hmm, that's an interesting question... on the one hand, i wouldn't really expect uf to be that much faster than java since most java implementations have just-in-time compilation, which has been tuned for a number of years and i suspect uf's scripting is interpreted*; on the other hand i'll bet uf has built in functions for dealing with complex numbers and quaternions (which will be a real boon for you given what i said in your cubic/quintic mandelbrot set thread), which will use fast c/c++ routines internally. depending on the mix of native vs scripted operations the results can go very much either way...

unfortunately there are too many complicating factors for a direct comparison, for example support code around the basic execution, like java's display classes or uf's gui rendering code.


* if it's actually compiled to machine code then it boils down to the efficiency of said compiler versus java's ridiculously conservative/safe execution (eg array bounds checking etc).

Julesruis that's quite obviously cheating
all of the other ways look better than that, it's just a 2D shape spun around one axis, like a lathe
look on Ultrafractal, there are several formuli for doing it in diffrent ways.