Background: I am doing some area statistics (how much of different properties is used for a specific type of use) based on ACAD-constructions. Both the from dxf converted shapes and the baselayer with all the properties weren't changed, but when I redid the calculations today in order to check for the source of some discrepancies, I stumbled upon several different numbers in my old and new statistics.
Therefore the question: How exact is the $area-function in QGis? Are errormargins by 10 to 100 square metres to be expected? (We're talking usual european field sizes, 50 hectars max)