https://issues.apache.org/ooo/show_bug.cgi?id=121421
--- Comment #12 from Pedro Giffuni <[email protected]> --- (In reply to comment #11) > Created attachment 79972 [details] > Alternative for Generating Sets of Results > > This is adjusted to generate long runs of the random values. Enough digits > are printed of the 0 < r < 1 random doubles so that where the values tail > off at the extremes of precision (and the output conversion software) can be > observed. > > An optional command-line parameter can specify the length of the sequence > you want. The default for the parameter is 1000. Interesting results > appear with runs of 1000000 or more. By then, each of the three short-width > generation functions will have recycled a few times, but not in > synchronization. Ignoring MAX_RAND is a very bad idea. I also found my implementation is incomplete: to avoid precision issues the complete implementation does some adjustments that involve substracting. That may be cause for microsoft's bug. The algorithm is rather old and thought for 16 bit architectures. There has been a revision in 2007 (which obviously didn't make it into Office 2003). I will update this soon. -- You are receiving this mail because: You are on the CC list for the bug. You are the assignee for the bug.
