There are two different kind of numbers in computer industry: integer and floating point. Integer (or long integer) is without decimal fraction and is always precise, but has limited range (approx. +/- 2 billions). Floating point has nearly unlimited range, but may be imprecise. This means, the number 5.0 may be stored as floating point 5.000...000 or 4.9999...999 or 5.0000...001 (the decimal fraction is up to 15 digits). To be sure to compare two floating point values (or integer with floating point), best to use the ROUND() function. See further details and hints in the FlagShip on-line manual (fsman) section LNG.2.6.4
Paul