Small update on Tabula, my static site generator - of course, it is all handmade - it is fairly featureful for a blog, meaning I am able to transport my old ghost blog content over entirely, with no changes, fudges or workarounds needed. Even the longer, feature heavy posts are being generated in 50-100µs. At this rate, Tabula should be able to generate about 1000 blog posts in the typical blink of an eye in a worst case scenario (100ms according to duckduckgo) :coolcat1:

I doubt I will reach 1000 blog posts written in my lifetime, so I'm cool with that :Crazy_Laugh:

Ported the dynamic collision detection and added ray cast shadows and flare projectiles:

Spent the last ~day and a half writing a static site generator with no libraries in Jai

Still have some quirks with markdown parsing to deal with, but happy with how it is shaping up in such a short time

Bonus feature; It has BYOCSS

Our Paint v0.5a is released!

• LZMA Compression
• Better X11 input device detection
• Much more stable in large canvases
• And MORE!

Get Our Paint on:
https://www.wellobserve.com/index.php?post=20260306205227
https://chengdulittlea.itch.io/ourpaint/devlog/1450204/our-paint-v05a-is-out-now

fixed a bunch of bugs in my aarch64 -> aarch64 jit this past week.
i can now run several complex programs without crashing, including Preview.app (showcased in the video). performance definitely needs some work in the long run, but for now, it's good enough :happy_triangle:

Updated the brush cursor outline to be the inverse of the canvas colors so you can see it better. Also color dropper is now on the side instead of being so big on the canvas. You can get the latest release here!
https://makinggamesinc.itch.io/spixl-pixel-art-editor/devlog/1445612/latest-release

Finished up porting the columns feature and interactive map editing:

I recently implemented the Amanatides & Woo algorithm for voxel ray casts. I wrote a small visualization to help me understand how it works. The horizontal green line shows the nearest boundary for the lines of Y and the red line shows the nearest boundary for X. The green cube is where the white vector would eventually intersect the green line. Same for the red cube and red line. The intuition behind the algorithm is that you always move in the direction of the nearest cube, so if the green cube were closer than the red cube with the green line above, then we'd increase +Y. If the red cube were closer and the red line to the right, then we'd increase +X.

Improved PBR maps, lighting and render-debug modes:
Strengthened roughness and AO maps
AO map darkens additive ambient lightings
Specular reflectance additive over rough surface lighting to give a more wet-over-rough feeling
Optimized GGX and added Phong, Blinn and Lambert shading modes
Added toggles for individual maps (AO, roughness, normal)
Added debug modes for maps and overall lighting
Added HUD lines for all render settings

I am prototyping a chat program. It sort of works.

finished my gamejam project, a homebrew game for the sega dreamcast
https://novicearts.itch.io/elysian-shadows-lost-in-transmission
This one was a doozy from a programming perspective. I implemented alot of RPG mechanics from the groundup (enemy AI, items, map loading, etc etc). Ended up feeling like I was coding a zelda game

making progress on an rpg I'm making for a game jam

A new article: "Choosing a Language Based on its Syntax?"
https://www.gingerbill.org/article/2026/02/19/choosing-a-language-based-on-syntax/

(oops, reposting because used wrong channel for previous message)
Single-file C header for doing 128-bit math operations for msvc/clang/gcc.
Includes full 128-bit signed & unsigned division, handling 128-bit quotient and remainder values.
Compiles to mostly branchless code with msvc & clang.
https://github.com/mmozeiko/overflow/blob/main/x128/x128.h

if you want 32-bit or 64-bit int to hex string in simd (sse2, ssse3, avx512, neon, wasm and rvv) or just in plain C swar, no loops: https://gist.github.com/mmozeiko/68c2d1ce466422b506b2c86e4f603f53

PBR shading and textures with diffuse, roughness, AO and normal maps.
Also made the lighting from the torch as a proper point light with a position around the player, and then gave it some breathing movement and a more appropriate lighting color.

Finally got around to improving the lighting, with a more appropriate and dramatic torch lighting. Before it was constant light intensity per scanline (horizontal for walls/ceiling and vertical for walls). Now is more appropriate spherical attenuation around the player position (as it was in the original Odin project way back then).
This makes it look more like an actual 3D ray casting/tracing rendering, as the lighting behaves as though there were camera-correct ray directions per-pixel.
Also then added some flame-like light flickering to get more of the feeling of walking around with a tourch.

Wanted to see if it would be possible to free the camera to look up and down, while still keeping it only pseudo-3D and still casting only a single horizontal line of rays across the screen. Of course it wasn't initially, because the vertical symmetry was heavily relied on - only the top half of the screen being computated with the bottom half using mirrored results. But ultimately, I've found that it is absolutely still possible (albeit not necessarilly optically correct though still looks reasonably plausible), by computing where the horizon line should go when looking up or down, and computing the full height of the screen with above and below the horizon line computed with respect to where it goes. Performance dropped a bit, but not by that much really, all things considered. CUDA kernel now uses the full screen grid though, as opposed to the half that it previousely used to.

Another big update to the Ray Cassting dungeon crawler project - removing the use of a RGB-float buffer and writing directly into u32 pixels more than quadruppled the performance on the GPU, and is now easilly around 2000-3000fps(!) Also added debug modes for Untextured, Depth, UVs, and Mip-Levels, and some of them on the CPU even get to 1000fps. So unsurprisingly the main bottleneck is the software texture filtered sampling, on the CPU (no SIMD is used, and all on a single thread). But GPU has no issue at at all, using the same software texturing code (no hardware texture units used). Can even be above 200FPS on my ultra-wide 5Kx2K screen.

Several years ago I wrote a ray-casting dungeoun crawler completely from scratch in Odin, that does everything including rendering (softare texturing with bi-linear filtering, mip-mapping and sampling) manually on the CPU (a VERY old Ryzen 1.2) on a single thread with streightforward code (no SIMD).
I recently brought it back to life with more recent version of Odin, and it runs about the same as it did, at a little over 30fps at less than 640x480.
UPDATE: Just now I looked up what optimization to set for Odin compilation, and using -o:speed Jumps it to over 300fps when not moving, and still over 200fps when moving (not shown in this video but posted another below). Also getting around 60fps at much higher resolution - cudos to Odin(!).
More recently I decided to try to back-port it to C++ using my SlimEngine base project, with an aim to later put the same rendering code on CUDA like I did with SlimTracin.

After some simplification and optimization, at the same resolution and on the same CPU, still single threaded and unvectorized, it now runs at over 200fps (and over 300fps on my work laptop).
I then got it working also on CUDA (RTX 4070) and on the same computer, and there it now runs over 600fp (and at 5K it runs over 60fps). Same code can be toggled between CPU or GPU at runtime (just like SlimTracin XPU).

Still doing the actual 2D ray-casting and generating vertical and horizontal rendering data on the CPU and that is resolution dependant.
The GPU (if/when used) is mostly just doing the software texture sampling and shading, so there's still room to improve, but pretty happy with the result so far 🙂

Small update on Tabula, my static site generator - of course, it is all handmade - it is fairly featureful for a blog, meaning I am able to transport my old ghost blog content over entirely, with no changes, fudges or workarounds needed. Even the longer, feature heavy posts are being generated in 50-100µs. At this rate, Tabula should be able to generate about 1000 blog posts in the typical blink of an eye in a worst case scenario (100ms according to duckduckgo) :coolcat1:

I doubt I will reach 1000 blog posts written in my lifetime, so I'm cool with that :Crazy_Laugh:

Ported the dynamic collision detection and added ray cast shadows and flare projectiles:

Spent the last ~day and a half writing a static site generator with no libraries in Jai

Still have some quirks with markdown parsing to deal with, but happy with how it is shaping up in such a short time

Bonus feature; It has BYOCSS

Our Paint v0.5a is released!

• LZMA Compression
• Better X11 input device detection
• Much more stable in large canvases
• And MORE!

Get Our Paint on:
https://www.wellobserve.com/index.php?post=20260306205227
https://chengdulittlea.itch.io/ourpaint/devlog/1450204/our-paint-v05a-is-out-now

fixed a bunch of bugs in my aarch64 -> aarch64 jit this past week.
i can now run several complex programs without crashing, including Preview.app (showcased in the video). performance definitely needs some work in the long run, but for now, it's good enough :happy_triangle:

Got a little update on my text editor I've been making for large projects and large files!

• It's now called Photon (cause it's as fast as light lol)
• The core text editing system has been completely rewritten to be fully threadsafe and lockless
  • Thanks to @starfreakclone for the help! Uses a similar data structure to his editor Fred
• It is now highly parallelized and even faster!
• It now support projects!
• In the demo below I have a copy of Unreal Engine's source code loaded as the project, 43,000 files (that includes Engine/Source/ThirdParty)
  • I can navigate the source tree to a file and load it instantly
  • I can fuzzy search for a filename in all the project files and load it instantly
  • The feature I'm most proud of, you can string search all 43k files in less than 16ms, giving you realtime, instant search feedback, updates instantly as you type
  • You can jump to any file and location instantly!
  • At the end, just for extra flare, I open a 1GB json file, which loads in about half a second, you can jump around in it, and search the 1GB file at realtime instant speeds

Updated the brush cursor outline to be the inverse of the canvas colors so you can see it better. Also color dropper is now on the side instead of being so big on the canvas. You can get the latest release here!
https://makinggamesinc.itch.io/spixl-pixel-art-editor/devlog/1445612/latest-release

I’ve been building a custom C + LLVM based language called Ura just for fun.
I finally got it to a point where it can handle full networked systems.
I put together a TCP client-server chat as a milestone.
There’s a unique satisfaction in watching two terminals talk to each other using a language you created from scratch especially moving from simple parsing to actually managing things like process IDs and sockets.

The goal for the syntax was to keep it close to the metal but make it a lot more enjoyable to write.

Repo: https://github.com/mohammedhrima/ura-lang/

Finished up porting the columns feature and interactive map editing:

I recently implemented the Amanatides & Woo algorithm for voxel ray casts. I wrote a small visualization to help me understand how it works. The horizontal green line shows the nearest boundary for the lines of Y and the red line shows the nearest boundary for X. The green cube is where the white vector would eventually intersect the green line. Same for the red cube and red line. The intuition behind the algorithm is that you always move in the direction of the nearest cube, so if the green cube were closer than the red cube with the green line above, then we'd increase +Y. If the red cube were closer and the red line to the right, then we'd increase +X.

Improved PBR maps, lighting and render-debug modes:
Strengthened roughness and AO maps
AO map darkens additive ambient lightings
Specular reflectance additive over rough surface lighting to give a more wet-over-rough feeling
Optimized GGX and added Phong, Blinn and Lambert shading modes
Added toggles for individual maps (AO, roughness, normal)
Added debug modes for maps and overall lighting
Added HUD lines for all render settings

I am prototyping a chat program. It sort of works.

finished my gamejam project, a homebrew game for the sega dreamcast
https://novicearts.itch.io/elysian-shadows-lost-in-transmission
This one was a doozy from a programming perspective. I implemented alot of RPG mechanics from the groundup (enemy AI, items, map loading, etc etc). Ended up feeling like I was coding a zelda game

Hosting a website from home using no operating system using a networking stack I wrote from scratch and a RMII/MDIO driver I also wrote from scratch. http://gsan.whittileaks.com. Its HTTP only so make sure your browser does not automatically add https:// in front. Come break it 🙂

Full post here https://handmade.network/forums/wip/t/9116-lneto_-_heapless_userspace_networking_stack_in_go