It’s been a long time I wanted to announce and reveal this, and this morning is now the moment! This blog post announces the release of cheddar in version 0.1!
Cheddar is a GLSL superset language. What it means is that most of the GLSL constructs and syntax you’re used to is valid in Cheddar – not all of it; most of it. Cheddar adds a set of features that I think are lacking to GLSL. Among them:
- Some non-valid GLSL constructions made valid in Cheddar to ease the writing of certain shader stages.
- A more functional approach to programming shaders on the GPU.
- Structures, types and GLSL-specific constructs sharing.
- Imports and modules with live reloading and transitive dependencies.
However, Cheddar is not:
- A new shading language that abstracts over any kind of shading language (i.e. unlike Cg for instance). However, since it should be possible for GLSL to be transpiled to SPIR-V (hence Vulkan), it should be possible to transpile Cheddar to SPIR-V.
- A dedicated DSL to a specific graphics problem. Cheddar is very similar to GLS (actually, that’s why it’s called a GLSL superset). If you’re looking for a solution to ease writing shaders in a way that you don’t have to learn shaders, Cheddar won’t help you much with this.
The documentation is still a big work in progress but most of it should give you enough information to get started.
A bit of backstory
Cheddar was imagined and designed while I was working on spectra, a work in progress demoscene crate I’ve been working for a while now. I released two demos thanks to spectra – this one and this one. Because those demos are very meh to me, I decided to enhance my tools. Among them was the need to write shaders a better and easier way. Then Cheddar got born. I used Cheddar while preparing other demos, and I was talking about it on IRC and Reddit, people seemed to be interested – I even had a friend writing demos in C++ interested!
So here it is. Please, provide feedback if you try it, should you like it or not!
Disclaimer: the current state of the language is pretty experimental and unstable. There’s no semantics checking, for instance.
You can read the full tutorial and design document on the official documentation page.
Thanks for having read me, and as always, keep the vibe!