managarm VS DataFrame

managarm is pretty clean from what I remember.

Somewhat related, Managarm[0] also attempts to be Linux compatible, but has a microkernel, multiserver design.

0. https://github.com/managarm/managarm

Neat idea. There is some research done on fully async OSes like https://github.com/managarm/managarm

If you haven't already, I'd hang out with the Managarm crew. Prety friendly, and what they're doing with C++ and async is very cool!

But well, maybe you should look at Managarm.

You can find more information on the project on our github page here. It is mostly a hobby project, so me and the other developers like to work on it in our free time.

Not hosted on rust-osdev but the Managarm project (an OS that focuses on async and that is primarily written in C++ so far) recently gained a rustc port and ports of some basic utilities written in Rust (ripgrep, exa). Support for Rust drivers is in the works (but still requires adding support for some libraries first that wrap the low-level system call interface).

If you have never seen an async OS, then you must’ve missed Managarm, a fully asynchronous operating system written in c++20. It’s a cool thing to check out, runs on x86_64 (a port to a raspberry pi4 is in progress) and has an extensive userland with various Linux applications running on a custom written libc. In case you’re curious, the repo can be found here: https://github.com/managarm/managarm

For a while, I have been hearing that Polars is so frighteningly fast that you shouldn’t look directly at it with unprotected eyes. So, I finally found time to learn a bit about Polars and write a very simple test/comparison for C++ DataFrame vs. Polars.

I have worked on C++ DataFrame for the past 5+ years in my spare times. It is comparable to Pandas or R data.frame, although it includes a lot more functionality.