Hi Robert!

[..]
 RK>  Been there, done that.. Screen writes use memory, and hence can be
 RK> stopped  by the OS, direct port I/O is another matter entirely. 
Nope. Read some literature about the design of the 386. OS/2 uses all of
the offered features.

 RK> The point was and is, though  virus attacks have not been a big problem under
 RK> OS/2, the key point is they  haven't been YET. 
Unless you bring some evidence, you can stop rumouring about this topic -
your credibility regarding this topic is slowly sinking. I know fairly well
(PF knows surely _very_ well) about possible security leaks in the OS/2
system and the big problems virus authors would have when programming a
virus for the OS/2 system.
However, I agree to you in principle. I claim as well that there is a
possibility to write a virus for OS/2, but compared to DOS OS/2's
architecture is vastly more effective than every existing virus scanner.

It is (almost) impossible to
a) bring "unknown" code into execution -> no hidden processes
are possible
b) infect other systems, "growing virus population"
c) hide the origin of possible destruction (means the virus will always be
easy to identify)

Without extreme efforts it is impossible to write an advanced,
"effective" virus for OS/2 !
     __
     |_)
cu/2 |atrick

--- FleetStreet 0.98wb