AABENSON@MTUS5.BITNET (12/12/90)
Does anybody have a program that generates a 3200 color version of one of
these? That would be NEATO! (And probably take a WEEK to generate! Ha ha!)
- Andrew.
Please email responses to Internet: aabenson@balance.cs.mtu.edu
Or if you're more of a relaxed-pace person, you could email it to
Bitnet: AABENSON@MTUS5.BITNETTHROOP@GRIN1.BITNET ("Throop,Henry B") (12/14/90)
aabenson writes: > Does anybody have a program that generates a 3200 color version of one of > these? That would ne NEATO! (And probably take a WEEK to generate! Ha > ha!) A Mandelbrot set plot is not the sort of thing that would benefit much from 3200 color, because one is still limited to 16 colors/scanline. Because the Mandelbrot set is a contuinuous set and the colors follow each other in order, many of the colors must be the same from one line to the next. If they are not, the same level sets would be different colors on different lines, which would make it meaningless. If the colors are the same from line to line, then obviously you won't be using 3200 colors. It might benefit a bit, but not significantly. If you did have some way to overcome the 16 colors/line restriction, generating with 3200 colors, aside from the slight time to get the palettes right, shouldn't take any longer than with 16 colors, or 2 colors, because one just takes the value of that level set mod number of colors to calculate the color number. I have a (16 color) Mandelbrot/Julia program I wrote awhile back that works pretty well, but I gave it up because TML Pascal II is impossible to work with. It uses integer multiplication (faster than floating point, even from a high-level language line Pascal), lets you zoom in with the mouse by just selecting an area while it's plotting (and then zoom back out wher you can continue that plot), and some other nifty features. I also had it set up so call up the local Unix system, send the numbers to calculate to it, and then plot the results on the screen. Even with the 2400 baud bottleneck, it's a whole lot faster than having the gs calculate each point. I was working on making it do this in a pseudo-parallel processing style, using a network of gs's hooked up over AppleTalk, but about then TML decided to stop working completely, so I gave it up. -- Henry Throop THROOP@GRIN1.BITNET throoph@jacobs.cs.orst.edu