Here I Am

From: Brian Chase <vaxzilla_at_jarai.org>
Date: Wed Jan 23 09:54:16 2002

On Wed, 23 Jan 2002, emanuel stiebler wrote:
> Jay West wrote:

> > There is something to be said for not just trying to
> > run on everything under the sun - and that directly translates into
> > stability of the OS.
>
> So, suddenly windows runs stable, just because they support x86 only ?

Yeah, the assertion that porting to a wider variety of platforms
decreases overall stability isn't necessarily correct. One of the goals
of NetBSD is to run on as many platforms as possible. What I've learned
through my involvement with NetBSD over the past few years is that this
wide support of platforms is now only secondary to a more significant
/root/ philosophy of the OS development efforts, and that is the focus
on the design of highly portable code.

The end result is that the various kernel subsystems are clean and
extremely well abstracted-- especially in comparison to the competition.
Code reuse across the platforms is higher than I've seen anywhere else.
It's still not perfect, but it strives to be as point of principle.

The process of throwing as much different hardware at NetBSD as possible
also means that the OS gets exercised in significantly more ways than
any other system out there. This is a great way to find portability and
design flaws quickly. Additionally, the constraints you face in trying
to create a middle ground between all the systems, with a relatively
small group of people, forces you to use good software development
practices.

It's a huge task to span from the support of 0.9 MIPS MicroVAX-IIs to
that of the fastest Alpha and Pentium processors. The NetBSD code has
evolved to deal with damn near anything. I think taking this long-view
will serve to the advantage of NetBSD in the future.

-brian.
Received on Wed Jan 23 2002 - 09:54:16 GMT