Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 exptools 1/6/84; site ihnss.UUCP
Path: utzoo!watmath!clyde!cbosgd!cbdkc1!desoto!cord!hudson!ihnp1!ihnp4!ihnss!knudsen
From: knudsen@ihnss.UUCP
Newsgroups: net.micro.6809
Subject: More on C floating point vs BASIC09
Message-ID: <2324@ihnss.UUCP>
Date: Mon, 3-Dec-84 19:16:10 EST
Article-I.D.: ihnss.2324
Posted: Mon Dec  3 19:16:10 1984
Date-Received: Tue, 4-Dec-84 19:47:15 EST
Organization: AT&T Bell Labs, Naperville, IL
Lines: 14

((( mixed marriage ]]]  but well balanced...

My apologies -- I rooted around in the Basic09 manual this weekend and yes,jej,
B9 uses 40-bit floats just like RS Basic.  That makes C look even slower.
Also, I tried to speed up the C programs by redeclaring all the floats to
doubles, so as to avoid the conversions.  Absolutely no visible change.
	But I think you other guys on the net have pinpointed the reason--
C insists on doing everything in 64 bits, needed or not.  That could
quadruple the mult and divide times!  Yes, the little white bible sez it's
all done in double -- I had mistakenly thought that only function arguments
were expanded to doubles.
	Seems awful stupid, but I guess C saves a little library space by
not having to maintain 32-bit operators.  Hurts to see something this wrong
in my favorite language ... mike k