[Date Prev][Date Next][Subject Prev][Subject Next][ Date Index][ Subject Index]

Re: Re Keystrokes: Possible BIOS compatibility (OT)



--- Robert Holmgren  wrote:

>> operating systems ... take *active measures* to
> preclude
> > an app being able to [crash the machine]. That's
> part of the
> > point of their design.

> You're right, but what can happen (does happen) is
> that an old DOS app grabs
> nearly 100% of the CPU and then that app locks up
> (or vice versa). There just
> aren't enough spare CPU cycles to break out (so that
> you can kill the app).

I think you've filled in the missing 'Why' of what I
had observed. In that situation, I don't suppose it
would matter what utility one attempted to use. The
point I was trying to make to Patricia was that OS
design had evolved to the point where the OS could
"say" to an app 'No, you can't go there,' or 'No, you
can't do that,' if it was attempting the sorts of
things that would formerly kill DOS or W98. In that
case (much more common than the overloaded CPU
scenario, I think), the app dies, but the OS keeps on
trucking. I don't know what the OS designers might be
able to do about the latter problem, but perhaps the
dual-core CPUs coming onto the market will have some
bearing on this. I also wonder whether running such
an app inside a virtual machine might tend to "wall
off" this effect.

In regard to the one example I cited with OS/2, that
was really a weird sort of bank shot. An app I was
trying out at the time -- a port of the Opera browser
-- could kill my video driver *if* I used certain
browser features, and bring the OS down indirectly.
Changing video drivers put a stop to that, but I
tossed Opera after that, just on general principle.


Jordan