Very Computer

Why is xfractint so darn slow? Heck, DOS Fractint under DOSEMU is faster!

--
Fredrik Sandstrom, Finland

Quote:>Why is xfractint so darn slow? Heck, DOS Fractint under DOSEMU is faster!

Yeah, it takes almost 8  seconds to render a mandlebrot here.

d-e-a-d-i-c-a-t-e-d---[ http://www.conservatory.com/~fishbowl ]-----l-i-g-h-t
Some people are born mediocre, some people achieve mediocrity, and some
people have mediocrity thrust upon them.  -- Joseph Heller, "Catch-22"

: >Why is xfractint so darn slow? Heck, DOS Fractint under DOSEMU is faster!

: Yeah, it takes almost 8  seconds to render a mandlebrot here.

What version are you guys using?  Version 3.0?  Or an older version?

byron
--
Mathematicians often resort to something called Hilbert space, which is
described as being n-dimensional.  Like modern sex, any number can play
                        -- Dr. Thor Wald, in "Beep/The Quincunx of Time"

No doubt the biggest factor in reducing the speed is the absence of the hand
coded assembler routines that make the DOS Fractint so fast.

Hmmm.  I suppose that since Linux runs on Intel machines, it would be possible
to make a special Linux version of xfractint that would make use of the
80x86 assembler.  MUCH of the asm code would have to be rewritten for
32 bit addressing as all of it is currently written for 16 bit real mode.
Still, it would be possible...

-wes loewer (a main contributing author for fractint 19)

: Very unloaded system:  AMD486DX100, 20MB, XFree863.1.2, S3 video,
: 1.3.28 kernel, Xfractint 2.03...  Takes about 8 seconds for all three
: passes to render a mandelbrot.

On my very similar system (AMD486/100, 16MB, XFree86 3.1.2, S3 video,
NCR SCSI-controller, 1.2.13 kernel & Xfractint 2.03) it takes only four
seconds to render the default mandelbrot. Of course, if I zoom to a place
which needs lots of iterations (=a lot of black), the time goes up.

--
Terje Bergstr?m