Well, I'd really like to get reimbursed for this.

If I post the bug here the chances of that drop to near zero.
I will say this about it though:
If you consider calculating the wrong sign a math bug,
then this is a math bug. In certain situations, that is the problem.

'Weekly claims of Excel Math bugs' that don't pan out,
wow, that's kind of amazing.

I think you'll all agree with me on this though, I've had 4 years of college
calculus etc, and the consultant has a degree in physics. And Mathematica
calculates the formula differently (the right way).



"Jerry W. Lewis" wrote in message
...
To expand on Ron's comment, basic math, exponential, log, and trig
functions are almost surely done by the CPU, rather than Excel
reinventing the wheel. It is extremely unlikely that there is a math
bug here, though on average there are 1-2 such mistaken claims per week
from people who don't understand the implications of finite precision
binary math (done by almost all software and hardware, not just Excel).

Another common problem is that text digits that look like a number in
input cells will either be ignored or cause an error, except for the few
cases where Excel's evaluation rules would coerce the text to a number.

Probability distribution functions (including ERF() and ERFC()), Bessel
functions and GAMMALN() were implemented in Excel using inferior
algorithms, but I don't immediately recall any other algorithm issues
that have not been improved in 2003.

If the previous paragraph does not cover your formula issues, then
rather than a "math bug", it is much more likely that either your
formula is numerically unstable or you have made some kind of mistake in
using it. We cannot help you diagnose this until you provide details.

Jerry

fred wrote:

I paid a consultant for a bunch of formulas
and built a spreadsheet with them.
The results didn't match those of the consultant.
I thought some formulas must have been keyed in wrong, so we
went over everything 15 times. Days later, that was ruled out.

Next, we suspected the consultants math.
After several consultant-hours we nailed it down to
a math bug in Excel.

After nearly a decade I can't believe it still exists.
I tried it in Excel 2002 and 2003 -- both have the bug.
It's not an everyday thing,
but it's not rare or esoteric either.
When consultant found it, he was shocked and said
they "should be sued".
It's not arguable -- it's wrong, it's a bug.

Since it cost me $2000 to find it, I'd like to know
if Microsoft has any kind of pay-for-bugs programs.
Or maybe magazines that would buy this info.

Any ideas for how I can get reimbursed?

Fred