Yes, this is a long running "feature" of UniData. With a simple statement you can change the results of many of your calculations. Which setting should you use? Well that depends on your data and your calculations. Look at the PRECISION and FLOAT.PRECISION statements (if you don't believe that IBM realizes that this is a problem but doesn't know how to fix it - at least without "breaking" existing apps - look at the help for float.precision).
It's always seemed like the same issue Intel had with the original Pentiums - that they had to recall. Perhaps they didn't have the "spin-masters" that IBM has - or fewer customers willing to put up with this type of crap. Your best bet is to avoid the issue by scaling your numbers so you don't do calculations with decimal values. hth Colin Alfke Calgary, Canada > From: m.hutchinson > > Hi All, > > We have Unidata 7.2 running on Windows 2008 server making incorrect > calculations when numbers are close to zero. I understand there is a > 'WIDEZERO' parameter on Univers but I don't think there is an equivalent > on Unidata. Below we have an example program and output. > > When run this gives: > > BAL [0] > BAL > 0 > Does not match integer > MCN BAL <= 0 > > Has anyone come across this one or have any ideas? > > regards > > Martin Hutchinson _________________________________________________________________ Internet explorer 8 lets you browse the web faster. http://go.microsoft.com/?linkid=9655582 ------- u2-users mailing list [email protected] To unsubscribe please visit http://listserver.u2ug.org/
