Speaking of unreleased code, the fact that there is any makes me
worried that I may have to fork seL4 someday just because of the
rather closed development process, which is a mismatch for the fully
open, copyright-assignment-free development process (akin to that of
the Linux kernel) that UX/RT will have. UX/RT is meant to be an OS for
the real world and not yet another systems research project that is
completely ignored outside academia. Therefore I am going to do
everything I can to give it the best chance at real-world success,
which includes a development model with as few barriers to
contributing as possible.

The experiments Gerwin refers to predate the open-sourcing of seL4.

However, there is always going to be some in-house experimental code, especially student work, that never makes it out. 

We’re doing our best to make things more open, but we have very limited resources, resulting at times in a divergence between intent and reality. The forthcoming seL4 Foundation should help.

Forking is a very bad idea. If you modify the kernel, you’re breaking verification, and it’s no longer seL4. And it will be one job of the Foundation to ensure that only things that *are* seL4 are called seL4.

Gernot