Hi Richard,
Sorry for the delay. It was a wonderful trip - the Caucasus mountains are glorious at any time of the year, and I have a lot of friends there in Svaneti.
6. If..Then programming for L-systems: I've not read or seen any examples of this, I just know that it's possible. My ideas concern the variables built into the language, such as Angle, @ (controlling length ratios of line segments), ! (reverse direction), brackets for stacking, etc.
An example of If... Then would be the equivalent of "If f is at an angle of n degrees [from baseline, which is horizontal], do A", A being a new L-system string of the user's choice specified in the current L-system. One could have this new string kicking in if:
f reaches size n as a ratio of the initial f;
the total number of fs in the iteration reaches n;
f reaches angle n in relation to baseline (as above);
there are n copies of f "stacked" in layers directly on top of each other in the iteration;
any of the above for g instead of f;
the length or height of the entire iteration reaches n as a ratio of the initial f...
These are a few possibilties I've thought of; there may well be many others.
7. I'm not at all familiar with P.P.'s 3d extensions; I have written to him about my infinite tile finds at least twice and received no answer. I revere his work, though, and have a copy of his first L-systems book. (It's not yet come out of storage along with most of my possessions which I tucked away when I left Georgia for Canada in 2005.)
3d L-systems - should these still be line-based, or somehow planar, or tube-based, or what? I don't know and haven't seen any of whatever 3d LS programmes might be out there. Sorry, I can't provide too much thought about this one.
Tony Hanmer
On 5/24/07, Richard <legalize@xmission.com> wrote:
Tony,
Now that you are back from your trip (I hope it was fun and relaxing),
can you elaborate on these points you mentioned earlier?
> In article <
546ce4c30705012157g171b82f1q2992120db09e8a39@mail.gmail.com>,
> "Tony Hanmer" <a.hanmer@gmail.com> writes:
>
> > 4. Removal of some limitations (I've mentioned this before; not the highes
t
> > 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).
>
> This can be done for the next beta.
>
> > 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.
>
> Can you provide some sort of specification of how this would affect
> the L-system grammar? I have seen probabilistic L-systems and other
> extensions of continuous parameters into L-systems done by
> Przemyslaw Prusinkiewicz in his SIGGRAPH papers and books, but I don't
> recall him introducing a conditional extension.
>
> > 7. 3d is perhaps not more than a dream or a distant goal, if anything.
>
> Do you mean 3D a la P.P.'s extensions? Can you elaborate on this a
> little more?