EVERHART%ARISIA.DECnet@GE-CRD.ARPA (08/06/88)
After seeing Carl Lydick's message, I guess I should have realized that setting terminals /slaved might be misunderstood. A terminal set /slaved on RSX is not slaved TO anything. It just is a device, whose characteristics and ownership are a separate issue. Thus, a terminal can be available to the world, yet have typeahead and the like enabled. This is what VMS can't do in this fashion. BTW a terminal can be logged in and slaved also, and even run tasks but with limited ability of users to break out of command files. Allocating a terminal under VMS is not equivalent, since there has to be someone to allocate it TO, and nobody else (w/o privs at least) can get to it. Setting it /notypeahead is also inequivalent, since typeahead for that terminal is no longer possible. The /notypeahead solution is frequently used, but only because that's the best one can do. I find it continually surprising that the /slaved characteristic was left out of VMS terminals. /slaved means that loginout will NOT be run on a terminal when input is detected. In RSX it also restricts special interpretation of control C, but that is already controllable in VMS. Problem is, there doesn't exist currently a mechanism in VMS that JUST tells VMS not to fire up loginout on a terminal without screwing something else up to do it. Once we remember that terminal ports get used for lots of things other than terminals, this becomes important, and I'd dearly love to see it. Glenn Everhart%Arisia.decnet@ge-crd.arpa