Path: utzoo!attcan!uunet!aplcen!samsung!usc!rutgers!cmcl2!lanl!opus!afoiani
From: afoiani@nmsu.EDU (Anthony Foiani)
Newsgroups: comp.graphics
Subject: More Mandelbrot Stuff
Message-ID: <AFOIANI.89Nov25202207@dante.nmsu.EDU>
Date: 26 Nov 89 03:22:07 GMT
Sender: news@nmsu.edu
Distribution: comp
Organization: New Mexico State University, Las Cruces, NM
Lines: 33


I saw the program FRACTINT mentioned a few articles back.  This is an
excellent program for the IBM PC-types.  The Mandelbrot implementation
it uses is the familiar iterate-until-magnitude-greater-than-2
algorithm, but it is implemented in 16- and 32-bit fixed-point(?)
and/or integer math.  thus, when running it on a 286 or 386, it can do
a full 360x480x1024iter Mandelbrot in about 30 seconds.  [well, that's
on a 386SX @16.7MHz]

but that is still pretty fast.  they did the same thing to their julia
set [even faster than the MB set!], and to their newton-solution
technique.  they still have a floating-point implementation of julia
and MB sets built into the program, for the sole purpose of:

     "If you feel the need to reminisce about the days when 
      men were men, computers were computers, and fractals 
      took forever to generate."

sigh.

this program also includes lambda sets, lambda-sine, newton,
newton-basin-of-attraction, and 'plasma' clouds.  in addition, it
supports a large amount of video modes, cycles colors, can save any
picture do disk, and can map any saved picture onto a sphere or do
other nifty 3d projections.

the C source code is avalable, but i think it is currently being
distributed over Compu$erve... i haven't looked that hard tho.

happy hacking,
--
tony foiani  (afoiani@nmsu.edu)         "And remember...don't lose your
a.k.a. Tkil  (mcsajf@nmsuvm1.bitnet)     head..." -Ramirez, HIGHLANDER