https://bugs.documentfoundation.org/show_bug.cgi?id=166631
--- Comment #6 from Steve Magruder <[email protected]> --- (In reply to Alfio Littletree from comment #5) > Welcome Steve. I tested your attachment with LibreOffice version 25.2.3.2 > (X86_64). > Calc do not produced zeroes with your formulas. > However, reading your original formula, I realized that it can produce zero > as a normal mathematical result. Indeed, in the last IF(;;) you left zero as > the third argument: zero is default when omitted. This sixth case, based on > the comparison conditions you originally used, occurs when a floating point > CPU requires to round the binary representations of your prices. > If you want to avoid zero prices, you must always use the second formula, > the one you called workaround. It is mathematically different, it cannot > produce zero. > The only flaw with spreadsheets is that they use the same calculation tools > to do monetary calculations as they do for length or weight calculations. If > fixed-point cells will be added in future years, the cells where you will > put your currency symbol will work as expected in your original formula. I see your point and have already come to this realization as my price calculations (which provide the first set of prices for normalizing) won't always give me flat prices to begin with for normalizing. So I am using something close to my workaround for production purposes. That said, based on the price data I tested with in this situation, it should not have produced any zeroes, and indeed, it didn't after I followed the process for resetting the UserProfile. It was an actual glitch, although if it cannot be reproduced, it also cannot be called a bug. -- You are receiving this mail because: You are the assignee for the bug.
