Hacker News Reader

Comment by Shared404 | Hacker News Reader

Is Redox something that might be a viable daily driver at some point? I've been watching for a little while, and hope it gets there.

Also, in what way's does the project need help?

Edit: FWIW, I think this or something similar to it is the future of OS development.

View on HN

At some point? Maybe. It would be in quite a long time. Right now writing drivers, porting software, and improving documentation are areas to help.

> At some point? Maybe. It would be in quite a long time.

I'll look forward to it.

> Right now writing drivers, porting software, and improving documentation are areas to help.

I'll take a look and see what I can help with.

How mature are the network stack and FS? I see a ZFS inspired FS named TFS, very cool! I’m assuming it can run headless but see it’s got a GUI from the book.

> FWIW, I think this or something similar to it is the future of OS development.

Disagree. It is mostly a Unix clone in rust with few novel concepts. Not that it is not important, and no offense meant to developers.

But why do you disagree? What does your comment bring to the discussion?

Edit: removed italics tag.

It is not a whole lot of new concepts as rust crowd likes to portray it.

Perhaps GP was referring to microkernel design as being the future of OS development?

This is more what I meant. It just seems like Redox is taking the lessons we've learned over years of development, and being willing to break compatibility in order to implement them.

That being said, I don't know much about OS/kernel development.

So is Google fuchsia - it is even capability based. (But Google is not cool here around HN circles, despite their contributions to open source and fostering one of good engineering environments, they aren't cool because they don't shout some Privacy buzzword like Apple.) Also it is not written in rust, which is webshit-favorite-of-the-month.

Redox is still Unix like, it moderately changes the file paradigm. Microkernel Unix clones are not new. They mention Minix as inspiration while there is no mention of L4 which did lot of work in microkernel performance.

Google's fuchsia C kernel isn't particularly interesting IMO. its essentially LK (little kernel) but with 100s of syscalls.

A couple of weekends ago I was playing with seL4 on RISC-V and Rust under qemu, and it felt much better balanced overall than my recollections of LK.

At least it is capability based instead of yet another Unix clone. Not that any one of these are significantly better. But the rust crowd heralds redox as very innovative OS.

What C kernel? It looks pretty much C++ to me.

https://fuchsia.googlesource.com/fuchsia/+/master/zircon/ker...

LK is C. It looks like they ported it to C++ for Fuchsia. https://github.com/littlekernel/lk/tree/master/kernel

Yeah, but Fuchsia is what OP was talking about and LK only existed in its original form during the early development of Fuchsia.

There used to exist a document about porting everything to C++, which I cannot find anymore.

Ah, the traditional future of OS development.

40 years in the making.

As opposed to Linux?

Who said Linux is a revolution in OS design? The big thing about linux was free availability and development model.

Reply on news.ycombinator.com