Tools: Problems with Excel 2007

[Peter Neilson]

Testing (as my audience already know) is best performed at the joints. 
Two to the 15th and to the 16th are both joints, as is two to the 31st 
and to the 32nd. Numbers in the vicinity of 32,768, 65,536, 
2,147,483,648, and 4,294,967,296 are thus all particularly interesting. 
Numbers near zero, one and two, and the negatives and reciprocals of 
those mentioned are also good joints.

Probably some tech writer originally found the problem, but was not 
believed. Oops, my excess cynicism is showing through. Down! Down!!

--pn


Geoff Hart wrote:
> Thought this might be of interest to early adopters who have migrated  
> to Office 2007:
> http://news.office-watch.com/t/n.aspx?a=556
> 
> My concern would be that if the process used to validate the  
> calculation and display engines used by Excel allowed this specific  
> error to creep in, there may be a more fundamental flaw waiting to be  
> discovered. Thus, calculate and view Excel 2007 reports cautiously  
> until the dust settles!