From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=0.7 required=5.0 tests=BAYES_00,INVALID_DATE,
	REPLYTO_WITHOUT_TO_CC autolearn=no autolearn_force=no version=3.4.4
Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!lll-crg!ames!ucbcad!ucbvax!nosc-tecr.UUCP!contr47
From: contr47@nosc-tecr.UUCP ("CONTR47")
Newsgroups: comp.lang.ada
Subject: timer question
Message-ID: <8612160033.AA15076@ucbvax.Berkeley.EDU>
Date: Mon, 15-Dec-86 10:55:00 EST
Article-I.D.: ucbvax.8612160033.AA15076
Posted: Mon Dec 15 10:55:00 1986
Date-Received: Tue, 16-Dec-86 22:56:11 EST
Sender: daemon@ucbvax.BERKELEY.EDU
Reply-To: "CONTR47" <contr47@nosc-tecr>
Organization: The ARPA Internet
List-Id: <comp.lang.ada>

Which of the following 2 ways do you find most useful
to specify a periodic interrupt?
1. Specify the time interval between interrupts.
2. Specify the interrupt rate per second.
   Before answering consider some examples:
a. Time interval resolution is 10 microseconds and
you want 64 interrupts per second thus you desire
an interrupt every 15_625 microseconds but cannot
specify the last digit so you either live
with the inaccuracy or if accumulated error is a problem
you must insert a leap interval to make accumulated
time work out in the long run average.
b. Interrupt rate resolution is inverse powers of 2
thus you can have precisily 64 interrupts per second
but interval timers are now not nice decimal
numbers. (reminds one of model number maybe).
     Is there a need for precise rates (e.g: 64 per
second) for mathmatical reasons? Is this need
strong enough to suffer the model number kind of
problem with interval timers (delays)?
    If you really can't choose but must have it
both ways please say so. It adds more complexity
to the runtime environment but if it's required
then it's required.
regards, sam harbaugh
---------------------
------