From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.5-pre1 (2020-06-20) on
	ip-172-31-74-118.ec2.internal
X-Spam-Level: 
X-Spam-Status: No, score=0.5 required=3.0 tests=BAYES_05,TO_NO_BRKTS_PCNT
	autolearn=no autolearn_force=no version=3.4.5-pre1
Date: 17 Oct 92 20:10:50 GMT
From: wupost!m.cs.uiuc.edu!gillies@gumby.wisc.edu  (Don Gillies)
Subject: Re: What is "real-time"?
Message-ID: <1992Oct17.201050.22006@m.cs.uiuc.edu>
List-Id: <comp.lang.ada>

A real-time computation is a computation whose correctness is not only
a function of the logical correctness of the program itself, but also
a function of the time at which inputs are read and answers are
produced.  There are an infinite number of ways to specify timing
constraints on a computation; deadlines are just one of many ways.
For examples, here are some other ways:

	temporal constraints -- the computation values are not allowed
to become "stale", i.e. the input from a previous computation will
become stale and useless and must be acted upon by the next
computation before the values are stale.  Sensors typically provide
temporal data.

	distance constraints -- minumum or maximum distance between
two computations must be maintained.  These can minimize response time
in a process like a keyboard-polling process.  periodic constraints
are not an efficient way to do this.  For instance, if you want to
poll the keyboard every 50 milliseconds, you need a 25 millisecond
period (just in case a polling task executes at the start of period
#n, and the end of period #(n+1)).  But with distance constraints you
can specify a relative maximum separation of 50 milliseconds between
task 1 and task 2.  This can reduce system load by 50%.

	timing constraints -- these are the traditional types of
constraints, such as release-times, deadlines, and periods (for
--