3 ms·
What, beyond the device drivers, even is so valuable in the Linux kernel that we would want to add Rust to it rather than develop a new kernel from scratch in R
by qwerty456127 5y ago
What, beyond the device drivers, even is so valuable in the Linux kernel that we would want to add Rust to it rather than develop a new kernel from scratch in Rust?
- philipswood 5y agoIs that a hint of Pythonic hubris?
- qwerty456127 5y agoIt's the decades of perception of the Linux kernel as an insanely bloated pile of legacy code in a legacy language (I would rather go modern C++ the way SerenityOS did, not necessarily even Rust) + the driver problem (new OSes have little option besides borrowing what little there is in NetBSD which seemingly has the best driver model) + optimism caused by numerous alternative OSes projects + plus some functional programmer hubris. I have tried quite a number of alternative OSes and the only thing I can see a problem with are hardware drivers.
- ilaksh 5y agoMaybe a new OS project with a focus on directly funding things like driver ports. Funnelling donations to something like Ethlance.
- dboreham 5y agoBecause at best that new kernel would be as popular as BSD?
- qwerty456127 5y agoHm. Okay. Why not? I would even prefer a world of less popular kernels to one single kernel everybody uses. In fact some "default Linux" with systemd et. al. becoming too popular already seems worrying.
- ezekiel68 5y agoTo be fair -- there was no evidence anyone should have thought any different about Linus's little minix rip-off project back in 1991.