🚀 go-pugleaf

Christopher C. Wood (christw@meaddata.com) wrote:
: In article <30ECC158.5289@osu.edu>, Jeff Jasper <jasper.7@osu.edu> writes:
: |> > I see.  A brand-new fully preemptive, protected-memory microkernal
: |> > operating system backward-compatible with todays' applications is
: |> > "little new technology".  Whatever you say, cap'n.

: |> Gee, kinda sounds like Windows NT.  Copeland isn't fully preemtive, 

: Copland (no 'e' -- its namesake was Aaron, not Stewart) is fully
: preemptive, ...

With regard to not fully preemptive, I think the previous poster was
referring to Copland and its applications as a whole, not merely the
kernel buried deep inside the OS. 

: ... in that nothing short of a kernel crash will circumvent
: the preemptive tasks from executing.  Legacy applications can only
: block other legacy applications (and even in the event of a
: misbehaving legacy application locking up the user interface, the user
: can force the blue-box to restart, and Copland-savvy applications will
: continue on as if nothing ever went wrong)...

Good, but even Copland-savvy applications are limitted in how they can 
use preemptive threads. This is why I believe some people argue Copland 
is not truly preemptive. 

: Unlike WinNT, Copland won't be a resource pig.

Of course not, 16M PPC systems will be standard by the time it arrives. 
NT was released when standard systems were much more modest, hence it's 
reputation. If NT were introduced at the later date Copland is arriving 
at it wouldn't be considered a "pig" either.

Tony
-- 
------------------
Tony Tribelli
adtribelli@acm.org