mas (mas769@hotmail.com) on 5/15/06 wrote:
>
>Don't know if you saw Tanenbaum's latest defence of
>microkernels but here it is if not.

I think he - and even more so Shapiro - kind of end up
discussing past the point.

Claiming that "embedded" OS's are secure, for example,
is laughable. The whole point of an embedded system is
that it doesn't even allow (much less support) the
huge majority of things that make security hard.

Shapiro (and to some degree Tanenbaum) also makes the
mistake of equating different address spaces with the
notion of modularity. They have nothing to do with each
other. You can be modular without using hardware to enforce
it, and you can generate a horribly messy system where
two processes are intimately aware of how each other works
even if they are separated by MMU boundaries.

As I already alluded to (and expanded on in my reply to
myself), I actually think using "designer languages" may
be a much better solution to the modularity problem than
the microkernel approach. The C model is very much one
where you have to do all the modularity by hand (which
Linux does, btw - don't get me wrong).

That said, I still think Tanenbaum's book on OS design
is one of the best ones around. So I'll happily disagree
with him, and I can still give the man credit for being
a big reason for getting involved and interested in UNIX
in the first place!

Linus