5 unstable releases

0.6.0 Sep 14, 2024
0.5.1 Sep 8, 2024
0.5.0 Sep 8, 2024
0.4.1 Sep 5, 2024
0.4.0 Aug 3, 2024

#218 in Graphics APIs

Download history 5/week @ 2024-08-09 82/week @ 2024-08-30 285/week @ 2024-09-06 182/week @ 2024-09-13 18/week @ 2024-09-20 24/week @ 2024-09-27 4/week @ 2024-10-04

273 downloads per month

MIT license

40KB
932 lines

Taika Logo

A low-cost abstraction layer on top of wgpu and winit to make their APIs more ergonomic.

This is very much a work-in-progress. This was originally meant to replace LoiRen, the renderer in loitsu, but the projects kinda got split off. Might still happen one day though. Currently most settings are hard-coded this will change.

State

Taika is early in development, meaning large API changes are bound to happen. However it is currently being used for a production ready game which serves as a good testbed for the library.

Goals

  1. Simplify window creation
  2. Introduce "RenderPasses" and "RenderPipelines", which are common tropes in game engines.
  3. Make API changes in WGPU and Winit less frustrating by providing a semi-stable API. API changes will still happen though.
  4. Give full access to WGPU

In addition to these goals taika also includes some common utilities mainly targeted towards game-development. Taika also includes a super basic form of asset management. It is designed to be built upon, not to be a full-fledged asset management system.

What taika doesn't do:

  • Input-handling, you can do this yourself by listening to the winit events that are passed through to your event-handler
  • Audio, use other libraries
  • Make rendering easy. You still have to write shaders, and implement the drawable trait, to actually issue the drawcalls to the GPU. Taika doesn't make any drawcalls by itself

Notes

  • The naming of rendering::RenderPass and rendering::RenderPipeline is a bit confusing at they are also used in wgpu.
  • No examples currently!

Platforms

Actively tested on Windows, Linux (wayland and x11) and macOS.

Web might happen one day

Dependencies

~8–44MB
~704K SLoC