News for hackers. Stuff that matters.
Re:The Fiasco microkernel by microcolonel on Thursday October 11, 2018 @02:41PM (#18195997)
I think there is a lot of promise in pulling mature components from Linux, OpenBSD, and Solaris/Illumos (or whoever does SunOS these days) across that paravirtual barrier, and progressively tailoring them toward the microkernel environment; while testing them using the systems they originate from.

Over time, maybe a microkernel could become the basis of most of these former monoliths.

Re:The Fiasco microkernel by nickpsecurity on Saturday October 13, 2018 @12:18PM (#18208685)
The Dresden and Genode folks do it a bit different: just virtaulize the OS as a user-mode application. Then, some add ability to reuse device drivers across VM's to get the drivers specific OS's (esp Linux) support. From there, you can slowly pull code out as standalone or native applications as you described. There's usually middleware (eg Camkes) that lets the standalone code talk to code inside the VM. I'll give Genode as an example of mixing native and VM code.

https://genode.org/documentation/general-overview/index