"Jan Dvorak" <jan### [at] centrumcz> schreef in bericht 
news:47cb13c8$1@news.povray.org...
> Warp napsal(a):
>> Jaap Frank <jjf### [at] casemanl> wrote:
>>> Because triangles are not solid objects but surfaces only, you can't 
>>> carve something out of it.
>>
>>   No primitive in POV-Ray is solid. Everything is just a surface.
>>
>>   What "solid" means in terms of CSG is that some points can be defined
>> to be "outside" the surface and others "inside" it. Thus "closed surface"
>> would be a much more accurate term than "solid".
>>
>>   The problem with eg. the triangle primitive is that it's not a closed
>> surface, but on open one. Thus it can't have a defined interior (ie. 
>> there
>> aren't points which can be said to be "inside" a triangle).
>>
>>   OTOH, a closed triangle mesh can be defined as being "solid" because
>> POV-Ray can determine whether a point is inside this closed mesh or not.
>> This can be done by using the inside_vector feature of meshes.
>>
> well, umm, thanks. And remember the sun doesn't move. But that doesn't 
> stop us from saying it does. I often say the sun has moved behind the even 
> though there is no such thing as a horizon and the doesn't move at all 
> (well, it does). It is because the result is the same (up to some feeble 
> effects).
>
> By saying "a sphere is solid" i imply "with respect to the context" and I 
> mean "a sphere behaves just like it was solid" i.e. "you can do CSG with 
> spheres". (actually, the sphere *doesn't* behave, it has no brain).
> What I'm trying to say, people often describe what do objects do 
> (interface) rather than how do they do it (implementation).
>
> Your comment would be appropriate as a response to someone claiming the 
> raytracer considers the whole object interior when raytracing it. IMO your 
> comment is inappropriate here I don't mean to be rude, I simply find it 
> strange that you defend something that's not being denied. If not 
> inappropriate, your comment is not very hepful (neither is mine, I admit)
>
> -- 
> You know you've been raytracing too long when...
> you ever saw a beautiful scenerey and regretted not to take your 6" 
> reflective ball and a digital camera, thinking "this would have been a 
> perfect light probe"
> -Johnny D

What Warp means, I think, is that you can make a 'solid' octahedron
by placing eight triangles between the points
<1,0,0>,<-1,0,0>,<0,1,0>,<0,-1,0>,<0,0,1>,<0,0,-1>.
It's a closed object, so you now can treat it as a 'solid', given an inside
vector. You can use this object in CSG. I hadn't thought about that.

Nevertheless Patrick wants rounded edges, so there is some more work to do.
Maybe he can start with this object and figure out how deep he has to cut
off the rims and points to put his spheres and cylinders for the rounded
edges. That's not easy, for this has to be so deep that the newly formed
stripes have precisely the right width.
At the moment I don't see how to calculate this.

Patrick,
I think you forgot that prisms are made along the Y-axis and you
define the points in the X-Z-plane. Later you have to rotate the prism.
If you use your selfmade rounded triangle eight times, then all edges are
double and all angelar points are quadruple so you get coincident surfaces
and weird errors in your surfaces.
I think the best way is to make eight different rounded triangles the
way you did and omit the parts that are not needed: make a drawing of
the beast and put it then together.

Maybe you can use conic_sweep to make pyramids with a summit that
will meet all other pyramids in the center of the octahedron, but I'm not
sure about this, for I have never used conic_sweep before and the
manual is not clear about this.
If you make a 3D-drawing with the points I gave, you will see what I
mean (I hope :) ).

Success,

Jaap

Post a reply to this message