Jason Fesler wrote in a message to Mike Bilow:


 MB> In theory, this should never happen.  The keyboard generates a
 MB> hardware interrupt, IRQ1.  The system should service that immediately

 JF>  May be true, but that type of lockup has happened here... 
 JF> The only external switch I had to break out of it,
 JF> unfortunately, was conneted to the "Reset" pins. ;-).

That, Jason, is one interrupt that can't be ignored... ;')

 JF>  Some windows were actually still processing data [those I
 JF> set as 'critical']; I could still see the modem activity
 JF> happening.  I just waited until the current call was done,
 JF> then hit that switch.. 

Jason. I've been there and done that... I'm amazed at those situations,
when they arise, that hose the system so completely and yet all (if only a
handful) of the background tasks still run.

If nothing else, this speaks well of OS/2's crash protection... the OS
dies, but the sessions still run... :')

Jerry McBride

--- timEd/2 1.01+