@terpatin.bsky.social
👨💻 Gamedev hobbyist.
✨ Likes to fool around with shaders.
🎮 Focus on GameMaker.
✨ Likes to fool around with shaders.
🎮 Focus on GameMaker.
Just to clarify; look at Step-event time, and not Draw-event time, which just happens to be the first.
February 12, 2026 at 8:37 AM
Just to clarify; look at Step-event time, and not Draw-event time, which just happens to be the first.
Maybe the game logic is done within shaders 🤔
January 8, 2026 at 8:02 AM
Maybe the game logic is done within shaders 🤔
Cool!
In the GLSLES version GameMaker currently uses, bit operations are not available, which I would otherwise used.
The rgba16float format has 64 bits, but 4 bits are unused to avoid dealing with infinities, which would break encoder.
In the GLSLES version GameMaker currently uses, bit operations are not available, which I would otherwise used.
The rgba16float format has 64 bits, but 4 bits are unused to avoid dealing with infinities, which would break encoder.
November 22, 2024 at 11:32 AM
Cool!
In the GLSLES version GameMaker currently uses, bit operations are not available, which I would otherwise used.
The rgba16float format has 64 bits, but 4 bits are unused to avoid dealing with infinities, which would break encoder.
In the GLSLES version GameMaker currently uses, bit operations are not available, which I would otherwise used.
The rgba16float format has 64 bits, but 4 bits are unused to avoid dealing with infinities, which would break encoder.
One curiosity I made some time back. If one is using rgba16float textures, but for some reason want to store six 10bit values instead, here are encoder functions: github.com/HannulaTero/...
Encode-six-values-to-rgba16float/shaders/shd_encoder_three10bit/shd_encoder_three10bit.fsh at main · HannulaTero/Encode-six-values-to-rgba16float
For GameMaker to be used with web exports, for encoding six 7bit values in rgba16float. - HannulaTero/Encode-six-values-to-rgba16float
github.com
November 21, 2024 at 7:00 AM
One curiosity I made some time back. If one is using rgba16float textures, but for some reason want to store six 10bit values instead, here are encoder functions: github.com/HannulaTero/...
Thanks 😊 I am still keeping the repo private until I'm ready to actually release it 😌
November 21, 2024 at 6:14 AM
Thanks 😊 I am still keeping the repo private until I'm ready to actually release it 😌
November 20, 2024 at 10:34 PM