Using Excel 2007 may not add up

Shades of the Intel Pentium floating point bug (not only do I remember that bug, I even recall receiving a replacement Pentium). A Microsoft developer has confirmed on the Excel blog yesterday that there's an issue with certain calculations in Excel 2007 that result in the number being displayed improperly (as opposed to being calculated improperly; you could multiply the result by a number to see the value itself was correct). Prior versions of Excel are not affected.

The key here is that the issue is actually not in the calculation itself (the result of the calculation stored in Excel’s memory is correct), but only in the result that is shown in the sheet.

So what, specifically, are the values that cause this display problem? Of the 9.214*10^18 different floating point numbers that Excel 2007 can store, there are 6 floating point numbers (using binary representation) between 65534.99999999995 and 65535, and 6 between 65535.99999999995 and 65536 that cause this problem. You can’t actually enter these numbers into Excel directly (since Excel will round to 15 digits on entry), but any calculation returning one of those results will display this issue if the results of the calculation are displayed in a cell.

As I previously wrote the recomputation engine on a spreadsheet (it was a major player at the time, but I'm not going to tell you the name as it'd date me ) I can tell you that recomputation engines as well as the display engines of spreadsheets are no simple task. You are not just trying to get the answer correct, but quickly (quicker than competitors, if possible) and with minimal recomputations of individual cells.

That said, I find it amusing that they say they "take calculation in Excel very seriously." Well, duh, it's basically a calculator.