Newsgroups: comp.realtime,comp.os.qnx,comp.os.ms-windows.advocacy,comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!uunet!newsflash.concordia.ca!canopus.cc.umanitoba.ca!mona.muug.mb.ca!rgallen
From: rgallen@muug.mb.ca (Rennie Allen)
Subject: Re: Real-time systems: Windows-NT or QNX
Message-ID: <1994Oct17.230930.6848@muug.mb.ca>
Organization: Manitoba Unix User Group, Winnipeg, Manitoba, Canada
References: <37q48e$mog@ixnews1.ix.netcom.com> <DERWAY.94Oct17100210@alumni.ndc.com>
Date: Mon, 17 Oct 1994 23:09:30 GMT
Lines: 23
Xref: glinda.oz.cs.cmu.edu comp.realtime:7204 comp.os.qnx:2259 comp.os.ms-windows.advocacy:40707 comp.robotics:14476

In <DERWAY.94Oct17100210@alumni.ndc.com> derway@ndc.com (D. Erway) writes:


>This is absurd.  What do we mean when we say "This OS has real time support"?
>I believe the readers of this news group expect guaranteed worst case
>interrupt latency, and deterministic system calls, and rescheduling.

>If either OS/2 or NT had this, you can bet we'd all know it.

>Real time does not mean real fast.

I beg to differ with this statement.  True; the *definition* of real-time does
not mean real fast, but actual real world requirements of real-time typically
do.  How many real-time apps have you seen lately where worst case time is
62 minutes.  Given this constraint I would say that just about any computer
system is deterministic (except maybe the AS/400 :-).  People working in the
real world who do real-time also need speed.  I would accept the statement
"real time does not *only* mean real fast"

email: rgallen@muug.mb.ca
QUICS: rgallen (613) 591-0934
Voice: (204) 367-2311       
Fax:   (204) 367-4407
