy-uninstall
<{^_^}>
Fancy way to uninstall packages, needs fzf installed: nix-env -q | fzf | xargs -I{} nix-env -e {}
jasongrossman has joined #nixos-chat
ottidmes has quit [Ping timeout: 244 seconds]
jackdk has quit [Ping timeout: 268 seconds]
sir_guy_carleton has quit [Quit: WeeChat 2.2]
ottidmes has joined #nixos-chat
hedning has joined #nixos-chat
__monty__ has joined #nixos-chat
ottidmes has quit [Ping timeout: 272 seconds]
__Sander__ has joined #nixos-chat
<lejonet>
gchristensen: erlang is wonderful, when you do stuff its designed to do, like distributed computing, many small things that are stateless etc etc, but once you start going outside the rather narrow scope of what it was actually designed for, it'll become a bit more hairy (i.e. as ivan said, parsing stdin and child handling and such)
<joepie91>
lejonet: from what I've seen, this is true for many functional languages :P
<lejonet>
joepie91: indeed
<lejonet>
joepie91: they have a tendency, maybe more than imperative/object oriented language?, to be quite purposefully built around a specific or a few specific usage areas
<lejonet>
Like erlang, its designed for fault-tolerant, seamless distributed computing of message-based stacks
<lejonet>
(and high availability)
<joepie91>
lejonet: I mean, I'd say that most languages were designed that way; just you need an escape hatch or a solid interop mechanism between languages or *something* to deal with the cases it wasn't designed for, and that sorta comes for free when designing an imperative language but not when designing a much more constrained functional language
<joepie91>
if you look at the history of pretty much any popular language, it probably has an origin along the lines of "I needed to scratch a specific itch"
* joepie91
has got to go now, though
<lejonet>
joepie91: fully agree there
<ldlework>
I love BEAM but I don't love Erlang per se.
<lejonet>
ldlework: same here, I like what Elixir has to add, runtime-wise, to OTP and BEAM, but I don't care much for the syntax of either erlang or elixir
<ldlework>
Yeah Elixir is more Ruby than proper ML and that basically breaks my heart.
<lejonet>
same here
<lejonet>
sure, iex is a lot nicer to do ad-hoc programming in, than erl, but both are headdeskable to use
<ldlework>
I once saw a project trying to bring F# to BEAM and I almost cried when I saw it went no where and is dead.
* lejonet
has never looked at F#
<ldlework>
It's a lovely competent contemporary ML. It's not... Haskell (yes, I know Haskell is not techncially an ML) but it is still quite good.
<ldlework>
Expression based syntax, currying, pipelining, overloadable operators, pattern matching, monad syntax support
<ldlework>
One of my favorite technical books of all time is Reppy's Concurrent Programming in ML and I found a faithful CSP library called Hopac for F#, so I went through that book porting all the examples to F# and it still stands out as one of more fun programming things I've done in the last decade.
<lejonet>
I might have to give it a look, is there a foss runtime for it (aka can it run on linux?)
<ldlework>
Yeah
ottidmes has joined #nixos-chat
<lejonet>
Sweet :D
<ldlework>
The dotnet core is open source and can run on Linux - but there is also Mono of course, which has been top-notch for like a decade or something.
<lejonet>
I'm doing my best to get back into to "real programming" again (been scripting for years, but never anything above that really)
<ldlework>
F# actually has a very lively web community who are using Fable a F# to JS compiler to do isomorphic web-development
<lejonet>
cool :D
<lejonet>
so that you don't have to touch JS yourself :D
<ldlework>
Yeah they have some slick Elm-like framework with reactive components on front-end being driven by GraphQL and stuff.
<ldlework>
It is all impressively modern.
<ldlework>
Once I moved to NixOS I became too confused about Nix itself to figure out how to get that build-chain working because there is not great Nixpkgs support for dotnet stuff.
<ldlework>
But now that I'm more experienced with Nix perhaps I'll go back sometime and give it another try.
<ldlework>
I should probably just learn Haskell though.
<lejonet>
haha :P
<ldlework>
F# isn't pure or lazy though. And actually, you can just stick imperative code or for-loops anywhere you want.
<ldlework>
It is really interesting that way.
<ldlework>
It's like a respectable ML, but you can also write classes and be imperative. Quite unique.
<lejonet>
That is quite unique
<lejonet>
Hows the performance of it in relation to other functional languages? Like erlang or haskell
<ldlework>
I don't have a clue
<lejonet>
Sure, in many cases nowdays, that is not an issue anyway, but always nice to know, like pre ruby 1.8, it was highly relevant for regex usage in relation to most other languages, because it was slower than molasses in pre ruby-1.8
qyliss has quit [Quit: bye]
qyliss^work has quit [Quit: bye]
qyliss has joined #nixos-chat
qyliss^work has joined #nixos-chat
__monty__ has quit [Read error: Connection reset by peer]
__monty_1 has joined #nixos-chat
__monty_1 has quit [Remote host closed the connection]
__monty__ has joined #nixos-chat
__monty__ has quit [Read error: Connection reset by peer]
__monty__ has joined #nixos-chat
<manveru>
aszlig: is there any way i can share modifications needed for steam games so they're applied automatically?
<aszlig>
manveru: err, what's the context?
__monty__ has quit [Remote host closed the connection]
<manveru>
just patched the Dead Cells start script because they set LD_LIBRARY_PATH=. and LD_PRELOAD=
<manveru>
so instead of `LD_PRELOAD= LD_LIBRARY_PATH=. ./deadcells` it should be `LD_LIBRARY_PATH=$LD_LIBRARY_PATH:. ./deadcells`
<manveru>
but not sure i should recommend that to the game devs, might break for other platforms
<manveru>
so was wondering if we can add nixos-specific tweaks after a game is installed?
<manveru>
also i'd like to get all the Zachtronics games to run, but they fail without any output at all... :|
<aszlig>
manveru: ah, i have started with Dead Cells in vuizvui at some point
<aszlig>
the idea was to directly package hashlink
<LnL>
the darwin stdenv also increased >100m since I started tracking it
<andi->
I have been thinking of just introdusinc g anew argument to `makeTest` that allows overriding the qemu version that is being used.. Not sure if that would be an elegant solution to the problem.
<samueldr>
andi-: what's the issue it would solve?
<samueldr>
I mean, there's the PR you linked, but is that insufficient?
<samueldr>
ah, logging-in into tests
<andi->
samueldr: it would break nixos-rebuild-vms according to xeji. Not really break but remove features that users might have goten used to (gtk interface for qemu, …)
<samueldr>
yeah, you're right, so it's probably the good solution then?
<samueldr>
tests test with a testing-specific qemu, and can be overriden for nixos-rebuild
<andi->
yeah, just not sure how to properly do the overriding there without moving too much stuff around