Update on 04-May-2017: This is the original README from circa 1999 -- I am slowly trying to update this package, but for now enjoy this historical artifact!
FractalMate (nee Fractlet) is a simple application/applet that draws figures by recursively replacing the individual line segments of the original wireframe pattern by copies of the pattern. The resulting figure is called fractal, as its properties do not change when zooming closer in, ie. no matter for close you look, the same patterns and figures can be found at every level.
FractalMate is written in pure Java, and in theory should be run on any Java-platform. It has been developed on a Red Hat based x86 Linux PC using Sun's JDK. So far, it has been succesfully tested on following Java platforms:
- x86 (Red Hat) Linux, JDK's (1.1.7) Java VM
-
- "" -, Kaffe-1.0b4
- Netscape communicator 4.5/4.6/4.08 on Linux x86
- Netscape communicator 4.5/4.6/4.08 on Windows98
- IE 5 on Windows 98; update packages (including new Java VM) installed. Runs well; also, IE 4 and un-updated IE 5 were tested at some point; there were some small problems, but I have not been able to test whether there are still problems (after all, the program code has changed significantly since then.
I would be grateful, if someone could test out other platforms, so I could indicate which systems can run FractalMate without problems, and also fix (or by-pass) problems that only affect certain platforms.
The important thing to note is that you definitely need java 1.1 (or higher; 1.2 should be ok AFAIK); 1.0.2 will not do. Older versions of Netscape (pre-4 and some early 4.x versions) and IE (at least before IE 4?) will not do.
There have been following problems on the platforms mentioned earlier to run FractalMate. These include:
- Kaffe's AWT functions bit differently than JDK's AWT, and thus there are some cosmetic layout/drawing problems with UI. Functionally FractalMate appears to work, though, and it is possible that these problems may be fixed soon (by new versions of Kaffe and/or platform-dependant fixes to FractalMate).
- Netscape Communicator (4.5, 4.6) on Windows9x has a JIT (Just-In Time compiler), that used to have problems with some code in FractalMate; there were some internal errors... Fortunately I was able to by-pass them. All in all, Communicator's JVM is very buggy, and it's unfortunate Netscape/Symantec were not able to produce a better VM. A shame really.
FractalMate is copyright by me (Tatu Saloranta, tatu.saloranta@iki.fi), but the source code is itself is open source, in that its freely available, and you are free to use it for non-commercial uses without a specific permission. You can modify the code, re-distribute the modified code and all (for non-commercial purposes). If you use it, or distribute it (as is or modified), the only thing I ask is that the original source of code is mentioned (a http-reference or a link to the author's home page should do nicely).
For commercial use, a written permission is needed. Not that I think there's all that much commercial potential for selling fractal image things... 8-) Nevertheless, I am not a free software bigot (not against commercial software), and so getting the permission should be straightforward.
The program code and the application itself is provided 'as is'. If you manage to hurt yourself by using the application, you must be a real genius. :-p
I wrote FractalMate for fun. I hope you use it for fun, or education or something. Some fractals are oddly beautiful, and probably could be used for producing some nice graphics. I would appreciate it if:
- You create some nice 'new' fractals, and send them to me. I will probably dedicate a web-page to present such fractals, along with the proper credit for the authors
- You use FractalMate for something interesting (in addition to just using it for creating new fractals); I'd like to hear about this too. :-) Perhaps FractalMate could be used for explaining some mathematical things; fractal generation through complex number arithmetics (Mandelbrot-set etc) is not as intuitive as FractalMate's method, IMO.
- You find bugs from the application; perhaps even know how to fix them. Please send me the bug-reports and/or fixes!
- You think of some improvements to FractalMate; I might be able to implement them... Email suggestions!
- You are an optimization guru, and know how to speed up the fractal calculation or drawing... I have tried to optimize it a bit, and think it's relatively fast even as is... But I know there's no such thing as perfectly optimized code . :-) If so, send me a patch, or explain how I could improve the efficiency (things done so far: object allocation uses an object pool; made fractal generation 25-30% faster, tried replacing floats with doubles, made calculation ~5% slower, probably due to higher memory overhead; removed recursion from the drawing, no significant change in drawing time)
I have some ideas that I might be able to implement in the future. Such ideas include:
- Use of curves (bezier, b-splines) for fractal patterns. It is possible that I will try to implement this for version 2.0, if I ever have the time to do that...
- 3-d'ish manipulations for the image, ie. surface mapping the resulting image to a non-rectangular 2-d area.
- Perhaps even do a 3-d version of FractalMate. User interface for such a thing would need more thinking, and there are also technical problems for fractal generation... There are many ways to 'add one more dimension'.
- Use of more than 3 separate fractals; each fractal segment could define which fractal to use for replacements.
- Scenes that would consist of actual fractal images... Think of a forest that consists of trees, each being a single fractal figure.