Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site bbncca.ARPA
Path: utzoo!linus!vaxine!wjh12!bbncca!sdyer
From: sdyer@bbncca.ARPA (Steve Dyer)
Newsgroups: net.unix-wizards
Subject: Re: Real-time needed; FTIME buggy. - (nf)
Message-ID: <150@bbncca.ARPA>
Date: Thu, 6-Oct-83 14:39:00 EDT
Article-I.D.: bbncca.150
Posted: Thu Oct  6 14:39:00 1983
Date-Received: Sun, 9-Oct-83 17:39:12 EDT
References: <131@hscfvax.UUCP>
Organization: Bolt, Beranek and Newman, Cambridge, Ma.
Lines: 11

Ftime's concept of milliseconds, at least on machines without millisecond
clocks, is incremented in values of HZ/1000, where HZ is usually 50 of 60.
You probably want to see about getting the "nap" device driver/system call
if your applications really require <1 sec sleeps.  I don't have it myself,
but others might give you pointers.

/Steve
decvax!genrad!wjh12!bbncca!sdyer
...!ima!bbncca!sdyer

P.S.: Looping in ftime() is pretty antisocial!