My take on (GNU/)?Linux Re: [PLUG] Another topic: the PLUG FAQ

Michael C. Robinson michael at robinson-west.com
Fri Nov 21 17:11:01 UTC 2003


On Fri, 2003-11-21 at 16:41, Ed Sawicki wrote:
> On Fri, 2003-11-21 at 14:29, Michael Rasmussen wrote:
> > On Fri, Nov 21, 2003 at 12:59:28PM -0800, Ed Sawicki wrote:
> > > > Linus did what
> > > > Richard was unable to do: create the kernel.
> > > 
> > > Apples and oranges. Stallman wanted a microkernel - a more
> > > difficult thing to do and an area where there are fewer
> > > people capable of helping out.
> > 
> > Not apples and oranges.  Part of system architecture is choosing
> > what works.
> 
> Another part of system architecture involves meeting design goals.
> It may not matter if it works if it doesn't achieve the goals. I'm
> not privvy to Stallman's goal back then but suppose he wanted the
> reliability that well-written microkernels ensure and monoliths
> like Linux cannot?
> 
> Ed

Can smoeone elaborate and what makes it difficult to do an
architecturally pure monolithic kernel verses the micro kernel
alternative?  I think I've heard that Linus likes the speed
of a monolith over a micro kernel.  I was looking at a book
by Randolph Bentson the other night on Linux that says a true
micro kernel may be able to have interfaces for other operating
systems, if I read it right.  Threading is apparently a big deal 
now.  I really don't know what it means for computer architecture 
and computer programming.  I seemed like people got too excited
about hyper threading on the Pentium 4 in my cs200 Nix class
along with let's talk about XP instead of Linux.  Really 
irritating.  I didn't learn anything meaningful about threading
in that class other than perhaps to ask about it ;-)

-- 
-----------------------------
Michael C. Robinson

Associates of General Studies 
PCC March 2003.
-----------------------------




More information about the PLUG mailing list