#runa #wayland #wayland-compositor #parser #compositor #spec

runa-wayland-spec-parser

Parsing and representing wayland protocol files

1 unstable release

0.1.0 Mar 29, 2023

#1156 in GUI

Download history 37/week @ 2023-11-02 33/week @ 2023-11-09 42/week @ 2023-11-16 44/week @ 2023-11-23 22/week @ 2023-11-30 35/week @ 2023-12-07 34/week @ 2023-12-14 34/week @ 2023-12-21 22/week @ 2023-12-28 36/week @ 2024-01-04 33/week @ 2024-01-11 26/week @ 2024-01-18 29/week @ 2024-01-25 14/week @ 2024-02-01 31/week @ 2024-02-08 52/week @ 2024-02-15

126 downloads per month
Used in 9 crates (3 directly)

MPL-2.0 license

20KB
526 lines

runa - wayland compositor toolbox

Table of contents

Demo

https://user-images.githubusercontent.com/366851/228325348-7b988f84-9837-4f5b-8fcd-d2085ea7edea.mp4

Project status, goals and plans

This project is still pretty much a "tech demo" in this stage. What you can see above is all that works. And it took a lot of short cuts to even get that far. (And please don't run cargo test.)

Creating a wayland compositor library from the ground up is a daunting task, and I don't think I can do this just by myself. That's why this project is announced in this state - because I wanted to attract interests, and potentially collaborators to this project.

I tried to document the code base in its current state as best as I can. Have a look at the documentation, or the code base, and get in touch if you are interested:

  • runa-core: base traits and types for a wayland compositor.
  • runa-orbiter: wayland protocol impls based on runa-core.

Chat room

I am currently hanging out in the #runa channel on libra.chat.

Roadmap

0.0.2

  • Have unittests for easily testable components.
  • All of the core and stable wayland interfaces implemented. No more "not implemented" error.
  • Fix interface versioning

0.0.3 - 0.0.4

  • Crescent able to run as DRM master
    • We likely want to create a crate for making accessing raw display devices and input devices easier. or use an existing crate. winit might become a viable option (rust-windowing/winit#2272)
  • Support hardware accelerated clients

0.1.0

  • Crescent can function as a minimal wayland compositor
  • A tutorial for writing compositors with runa

0.2.0

  • Xwayland support

Q&A

  • Why runa, instead of other wayland libraries like wlroots, or smithay?

    wlroots has a model that is not very suitable for a Rust binding - people has tried and failed at creating a binding for it. And even if I do manage to create a wlroots binding, it's not going to be ergonomic Rust.

    As for smithay, runa's design philosophy is pretty different. First of all, smithay was created quite a few years back, and Rust has evolved a lot in that time. runa tries to take advantage of new Rust features, most notably async, as much as possible. (this does unfortunately mean a nightly Rust compiler is required for now, at least until TAIT is stabilized).

    Also, runa took a page from other amazing Rust crates like smol and futures, and has a very modular design, unlike smithay which is everything inside a single crate. This way 1) users can bring their own implementation if what we provide is not suitable; 2) some of our crates can be useful outside runa.

Dependencies

~1.6–2.2MB
~41K SLoC