I've attached a modified version of the DigitalMars D Mandelbrot set
test program that uses D's builtin complex type. It should have the
same performance but the code looks nicer so I thought I'd submit the
update.
Note that test seems pretty fragile since I slightly modified the way
the initial seed "c" was computed and it must have resulted in a
difference in a bit or two since the resulting image files tested as
different by diff. Since the mandelbrot set is all about exploiting
small changes in initial conditions any difference in roundoff
behavior could result in a different image.
thanks,
-Ben
_______________________________________________
Shootout-list mailing list
Shootout-list@...
http://lists.alioth.debian.org/mailman/listinfo/shootout-list