gawk: Inexactness of computations

1 
1 15.4.1 Floating-Point Arithmetic Is Not Exact
1 ---------------------------------------------
1 
1 Binary floating-point representations and arithmetic are inexact.
1 Simple values like 0.1 cannot be precisely represented using binary
1 floating-point numbers, and the limited precision of floating-point
1 numbers means that slight changes in the order of operations or the
1 precision of intermediate storage can change the result.  To make
1 matters worse, with arbitrary-precision floating-point arithmetic, you
1 can set the precision before starting a computation, but then you cannot
1 be sure of the number of significant decimal places in the final result.
1 

Menu

1 
· Inexact representation      Numbers are not exactly represented.
· Comparing FP Values         How to compare floating point values.
· Errors accumulate           Errors get bigger as they go.
1