I'm interested at some point in taking on writing a random tree macro.
Probably mesh-based. I would just like to know, from those who have written
them, where do you begin? What's the general theory for going about it? I
assume there is much recursion involved. If you begin with one triangle in
the mesh, how does the macro figure out where to place subsequent vertices
as it builds the tree? Basically, I want to know just how far over my head
just a task would be. ;-)

Post a reply to this message

I use a good (and free) tree generator that exports to POV files :

ARBARO (http://arbaro.sourceforge.net/)

If you don't want to use it but you rather want to write your own macro, the 
theory is well explained in the PDF file :
http://www.cs.duke.edu/education/courses/fall02/cps124/resources/p119-weber.pdf

If it can help you, as Arbaro is open source you can download the source 
code and read it.

Good luck to your project


web.42ea5964ab8cade7e32e54c40@news.povray.org...
> I'm interested at some point in taking on writing a random tree macro.
> Probably mesh-based. I would just like to know, from those who have 
> written
> them, where do you begin? What's the general theory for going about it? I
> assume there is much recursion involved. If you begin with one triangle in
> the mesh, how does the macro figure out where to place subsequent vertices
> as it builds the tree? Basically, I want to know just how far over my head
> just a task would be. ;-)
>
>

Post a reply to this message

Hm, I guess the question I should have asked is more general... Supposing
that I want to create a mesh object which has random factors incorporated
into it's shape (not necessarily that of a tree), and that the mesh is
built at run-time through a loop of some sort, how do I make sure that it
remembers where it left off with previous triangles? How do I make the
vertices coincide if I don't know what they will be?

"bancquart.sebastien" <ban### [at] wanadoofr> wrote:
> I use a good (and free) tree generator that exports to POV files :
>
> ARBARO (http://arbaro.sourceforge.net/)
>
> If you don't want to use it but you rather want to write your own macro, the
> theory is well explained in the PDF file :
> http://www.cs.duke.edu/education/courses/fall02/cps124/resources/p119-weber.pdf
>
> If it can help you, as Arbaro is open source you can download the source
> code and read it.
>
> Good luck to your project
>

> web.42ea5964ab8cade7e32e54c40@news.povray.org...
> > I'm interested at some point in taking on writing a random tree macro.
> > Probably mesh-based. I would just like to know, from those who have
> > written
> > them, where do you begin? What's the general theory for going about it? I
> > assume there is much recursion involved. If you begin with one triangle in
> > the mesh, how does the macro figure out where to place subsequent vertices
> > as it builds the tree? Basically, I want to know just how far over my head
> > just a task would be. ;-)
> >
> >

Post a reply to this message

I cannot answer you because
1) I don't exactly see what you want to do (even if I have a little idea)
2) my POV-level is quite poor
but, just a hint. Maybe the "trace" function could help you

Post a reply to this message

Justin Smith wrote:
> Hm, I guess the question I should have asked is more general... Supposing
> that I want to create a mesh object which has random factors incorporated
> into it's shape (not necessarily that of a tree), and that the mesh is
> built at run-time through a loop of some sort, how do I make sure that it
> remembers where it left off with previous triangles? How do I make the
> vertices coincide if I don't know what they will be?
> 
> "bancquart.sebastien" <ban### [at] wanadoofr> wrote:
> 
>>I use a good (and free) tree generator that exports to POV files :
>>
>>ARBARO (http://arbaro.sourceforge.net/)
>>
>>If you don't want to use it but you rather want to write your own macro, the
>>theory is well explained in the PDF file :
>>http://www.cs.duke.edu/education/courses/fall02/cps124/resources/p119-weber.pdf
>>
>>If it can help you, as Arbaro is open source you can download the source
>>code and read it.
>>
>>Good luck to your project
>>

>>web.42ea5964ab8cade7e32e54c40@news.povray.org...
>>
>>>I'm interested at some point in taking on writing a random tree macro.
>>>Probably mesh-based. I would just like to know, from those who have
>>>written
>>>them, where do you begin? What's the general theory for going about it? I
>>>assume there is much recursion involved. If you begin with one triangle in
>>>the mesh, how does the macro figure out where to place subsequent vertices
>>>as it builds the tree? Basically, I want to know just how far over my head
>>>just a task would be. ;-)
>>>
>>>
> 
> 
> 
> 
> 
I think your "random factors" will have to be used to define some 
general structure which is then subdivided in an organized way.   For 
instance suppose you are doing a tree.  Suppose you have a way to 
generate a realistic set of node points that mark the advance and 
bifurcation of the branches.  These points can then be used to define 
vectors which "connect" them.  Points can then be located around and 
along these vectors in a regular grid as if the vector formed the axis 
of the branch.  It would then be easy to loop through these points and 
define a mesh of triangles before moving to the next "axis"

The variety of available splines, they can be curved as well as linear, 
might provide an alternative stepping stone between random generated 
nodes and a grid of vertices.

You might generate your random nodes in steps or all at once, but no 
doubt you will have to store them either temporarily or in total.  You 
will almost certainly use a structure such as an array or indexed file 
to do this.

Post a reply to this message