Let's gather interesting test cases for a unit test of the ComplexNumberPackage here. This may provide a more meaningful way to compare various implementation technologies. One battery of tests we need is all the permutations of primitive types, including positive, negative, and zero values. I also suggest we perform some interesting Julia set or EE calculations (maybe a whetstone?) from http://moon-soft.com:8000/program/C/C2/GhI-39/GADTERM/TERMS.TXT: : ''>Whetstone. A standardized test of overall computer system performance. Said to reflect the actual performance doing real world tasks. Written by the National Physical Laboratory (England's version of NBS) based on an analysis of 1000 ALGOL-60 programs, the 'Whetstone Instruction' is really an average of 1000 language constructs arrange in a loop, one interaction of which is considered to be "1K Whetstones", the accepted basic unit measurement. For example, a VAX-11/780 does about 1,133K B1D Whetstones/second . (By comparison, the VAX is also rated at about .25 MFLOPS). See Also: Benchmarks'' I think you mean Wheatstone bridge: http://www.wikipedia.org/wiki/Wheatstone_bridge. ''No, I meant a "Whetstone", as in the performance measurement: http://www.wikipedia.org/wiki/Whetstone. Being trained as a EE many years ago, I confused the benchmark ("Whetstone") with the circuit ("Wheatstone Bridge"). I edited my original line accordingly.'' ---- See also ComplexNumbers ComplexNumberTest ---- CategoryMath