Path: number1.nntp.dca.giganews.com!border2.nntp.ams.giganews.com!nntp.giganews.com!zen.net.uk!dedekind.zen.co.uk!newsfeed.tpinternet.pl!newsfeed.atman.pl!news.intercom.pl!f124.n480!f127.n480!f112.n480!f200.n2432!f605.n774!f500.n123!f2000.n106!f45.n396!f1.n379!f133.n167!not-for-mail
Newsgroups: fido.science
Distribution: fido
From: MIKE ROSS 
Date: Tue, 10 Aug 04 22:30:16 +0200
Subject: Pythagorean triples
Message-ID: 
References: 
Organization: Juxtaposition BBS, Telnet:juxtaposition.dynip.com
 111
 300
Lines: 28
Xref: number1.nntp.dca.giganews.com fido.science:295

"Jasen Betts" bravely wrote to "Miles Maxted" (09 Aug 04  07:26:21)
 --- on the heady topic of "Pythagorean triples"

 JB> Hello Miles.
 JB> more precision that that is dificult in interpreted basic, but in C I
 JB> have timer code accurate to better than 1 microsecond (by reading the
 JB> registers of the timer chip).

I was reviewing the procedure to read the timer chip registers and there
was a warning about directly reading the timer ports will upset the
count. Apparently the counter must be latched first before it is read.

Have you ever experimented with timing the video sync signals available
at the crtc ports? Those signals go low during the blanking and depending
on the rates can be pretty fine resolution. Not 1us but much better than
18.2ms.

Still another timing method is to use the music in background routines
and setting the tempo value for the finest time slicing. The limit is
around 200 beats per second using 1/32 notes, IIRC. That is a timing
period of about 1/64000, which is pretty good too.

The simplest delay is to use ON TIMER(1) and measure the average loop
count. Then use this value to time a programmed delay.

 M*i*k*e

... On a clear night you can see 1.72*10^19 miles.