Skip to content(if available)orjump to list(if available)

Show HN: An Open Source XR(AR/VR) Operating System

Show HN: An Open Source XR(AR/VR) Operating System

7 comments

·September 7, 2025

We're two college students building an XR(AR/VR) native Operating System with a custom kernel. We're also Open Source so feel free to check our GitHub Repository- https://github.com/manaskamal/XenevaOS .

The journey hasn't exactly been easy, we've been criticized by a lot saying that whatever we're doing is impractical and that we're too ambitious. Regardless, we've been committed to reach our goal.

Here to answer all questions and doubts. Answering one question beforehand because we know someone is going to ask it -

Q: Why use your own kernel/ Why don't you use Linux/ Why are you trying to reinvent the wheel?

A: Using our own kernel helps us get rid of the baggage of legacy codes, bring the most optimal performance on our target hardware (XR/AR/VR) and achieve more efficiency than what we would've achieved on an existing kernel.

We're not trying to reinvent the wheel, but just building Formula One racing tyres for it.

wkat4242

Very interesting. I do have a bit of a "vids or it didn't happen" feeling about this.

Cool idea to use your own kernel though it does sound like you could find yourself in perpetual development hell. And, don't forget all the sufficiently powerful SoCs are super closed. You won't be able to leverage any of their existing driver work and you will need some serious clout to get access to their documentation, with some really scary NDAs attached. However I'm sure you know this and took it into account. Very cool. I hope you will manage to get it to market!

onethumb

Related discussion:

John Carmack's arguments against building a custom XR OS at Meta

https://news.ycombinator.com/item?id=45066395 (11 days ago; 527+ points; 646+ comments)

Love to hear whether you agree or not and how your project is different?

potatolicious

> "A: Using our own kernel helps us get rid of the baggage of legacy codes, bring the most optimal performance on our target hardware (XR/AR/VR) and achieve more efficiency than what we would've achieved on an existing kernel."

This is kind of a non-answer, no? What baggage does it get rid of? What kind of performance optimization does it bring that cannot be fulfilled with an existing OS/kernel?

janice1999

Exactly this. The kernel seems like the least interesting part of an AR system. Also targeting x86, ARM and RISC-V for a new kernel is such a huge workload it makes no sense not to just re-use something already existing.

password4321

Related recent discussion:

MentraOS – open-source Smart glasses OS

https://news.ycombinator.com/item?id=45140381 (4 days ago; 200+ points, 120+ comments)

verdverm

Qualcomm is supposedly telling developers to target AndroidXR instead of Spaces

Is this building on that or a complete bottoms up writing of the full AR software stack?

adfm

What hardware is currently supported?

What open standards does it support? OpenXR, WebGPU, WebXR?

What industries are best suited? Games/Entertainment/ Sports? AEC?