[unios] Re: As I see it . . .

[Pieter Dumon]

From: Pieter Dumon <Pieter.Dumon@rug.ac.be>

> 
> From: OJ Hickman <hickman1@peak.org>
> 
> Does the kernel matter?
> 
> As long as the programming APIs are consistent what does the underlying
> kernel matter?
> I mean, for example, Linux and the Herd. They are both UNIX systems are
> they not? Can't a
> properly written C program be compiled to both? But Linux is monolithic
> and Herd
> microkernel. The APIs, you see, are the important issue.
>

It's Hurd, with a U.
But Hurd is no unix! It uses the first-generation microkernel mach. This
makes Hurd old before it is even written :(
But Hurd puts a Unix server on top of mach, providing a look-a-like POSIX
system on mach.  

> 
> No kernel?
> 
> I think this idea of breaking the kernel into swapable components is
> worth farther
> consideration. Change the basic concept of what a kernel is.
> 

I think so too. The basic thing you need is some layer on top of the
hardware wich rules the hardware. Even this layer (HAL) should be
run-time changeable. On top of this HAL, you can implement seperate
components like scheduler etc.
But, I think, code size will grow. You will need so much code to implement
this component system, that it is far easier to develop a good, tightly
coded microkernel.  

Pieter

----------------------------------------
 Pieter.Dumon@rug.ac.be               
                                      
 http://studwww.rug.ac.be/~pdumon     
 
 ICQ  : 12428974
---------------------------------------

------------------------------------------------------------------------
Help support ONElist, while generating interest in your product or
service. ONElist has a variety of advertising packages. Visit
http://www.onelist.com/advert.html for more information.
------------------------------------------------------------------------
UniOS Group
http://members.xoom.com/unios