Re: A short note on microkernels

[Jonathan S. Shapiro]

On Tue, 2006-04-25 at 00:19 +0200, Pierre THIERRY wrote:
> Scribit Jonathan S. Shapiro dies 22/04/2006 hora 11:11:
> > 1. There is a desire for pinning, but no workable design exists.
> 
> How do other systems deal with what we need pinning for without pinning,
> or why is their design of pinning not usable?

No design exists that deals reasonably with sharing and persistence.

> > 2. No particular scheduler has been selected. The current plan in
> > Coyotos is to continue with the Mercer/Tokuda capacity reserve
> > scheduler by default.
> 
> Won't the scheduler be selectable? At least at the compiling stage.

Yes, but this avoids the question, which is: what will the choice(s) be?

shap