This Month in Redox - December 2025
19 points by gnafuthegreat
19 points by gnafuthegreat
Congratulations to the great progress.
One thing for me is that It feels to me that all kernel eventually converge to something very UNIX/Linux like, in order to make software portable to them and have wider reach. I wish there was more true innovation for example pushing the Plan9 ideas further or leaning more into hurd-like microkernels. Kernels like Oxide's Hubris feel to me more interesting than another unix-like kernel.
Redox kernel implementation is a microkernel and is very unlike Linux. Plan 9 was an inspiration for Redox addressing schemes (I need to recheck if they are still there, they were removed/deemphasized, if I remember correctly). Redox chooses to provide a POSIX interface, but its implementation does not seem to be shaped by it.
There is also the social aspect: POSIX/Unix-like/Windows are also shared experience and shared vocabulary of their users, which would be just thrown away if innovating OS concepts from scratch. E.g. Fuchsia system interface design is stunningly clean and beautiful, but hardly anyone knows about it and hardly anybody programs on top of it outside of Google.
There's nothing wrong with innovative research designs, but the purpose of a non-research OS like Redox is to support and enable useful applications and hardware drivers. That's why at least some compatibility with an existing ecosystem is unavoidable.
In case its news to you: check out Genode/SculptOS! Its pretty cool and sufficiently different to traditional UNIX/Linux.