<testuser[m]>
maybe u forgor a return after warn()
<testuser[m]>
oh illiliti already pointed out the wrong type
an3223 has joined #kisslinux
<testuser[m]>
The author clearly confuses one problem for another. Yes, if you allocate elements in a linked list with malloc, you're going to have a bad fucking time as it will spread out yourallocations everywhere. This is a problem with malloc/free and not with linked lists. I don't use libc and don't even have a "malloc" function because the concept of treating allocations as side effects is horrible and stupid
<riteo>
how are things going, other than dylan disappearing from the public once again?
<midfavila>
No idea wrt: KISS
<midfavila>
I haven't used upstream in months
<midfavila>
personally I guess I'm alright. Just slowly making progress on maths and CS, as usual. You?
<riteo>
I've closed quite a lot of projects (minekiss and stuff) to focus on learning more C/C++
<riteo>
I've also read a bit that C book I *ahem* found on the internet
<riteo>
(only a little bit though, got distracted with other, more pratical stuff)
<riteo>
the biggest thing I've been doing is porting Godot to Wayland
<midfavila>
C is something I need to dedicate more time to. Ordered a bunch of books about Unix programming, so hopefully that'll give me ideas for good projects. What have you been reading?
<midfavila>
Oh, that's neat. Has it been difficult?
<riteo>
surprisingly, not that much
<riteo>
I mean, it's still not done, but, ignoring weird limitation imposed by Wayland, I've been really liking the whole API-from-protocol thing
<riteo>
s/limitation/limitations/
<riteo>
people hate fragmentation, but I really like the idea of making everything a different protocol. People have different needs.
<midfavila>
Hmm. I'm definitely in the opposite camp.
<midfavila>
Ideally everything would use formatted text communicated via identical IPC mechanisms, at least in my opinion
<riteo>
I don't see how that would go against the idea of different protocols for different things
<midfavila>
Well, depends on how you define protocol, I guess - but fair enough.
<riteo>
I haven't looked your website in a while, have you been collecting more weird hardware?
<midfavila>
I considered it, but no, not as of late
<riteo>
I had no idea wikipedia had a page for this
<midfavila>
now you do
<riteo>
wikipedia for some reason has pages for the weirdest or most "obvious" of things
<riteo>
like use case
<riteo>
why is there the need to have a whole page for "use case"
* midfavila
shrugs
<riteo>
oh wait, it might look at it in some weird technical sense
<riteo>
whatever
<dKstv>
Hi everyone
<riteo>
hi!
<dKstv>
Im new to kiss, enjoying it a lot so far
<illiliti>
dilyn: fyi, people at wlroots are reworking libinput backend. not sure if it's related to the bug you have, but there's a possibility
dKstv has quit [Remote host closed the connection]
dKstv has joined #kisslinux
dKstv has quit [Remote host closed the connection]
eudfty has joined #kisslinux
eudfty has quit [Remote host closed the connection]
<testuser[m]>
C is the only sane language left. It never changes and it lets us write code that just works. We can focus on getting things done and shipping rather than learning new features and Googling compiler errors. Learn once, ship forever.
<testuser[m]>
On my last workplace we argued if we dared to use stdint.h, since it was so new and fancy. I was a proponent, but there is something calm over discussing whether to use 23 year old additions or not.
<testuser[m]>
I almost had a heart attack when the twitter post stated that "true" and "false" are keywords in C23. However that seems incorrect.
<illiliti>
i wish we have go-style defer in c23
Rohan has joined #kisslinux
<Rohan>
hi guys
<Rohan>
any news about dylan?
<testuser[m]>
Hi
<testuser[m]>
No
<testuser[m]>
But you know what? Something happened that changed my life. I bumped into a Rustacean. And another Rustacean. And a few more Rustaceans. I found five Rustaceans. And they showed me what they did to become Rustaceans. Again, it’s not just about jobs, it’s about the good programming; ergonomics, safety, zero cost abstractions and efficiency.
<Rohan>
lol
<Rohan>
rust compared to c++ is better?
<omanom>
it's probably that i haven't used it very much / learned much about it... but rust is such a PITA to read and comprehend
<Rohan>
ive been thinking in learn rust for some projects
<Rohan>
but dont know if that be useful than other languages
<omanom>
depends on the project and how you define "useful"
<Rohan>
not that big just some CLI programms
<Rohan>
for practice
<testuser[m]>
Use rust or goolang
<illiliti>
rust is no-go absolutely
<testuser[m]>
Why
<illiliti>
pita to bootstrap
<illiliti>
extemely slow and inefficient build system
<Rohan>
anyone have a weston build out of the box?
<illiliti>
"move fast and break things" TM
<Rohan>
need some to test my waydroid package
<midfavila>
do they have non-x86 support yet
<midfavila>
or is memory safety still limited to the high-reliability world of PCs
<omanom>
musl is tier 2 looks like: "Tier 2 targets can be thought of as "guaranteed to build". The Rust project builds official binary releases for each tier 2 target, and automated builds ensure that each tier 2 target builds after each change. Automated tests are not always run so it's not guaranteed to produce a working build, but tier 2 targets often
<omanom>
work to quite a good degree and patches are always welcome!"
<omanom>
kind of sounds like "if it works, great, but we're not going to put any effort into it"
<midfavila>
fucking cat keeps jumping on my desk when I'm trying to work
<midfavila>
reeeeeeee
<testuser[m]>
Send cat.jpg
<riteo>
I don't understand, is there the possibility of them using GNU extensions?
<testuser[m]>
Gnu extensions in what
<testuser[m]>
Rust is in rust
<riteo>
how can you possibly not support a correct libc?
<midfavila>
because fuck you
<midfavila>
you're not relevant
<midfavila>
begone, brogrammer
<riteo>
I thought it was about bootstrapping, how else would musl not be garaunteed to supported
<riteo>
s/garaunteed to/ garaunteed to be/
<testuser[m]>
What's the least worst alternative to c
<riteo>
IMO, the best standard structure is one which has an extremely basic but basically frozen core, slowly going up in complexity with different extension outside of it
<riteo>
kinda like (very roughly) RISC or Wayland
<riteo>
although RISC has "official" extensions
<riteo>
IIRC
<noocsharp>
that's essentially want i want to do with my language
<riteo>
ooh, nooc is a real thing?
<riteo>
from the context I thought it was a joke, sorry
<riteo>
could you tell me more about it?
<testuser[m]>
noocsharp: i thought it was a c coompiler
<testuser[m]>
As of C++20, the right way to write a trivial getter in C++ looks like [[nodiscard]] constexpr auto GetFoo() const noexcept -> Foo { return foo_; }
<noocsharp>
it's basically a (crappy) toy compiler atm
<riteo>
have you thought about using QBE as the backend?
<riteo>
at least for testing
<noocsharp>
i've thought about it
<noocsharp>
what i'm probably gonna do with it is make it into an interpreter, and then write a compiler as a library in nooc
<riteo>
as a bootstrapping method or as a core language feature?
<noocsharp>
well there
<noocsharp>
's this language called jai which does a similar thing
<noocsharp>
except the compiler is implemented in the interpreter
<riteo>
mhhh, I see
<riteo>
so that'd make it both an interpreted and compiled language
<noocsharp>
the idea is you have a build "script" written in nooc which calls the compiler library
<riteo>
that's a very interesting approach, but won't it make the compiler very slow?
<noocsharp>
i'll worry about that if it actually becomes a problem
<noocsharp>
and if i get that far
<riteo>
but it's bound to be slow, isn't it?
<illiliti>
i'm shocked that in 2022 we still don't have a portable way to set symbol visibility
<illiliti>
i wish c23(or posix) do something about this
<noocsharp>
it's gonna be a pretty lightweight bytecode interpreter and a pretty small core compiler
<riteo>
wait wait wait
<riteo>
bytecode?
<noocsharp>
that's how interpreters generally work
<riteo>
oh, as a preloading step I supposed
<riteo>
s/supposed/suppose/
<riteo>
I thought it was going to be bytecode compiled manually before running
<riteo>
if it does that after parsing the script then it makes sense
<noocsharp>
no, the compiler will compile to native code
<riteo>
I see
<noocsharp>
but most of this is hypothetical at this point, right now i just have a compiler written in c which compiles to amd64 machine code
<riteo>
well, gotta start somewhere
<testuser[m]>
"Just put namespaces in C" YOU just put namespaces in C, goddamnit. It's hard enough getting existing practice used by 5+ compilers past the Committee, do you know what it fucking takes to put a WHOLE FEATURE in, that people DO NOT understand???
<riteo>
what happened
<riteo>
why is C23 getting talked about so much lately, is it a fresh thing?
<riteo>
even while looking at cproc it talks about c23 features
mahmutov has joined #kisslinux
<riteo>
omg are we sure that the `true`,`false` keyword thing was wrong?
<omanom>
there's supposedly another features update planned soon?
<riteo>
actually this new revision doesn't sound that bad (ignoring true/false, if that's true)
<riteo>
quite the opposite
<noocsharp>
what's the problem with true and false being keywords?
<riteo>
if they map to integers, hide their real value
<riteo>
if they don't, they add a new type into the core, if there isn't already a bool
dKstv has joined #kisslinux
dKstv has quit [Remote host closed the connection]
<noocsharp>
stdbool.h already hides their real value
dKstv has joined #kisslinux
<riteo>
yes, but it's a library
<riteo>
it's optional, and at discretion of the user
<noocsharp>
yeah, wouldn't want to pollute my prestine programming language with bloated built-in booleans
<riteo>
bruh, those aren't full-fledged booleans
dKstv has quit [Quit: dKstv]
<riteo>
are constants that can be added in an extremely trivial way
<riteo>
s/^are/they are/
dKstv has joined #kisslinux
<illiliti>
testuser[m]: which version of sway are you using?
dKstv has quit [Client Quit]
dKstv has joined #kisslinux
dKstv has quit [Client Quit]
dKstv has joined #kisslinux
dKstv has quit [Client Quit]
dKstv has joined #kisslinux
ehawkvu[m] has quit [Ping timeout: 256 seconds]
<dilyn>
illiliti: have you tried using -display sdl in your qemu invocation? qemu with just sdl2 works perfectly fine for me; I just specify -display sdl -vga virtio to get dynamic resizing
<dilyn>
also ofc they would be rewriting libinput backend for wlroots smdh. I suppose I just have to play the waiting game; I feel like there's no sufficiently good reason to backport these patches to pre-0.15
<illiliti>
dilyn: yep, tried that. nothing is working
ehawkvu[m] has joined #kisslinux
<dilyn>
that's very interesting. especially since I don't even have to invoke root to launch my qemu machines...
<illiliti>
try -enable-kvm and you will need root or rw for /dev/kvm
<dilyn>
ah yes I am in the kvm group
<dilyn>
forgot about that
<dilyn>
are you just building qemu with the standard recipe?
<illiliti>
yes
<dilyn>
curiouser and curiouser
<illiliti>
i even tried downgrading it to 6.1.0 and 6.0.0. neither is working
<illiliti>
gonna rebuild everything with -O0 -ggdb and debug
<dilyn>
XD all in
<dilyn>
might as well; I'm not certain what could be wrong. by all measures my system should be completely broken and yet...
<midfavila>
that's a mood
<dilyn>
*cries in ffast-math*
<testuser[m]>
funroll all loops
dKstv has quit [Quit: dKstv]
<midfavila>
wish I could recompile myself with -ffast-math
<midfavila>
smh
<dilyn>
sometimes I think I was built with ffast-math
<dilyn>
I didn't understand most of my senior thesis:'(
<jslick>
rohan learn all the languages that you are interested in. Don't let these people dissuade you. Unless you are doing it for something specific. The overlap of peoples' opinions of what makes a good programming language here is probably very narrow.
dKstv has joined #kisslinux
ehawkvu[m] has quit [Read error: Connection reset by peer]
<Rohan>
jslick thanks for advice
dKstv has quit [Client Quit]
ehawkvu[m] has joined #kisslinux
<omanom>
except java
dKstv has joined #kisslinux
dKstv has quit [Remote host closed the connection]
dKstv has joined #kisslinux
dKstv has quit [Client Quit]
ehawkvu[m] has quit [Read error: Connection reset by peer]
ehawkvu[m] has joined #kisslinux
ehawkvu[m] has quit [Ping timeout: 272 seconds]
midfavila has left #kisslinux [#kisslinux]
ehawkvu[m] has joined #kisslinux
mahmutov has quit [Ping timeout: 256 seconds]
<Rohan>
java i agree
<Rohan>
i just need it for flutter
<Rohan>
and SDK tools
illiliti has quit [Quit: leaving]
progenyx has quit [Quit: progenyx]
<riteo>
well, it's been fun, I'll go for now, bye!
riteo has quit [Quit: epic comeback moment]
mahmutov has joined #kisslinux
progenyx has joined #kisslinux
midfavila has joined #kisslinux
<midfavila>
bluuuuuuuuh
<midfavila>
lang keeps kicking my ass
<midfavila>
how do I become not shit at math dilyn
<dilyn>
choose a different sub-topic /shrug
<dilyn>
maths is difficult to be good at. I'm not very good at it
<dilyn>
I'm a far better logician than an Analyst, I'm a far better teacher than a student
<midfavila>
fuck
<phoebos>
wow git-add(1) has an interactive mode
Rohan has quit [Ping timeout: 250 seconds]
<phoebos>
wild
<phoebos>
o/ all
<phoebos>
busy in here today for once
Guest93 has joined #kisslinux
<Guest93>
Hi
<Guest93>
Is it possible to install the nouveau driver on Kiss Linux?
<dilyn>
testuser[m]: ^
ella-0_ has joined #kisslinux
ella-0 has quit [Ping timeout: 250 seconds]
Guest93 has quit [Quit: Client closed]
midfavila has left #kisslinux [#kisslinux]
dimKv has joined #kisslinux
dimKv has quit [Quit: leaving]
mahmutov has quit [Ping timeout: 250 seconds]
dimKv_ has joined #kisslinux
dimKv_ has quit [Client Quit]
icy has quit [Quit: ZNC 1.9.x-nightly-20210409-e0ffdddd - https://znc.in]