microkernels

Thomas M. Farrelly s720@ii.uib.no
Fri, 10 Sep 1999 19:35:01 +0200


Laurent Martelli wrote:
> 
> >>>>> "Beholder" == Beholder  <beholder@bespin.dhs.org> writes:
> 
>   Beholder> Francois-Rene Rideau wrote:
> 
>   >> I do agree that the most important point is to have the right
>   >> primitives; and that's perhaps what I'd like a no-kernel to be
>   >> about: focus the design attention on the semantics of primitives,
>   >> rather than on particular implementation mechanisms, that should
>   >> be hardware-dependent (which is I think a positive contribution
>   >> of exo/nano-kernel approaches, independently from their kernelful
>   >> aspect).
> 
>   Beholder> What primitives do you suggest?
> 
> I'd suggest something very close to Lisp or Scheme : objects,
> functions, you can call functions with objects as arguments and get a
> a result, you can define new functions as calls to already defined
> functions (lambda). The state of the system should be persistent : if
> you shutdown the system, and restart it, you should have the same
> state sa before the shutdown. That's all you need I think. (plus a
> number of "hardcoded" standard functions such as and, or, if, =, cons,
> car, cdr).
> 

Primitives in the sence of "hardcoded" mess up reflection - it's not
possible to alter their definition. And primitives are not strictly
nessacary anyway ( see TOOL ) .

So, loose the primitives.

 
===============================================================================
    Thomas M.  Farrelly     s720@ii.uib.no     www.lstud.ii.uib.no/~s720
===============================================================================