Maybe I misspoke but it seems the newsgroup gets fewer question on this
topic now. I had thought XL2003 had improved the "optimization" Was is
really as long ago as XL97! Sorry for the confusion.
best wishes
--
Bernard V Liengme
www.stfx.ca/people/bliengme
remove caps from email

"Jerry W. Lewis" wrote in message
...
Pardon my ignorance, but what changed in this regard with Excel 2003? I
am
only aware of the unfortunate (IMHO) fuzz factor introduced in Excel 97
(MS
calls it an "optimization")
http://support.microsoft.com/kb/78113
with the result that =a-b and =(a-b) are no longer guaranteed to give the
same result. It probably did avoid a few easy questions about "why isn't
my
result zero", but created a whole new genre of harder questions dealing
with
the fundamental (in)consistency of Excel's arithmetic.

Jerry

"Bernard Liengme" wrote:

Hi Jon,
Yes, I overlooked the mention of 2007! But as I said to OP, if is hard to
imagine a use of such small numbers and a y-axis transformation will work
even in XL2007. As you know better than I, MS have been trying hard to
avoid
the 'nasty' tiny numbers that should be zero but are not because of
'binary
round off errors'. We getting fewer questions about that topic with
XL2003
becoming more popular. So I cannot fault MS too much on this score!
Cheers
--
Bernard V Liengme
www.stfx.ca/people/bliengme
remove caps from email