> I can't see much point. It's a fair amount of work for very little
> gain.  (You're not going to save more than a few K of memory doing
> this, which is not a big win in a world where 16MB of memory is
> less than a hundred dollars.) I'd rather see the time put into things
> that will benefit a greater number of NetBSD users, such as merged
> VM/buffer cache.

Absolutely.  I understand that there are machines with 4 MB RAM or even
less, but even then, making the kernel a few KB's smaller isn't a big
win.  And if you really need a minimal kernel, you can config and build
one.  I would love to see the merged VM/buffer cache, more drivers for
all archs, support for multiple CPU's, etc.  And yeah, all the known
problems/bugs fixed. :-)

  -jm


-- 

                     ---> http://www.jmp.fi/~jmarin/ <---