I've just noticed the recent LLM-policy drama surrounding Rust, and to be honest, it's put me off learning and using it. I'll stop working on my Learning Rust series and focus on Zig or C instead. At least Zig has a strict anti-LLM policy in place.

I should have started with Zig... Now, I'm really angry with myself for wasting weeks writing and learning about it 😠

#Rust #Rustlang #Zig #Ziglang

Office People - now for Desktop PC.
Native Zig port using my Borowik Engine.

#ZigLang #GameDev #IndieDev #PixelArt #FirefoxOS

Codex is porting my old game (for Firefox OS) to my ZIG engine :D

Work in progress but it mostly works :)

#ZigLang #Codex #Porting

In my quest for using -fincremental in Zig I've so far found and fixed 3 fairly critical linker bugs. (Sent a PR for one of them, got to do it for the other two but I'm a little bit uncertain about the details on one of them.)

Seems I'm going to have to implement a few relocation types as well, which means I'm probably going to have to understand the implementation on higher level and not just specific local details.

I'm in for a treat. 😬

#ziglang #zig

Weeknotes 2026 W19: Slightly exploitative

Denis Defreyne’s weeknotes for May 4​–​10, 2026

denisdefreyne.com

raylib is fun. and the mach engine has some nice features. i have a minimal working example of them being used together and it has been great fun to hack on

#raylib #MachEngine #zig #ZigLang

Functional Programmers need to take a look at Zig.

I really do not like the direction Zig is going, personally.

With "direction" i mean the new std.Io type, and I mean, it's really cool that you can now do transparent, uncolored async where the actual implementation is handled by filling up a struct with functions, my main concern is just... the sheer size of that struct with functions, and how they're handled!

You can't just optimize away a field from a struct, as in, if I do not use a function in this set of useful functions, the language really has no easy way of knowing to just remove that function, without a lot of complexity and tree traversal stuff, and while this is actually a solved problem for certain high level languages such as C++, I just don't think that's the way forward.

Making structs into these... schrodinger's boxes where functions exist and not at the same time, introducing massive amounts of complexity to the compiler, to be honest, just doesn't sound right.

I saw someone around in the Zig space claiming that, perhaps, we actually should have a new keyword, or mechanism in Zig which would let us define interfaces with this kind of ideal behind them, a vtable that can and will be shaken off and optimized, which is not perfect but at least it won't be as involved as the previous idea.

Sigh, I miss when Zig was a simple language

#ziglang #zig
Functional Programmers need to take a look at Zig.