Microsoft Patches Excel 2007 Multiplication Bug

Red-faced Microsoft officials say they have fixed a problem that causes the latest version of the company's Excel spreadsheet program to produce significant multiplication errors.

"We are in the process of adding this fix to Microsoft Update so that it will get automatically pushed to users running Excel 2007 or Excel Services 2007," the officials wrote Tuesday on an Excel developer team blog.

The fix is available as a download, and also will be included in the forthcoming service pack 1 for Office 2007 -- for which a release date has not been set, the officials said.

In a posting last month, Excel developers said the flaw occurs during calculations that would ordinarily result in, or be close to, the number 65,535. Instead, Excel produces a result of 100,000.

The Excel team members said the flaw is the result of a floating point error -- the same sort of problem that infamously plagued Intel's Pentium chips in 1994. A floating point error occurs when a computer -- or a human -- places the decimal point in a string of numbers in the wrong spot.

Excel can store millions of floating point numbers. Twelve of them are capable of causing the error if the software isn't patched, according to the Excel developer blog.

The 16-digit numbers can't be entered directly into Excel because the program will round them off to its maximum, 15-digit display. But the error will occur if the numbers are the product of a calculation.

Microsoft's Excel problems come at a time when its Office franchise is under threat from free and low-cost alternatives offered by well-heeled rivals like Google and IBM. IBM recently said its free Lotus Symphony suite -- which includes a spreadsheet -- garnered more than 100,000 downloads during its first week of availability.