4 ms·
> Linus has made it incredibly damn clear what "no user space regressions ever" mean out of curiosity, how does a policy like this translate to microkernel OSe
by leeoniya 9y ago
> Linus has made it incredibly damn clear what "no user space regressions ever" mean
out of curiosity, how does a policy like this translate to microkernel OSes (e.g. Redox) where most everything runs in userspace?
- olewhalehunter 9y agoshhh just get out the popcorn and enjoy the rantings of an egomaniac who wrote a glorified DOS extender for an outdated time-sharing system
- sctb 9y agoLet's not join in with rantings of our own? https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- city41 9y agoDoes it need to? This is the rule for Linux development. Other kernels need not follow it if it doesn’t make sense.
- ams6110 9y agoYes, exactly. Linus does things his way. Others (OpenBSD and maybe other BSDs for example) view the kernel + userland as a complete release and you upgrade everything together.
- tonysdg 9y agoI've never programmed a microkernel OS (and I've only done minimal Linux kernel programming), but I would assume it's even _more_ important, as a single change would impact almost every component of the system.
- invokestatic 9y agoMicrokernels still need syscalls for usermode to interface with, so it is just as important. Kernels exist to serve userspace, not the other way around. Though, given the relative immaturity of Redox I assume this is less important compared to Linux.
- leeoniya 9y agoi think Redox explicitly tries to limit the syscall exposure as well relative to linux. (dozens vs hundreds) https://doc.redox-os.org/book/introduction/how_redox_compares_to_other_operating_systems.html https://doc.redox-os.org/book/introduction/how_redox_compare...