Gabriel Dechichi ✓

[video] that's how fast this can be in C btw this is *still* slow because you would not run layout on 5k elements anyway, you would virtualize, but the point of the benchmark is to check the layout algorithm scalability computers are fast.

[video] it's the same speed for web and native, video below. 9 out 10 times there is no valid reason for a game to run slower on the web then it does native. people just like to use this as an excuse for their slow code.

[image] no one on X seems to be looking at code anymore. but I just love looking at C code can't really explain why

[image] the C engine is finally ready for prime time, it's name is Kaigen. in the next few days I'll start posting more demos, and re-posting old demos, with playable links and intake forms for a closed beta. if you're a professional game or 3D app developer who wants to give your

[image] folks might think AI coding "changes everything", but attempts to get non-technical people to "write code" have been around for 15 years. they are always somewhat helpful, but programmers are very much still here, and I don't think we are going away :)

[video] testing the engine's asset system by importing tons of publicly available assets. that's 3777 assets loaded in 1.4 seconds. probably could be faster still. thanks @KenneyNL and @quaternius for providing so many high quality assets for free

[video] statemens like this are so dumb I don't even know where to begin. but I'll try. 🧵

[image] I'm glad more people are realizing this.

[text] that's *exactly* what I've been saying for the past year. C was always the best language for building ultra fast, high quality software. the challenge is just that C can be hard to write correctly. now with LLMs everyone should just be writing C

[video] First demo: The Minecraft Test. @kaigen3d works great with AI. The difference with Three.js is that when you're done it runs on Windows, Mac, iPhone, Android *and* the Web. (VR soon!) oh and it runs at 330+ FPS on a 6 year old laptop gpu. link to demo + code below🧵

[video] this guy is building his game to run *entirely* on the GPU btw everything is compute shaders

[text] this Thursday 9am PST / 12pm EST I’ll be on the @wookash_podcast talking about my C game engine. we’ll demo some of the code too! hope to see y’all there!

[image] literally what *every* single professional developer is thinking right now

[video] tried implementing dual quaternions on my animation system tonight and it totally worked on the first try

[video] people think crashes are hard to solve, but usually they are very easy to solve now *these bugs*, those are a pain to solve time to go figure out why I have corrupted vertices *only* for this asset 😅

[video] this is using my C game engine btw. I'm *so close* to release, people have no idea what this tech can do

[image] Unity: "you will be able to prompt full casual games into existence" also Unity:

[video] how many times can I open @kaigen3d before @unity opens once? looks like 10 times flat.

[video] designing and implementing a real programming language in less than a year is no joke. inspiring work.

[image] looks like I struck a nerve? 😉

[video] another app being built in my upcoming C engine, and that's *one day* of work. one of the advantages of using C is that it binds to everything, so this app is building native iOS UI in Swift, while all of the 3D rendering is done in C.

[video] this kid got access to @typesafeai Jev like yesterday and already built a system for granular expression control for his character app. all written in C btw

[text] *starting* your own startup is definitely easier than getting a job. *succeeding* at building a startup is 100x harder though

[video] you guys don't know this but I was deep into rubik's cube as a kid (you can search my name online). so I wanted to see if @typesafeai Jev could solve the rubik's cube by itself. it turns out it can lol, and it does it just like a human would

[video] I think I discovered a way to generate 3d animation from text / image / video that supports a lot of creative control which is not possible with current mocap driven text to animation models. still early but so far results seem quite promissing

[video] super excited to work with @mutuals_inc on this

[video] getting AI memory long-term storage and retrieval to work fast reliably has always been a challenge. @mutuals_inc is building an entire memory system on top of Jev to make this ultra fast *and* cheap.