Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site abnjh.UUCP
Path: utzoo!watmath!clyde!floyd!whuxle!spuxll!abnjh!ncoss
From: ncoss@abnjh.UUCP (Scott D. Allen)
Newsgroups: net.micro.apple
Subject: Benchmarking Microsoft vs Macintosh BASIC
Message-ID: <532@abnjh.UUCP>
Date: Wed, 11-Apr-84 08:46:45 EST
Article-I.D.: abnjh.532
Posted: Wed Apr 11 08:46:45 1984
Date-Received: Fri, 13-Apr-84 06:09:13 EST
Organization: AT&T Information Systems
Lines: 32

[Into the bit bucket]

	The latest issue of BYTE magazine, April 1984, has a preview article
of Macintosh BASIC, which I assume is now called MacBASIC.  The article was
very good in that it desrcibed the general operation of Macintosh BASIC.  The
idea of a semi-compiled language is very interesting.  Anyway, the author also
included some benchmark programs and times for this version of BASIC.  I put
the same programs through MICROSOFT BASIC on the Mac and here are the results.


TEST			Macintosh BASIC		Microsoft BASIC
			(seconds)		(seconds)

Empty GOSUBs		3.0			8.0
Empty FOR/NEXT loop	1.5			5.0
Midstring search	9.0		    	22.0
Real Division(by 3)	18.0			32.0
Integer Division(by 3)	3.0			xxx  *
Eratosthenes Sieve	31.5 **			1 min. 45.0 sec **


All times for Microsoft BASIC were done using Mac's clock which only gives time
in whole seconds.

*	I could not find an equivalent function in Microsoft for Integer 
	functions.

**	Macintosh BASIC used short variables for the Sieve and found 1899 prime
	numbers in 8191 numbers.  Using 2-byte integers (as Microsoft calls 
	them), I could only DIM an array for a maximum of around 6500.  That's
	the time presented.  By using regular integers, I could only use around
	a 1000 element array.