[1/2] I felt like this question deserved a more complete answer than "undefined behavior", so I made a full walkthrough to show exactly what part of the x86 execution differs from the other two, and to demonstrate how you go about investigating an anomaly like this.
Huge shoutout to @valigo, who **on the spot** joined the livestream, and showcased jai metaprogramming and crazy `or_return` macro! Badass programmer aura!
Members preview available (edited version out in couple of weeks!)
https://t.co/99n5f7JXdP
Tomorrow airs a lesson on State Management AKA intrusive lists AKA fat structs AKA plex AKA Large Arrays of Things™ ! Here is a preview :)
Of course with @antovsky
Hey! My #raylib talk at FOSDEM 2026 is already available online! 🚀
It's been my biggest talk to date and the one that took me more work, with +5000 locs of hand-coded C 😱
Hope you enjoy it! 😄
Talk: https://t.co/fHOxtCja0f
Interactive presentation: https://t.co/dEJJroERuC
Join @antovsky and myself for a special live episode, where we're gonna do a live code review :3
Project that's gonna be reviewed is a surprise!
Tomorrow!
https://t.co/WoduWRwoqg
Here is an incredible discussion with @Jonathan_Blow!
- design choices for @SinkingStarGame
- live demo of engine and tooling (!)
- jai metaprogramming showcase that's used for the game
- release (both jai & Order), and what games come next!
@Jonathan_Blow it was an honor! Thank you!
People in comments are talking Rust v C/C++.
imo the more concerning take here is the attempt to rewrite huge codebases from scratch.
In huge complex programs you end up with unintentional behaviour that other parts (of the same program, as well as downstream users) depend on.
By “fixing” it and rewriting things anew, you miss a ton of these and you ship an entirely new and incompatible program.
You force all the dependants to adapt to new reality.
And if you set to rewrite X year old tech, be pretty damn sure you know everything that people back then knew and planned for!
And to top it off, organisationally, after some time and effort spent in the rewrite it becomes “too big to fail”. No one will have the courage to bubble up the chain the honest truth, the sinking cost fallacy kicks in and everyone is sadly shipping a worse v2 of a functional v1. (Been there three times!)
If you want to improve the current state, don’t attempt from scratch rewrites for a complex piece of code of many states and outputs.
Instead develop tooling to help learn the codebase and its quirks (profilers and debuggers are good at this, you can run coverage on a prod like system to discover not used / legacy code)
You can write custom software to visualise the events that happen in between internal systems (and all 7 generations of it, that your product accumulated over years).
Works for all large systems, huge desktop applications, games, giant SQL data processing pipelines, webs of microservices (as well as that one not-so-micro-anymore-service which holds 40% of business logic), etc.
Not only will you discover tons of race conditions, unused code and tons of other issues with your codebase, but you will also help your fellow Distinguished Engineers to like it. Cause now they seem to dislike it to the point of getting rid of it!
Join @Jonathan_Blow for a LIVE interview about the development of @SinkingStarGame!
How the development looked like? How and where jai language made a difference? How the game systems have been built? What where the technical challenges in the game production? How wild jai metaprogramming looks like in Sinking Star?
Live Q&A at the end :)
Next Week LIVE Episodes Announced: 2 out of 2
Sometimes you meet a person who knows a lot, and you're taken aback. But rarely do you get the opportunity to talk to a mind like Mārtiņš Možeiko.
Mārtiņš can tell you about hundreds of obscure OS API calls, but he can also tell you about *undocumented* ones he found out by trying things out. What?
To top that he's low key, easy-going, not chasing the spotlight and casually helping out thousands of people, sharing his immense knowledge, answering questions for free on all sorts of forums.
Wow :) Enjoy the talk!
Just took @vkrajacic's filepilot's for a run. It makes Windows feel... modern ?
We should get together, buy Windows from Microsoft, and rewrite it.
link: https://t.co/Gg1gZkluvl
Credit to @rfleury and @Jonathan_Blow for promoting filepilot.
(This is not an ad, and I don't know the author.)
IMPORTANT message for everyone using Gmail.
You have been automatically OPTED IN to allow Gmail to access all your private messages & attachments to train AI models.
You have to manually turn off Smart Features in the Setting menu in TWO locations.
Retweet so every is aware.
On Saturday airs the episode with an absolute legend: Mārtiņš Možeiko
He's an unsung hero of Handmade forums, helping out tons of people and sharing his immense knowledge - all in his spare time!
Please look forward to the episode! :)
raylib, C
more low level, high efficiency adventures
been building an ink-like terminal UI system in ImGui-style immediate mode
the output of this library is just an array of characters, colors and meta info (bold, italics, etc)
this can be neatly packed as textures, uploaded to the gpu and drawn in a single drawcall using a font atlas and vertex/fragment shaders
literally reinventing the wheel but it feels like a renaissance
heavily WIP, minor issues present in layout, etc
#gamedev #indiedev
will eventually be using this with the amazing:
@ajmmertens #flecs
@raysan5 #raylib
@jsuarez #pufferlib
NEWS! raylib custom Software Renderer backend [rlsw] has been finally merged! 🚀
For the first time in this 12-years journey, there is a NO-DEPENDENCIES path available for raylib rendering, no platform library (direct Win32 implementation) and no OpenGL required! 🤯