http://web.ukonline.co.uk/members/robin.b2/olig/wisharc.htm That looks like *a* wishlist archive, but my wishes from after 1999 aren't on it. Is there another, later one? Here are a few wishes of mine, all for L-systems: 1. Vector output as an option, e.g. dxf. L-systems ARE linear-based fractals, after all. Vector output would allow exactness in L-systems rendering - something sorely lacking once screen details come anywhere close to maximum. I do already have a "mutant" version of Fractint with this capability, kindly sent to me by Robin Bussell. It works only in mono, no colour output, but has given me much enjoyment. I would suggest 256-colour output at least. (This option could even be included for tesseral and outline output types of bitmap fractals! Both allowing filled closed shapes. Think of that.) 2. If vector output is enabled, then other possibilities beg for implementation. Joining of line segments' endpoints to make closed curves which can be filled in a programme like CorelDraw is one. Removal of endpoints of adjacent line segments which are on the same straight line is another - a way of hugely reducing file memory size. Reducing of "layers" of redundant line segments after each iteration is a third. This is unnecessary in bitmap output - drawing a line on top of itself many times has no effect on the memory size of the file. But in Vector output, it does, of course, add to file size, possibly multiplying size many times over. There are many L-systems (like Hanmer Tiles) which have great numbers of layered line segments; keeping them as small as possible in vector output would be of great benefit. Line segment elimination could be done after each iteration has been calculated, before the follownig iteration - this slows down calculation, of course, but computers ARE getting faster all the time... 3. The possibility of line segments being curved instead of only straight! One could select from a list of curve equations, and the list could be added to as well. This would really transform L-systems. 4. Removal of some limitations (I've mentioned this before; not the highest priority, but not too difficult to implement either I think, once Fractint migrates from DOS). Namely, *.l file limit of 2000 entries (Hanmer Tiles are infinite in number...), and L-system programming string limit of 256 characters (I think that's the current limit) (Hanmer Tile programming can lead to very long strings). 5. Some kind of randomisation factor, user-specified as a percentage or range. This could vary either line length, or angle, or both, and lead to more natural-looking L-systems for rendering of plants, etc. Perhaps if a certain specific randomisation is attractive to the user, it could be saved and re-used. 6. IF-THEN programming. There are all sorts of possibilities for this, most of which I probably am not informed enough to imagine. Again, they would greatly extend L-systems programming. 7. 3d is perhaps not more than a dream or a distant goal, if anything. Tony Hanmer