Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!purdue!ames!amdahl!kevin
From: kevin@uts.amdahl.com (Kevin Clague)
Newsgroups: comp.sys.amiga
Subject: Re: MandFXP?
Message-ID: <20YMz88.Fs1010aEv9o@amdahl.uts.amdahl.com>
Date: 2 May 89 16:14:54 GMT
References: <10845@polyslo.CalPoly.EDU> <8818@polya.Stanford.EDU> <11608@s.ms.uky.edu>
Reply-To: kevin@amdahl.uts.amdahl.com (Kevin Clague)
Organization: Amdahl Corporation,  Sunnyvale, CA 94086
Lines: 74

In article <11608@s.ms.uky.edu> jgary@ms.uky.edu (James E. Gary) writes:
>In article <8818@polya.Stanford.EDU> rokicki@polya.Stanford.EDU (Tomas G. Rokicki) writes:
>>Nothing is as nice as MandelVroom.  Nothing.  An earlier version is on
><stuff deleted>

Thanks Tom.  (I payed him to say that ;-)

>
>Does MandelVroom use the recursive perimeter algorithm that RMand (a     
>relatively recent PD prog) uses? RMand is quite quick, but doesn't have
>all the bells and whistles that I would like to see. Briefly, rmand
>calculate the colors for the pixels forming the perimeter (the entire
>perimeter) of the screen and if the same, does a flood fill. This tends
>to flood the 'black' areas (those actually in the set) quickly, where
>most other algorithms spend most of their time. Anyone working on a 
>full featured Mandelbrot program should consider this algorithm. I haven't
>implemented it, but understand it pretty well and will be happy to 
>correspond via e-mail with interested parties.

MandelVroom does not use the recursive perimiter algorithm for a number
of reasons:
  1.  It is not as precise as the brute force method...  If the
      computer were indeed calculating a continuous perimiter (i.e. an
      infinite number of points along the perimiter) then it would
      be a good idea.  I indeed implemented it once long ago (non-
      recursively) and chose not to keep it.

  2.  This algorithm works nice for pictures with large amounts of
      the Mandelbrot set in it, but slows down the the non-Mandelbrot
      stuff.  I find large sections of black boring, so the algorithm
      would only get in the way for most of the really pretty pictures.

  3.  MandelVroom is a "full figured" Mandelbort/Julia picture generator
      already.  Some of it's features:
     -Uses configurable custom screen (HAM is only mode not supported)
     -Project oriented.  Multiple pictures on the screen at once.
     -Multitasking.  Can have as many projects open and generating
      (simultaneously) as you want.
     -Save/Load projects (can save partially generated picture, reload
      it and resume)
     -Borderless projects.
     -Zoom in/out as well as pan around.
     -Powerful picture coloration tools that allow you to recolor
      pictures while they are generating.
     -Online help for every gadget and menu item.
     -Orbits... Tom asked for these...
     -Five Math modes:
      -68000 fixed point, 68020 fixed point, FFP, IEEE, 68881 assembly.

     With five math modes for Mandelbrot and Julia generators, that
     makes 10 generator routines.  If we add another dimension for
     perimeter chasing that makes 20.  Ten with perimeter chasing.
     Ten without.

I've been very busy lately preparing to move into my first house,
so MandelVroom release has been slow in coming.  The release date
has been a month away for three or four months now, I hope it
really does happen soon.

MandelVroom will be released with source, so you can add it if you
want.  Good luck with some of the code.  Most of it is REALLY icky.

I've also been very distracted by Fract3D (a new 3D fractal generator
I've been working on.)  It produces HAM pictures that are really nice.

Kevin
-- 
UUCP:  kevin@uts.amdahl.com
  or:  {sun,decwrl,hplabs,pyramid,seismo,oliveb}!amdahl!kevin
DDD:   408-737-5481
USPS:  Amdahl Corp.  M/S 249,  1250 E. Arques Av,  Sunnyvale, CA 94086

[  Any thoughts or opinions which may or may not have been expressed  ]
[  herein are my own.  They are not necessarily those of my employer. ]