I have been familiarising myself with Fractint's L-system types for about 3 years now, and while I still have a long way to go, I have some ideas about how to further develop this part of the programme. It isn't, perhaps, used nearly as much as some other parts, but could still stand some modifications, which might attract a new set of users. The first improvement would be to allow for colour *vector* output, e.g. dxf or ai, instead of only bitmap output. L-systems are linear-based, not pixel-based. (I have a mutant version of Fractint sent to me by one of Fractint's developers which allows for dxf output, but without colours. This has enabled me to render L-systems as large as 1 GB, which fortunately compress to about 4% of their size when imported into CorelDraw and saved there!) A probability function would allow for more realistic-looking trees and other living forms to be emulated. For example, the option of setting a probability that any given part of the L-system string would occur, be it an angle change, size change or whatever. This seems to be a feature of some other L-system generators I have read about in books on fractals. I suppose that there would have to also be some way of ensuring that every somewhat "randomly" generated L-system could be produced more than once, e.g. by recording the actual variables used in each case of allowing the probability function. if...then, etc. could also be incorporated. If the line(s) resulting from iteration x reach or surpass an angle of y in relation to the horizontal or starting angle, do z: introduce an alternate text string command. If the total number of line segments resulting from iteration x reaches number y, do z; etc. A final modification, which is rather specialist so far but again which could enlarge the group of L-system uses, would be to allow *animations*. Currently, by using d instead of f, and setting angles instead of using Angle x, angles can be varied to the limits of Fractint's precision, allowing one L-system to mutate into another. The start and end parameters could be set, along with the intervening number of pieces to produce, and then the programme allowed to generate the whole set of pieces, which collected together would form a smooth animation from starting form to finishing. Actually writing a separate L-system for each piece would not then be necessary. I have many pairs of L-systems which would benefit from this proceedure. (IFS types and, I suppose, most fractal types could also be treated similarly. There are already a number of fractal animators out there; but can any of them handle L-systems?) Tony Hanmer _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp.