[GRASS-dev] Naming conventions and axiomatics [RFC]

[tlaronde at polynum.com]

On Sat, Aug 06, 2011 at 11:26:03AM +0200, tlaronde at polynum.com wrote:
>[...]
> 2. Extending to 3D.
> 
> >From what has been explained above, a question arise for extending the
> process to 3D.
> 
> An arc can be one dimension or two dimensions. But an "area" (face)
> could be two dimensions (not defining a solid) or three dimensions (one
> orientation more: "above" and "below"). So an area is not mandatorily an
> edge of a solid.
> 

To be more accurate and unfold more, an arc can be 0 dimension (a point;
nothing theoritically prevent a "point" from being defined by a series 
of distinct vertices; these would be only static, i.e. not travellable
through; orientation is not defined; in GRASS, an arc of dimension 0 has
a topeon of two identical vertices. But this is a [natural] 
implementation choice. Less "natural" could be imagined.), 
dimension 1 (travellable in sequence, in one direction or the other),
dimension 2 (a supplementary orientation giving a left and right), etc.

So a solid would be built from dimension 3 arcs, meaning that some faces
could be subzones of solid faces without being faces by themselves.

The implementation problem is that in GRASS the sign is used for saying,
for an arc, the direction from N1 to N2, minus being from N2 to
N1. For faces, a positive id is a face, a negative an isle. So
there would be a need to discriminate with another feature (another
dimension) the orientation of a face to not mix the orientation of
a perimeter (defining area and, perhaps, isle), and the orientation
of a solid face.

-- 
        Thierry Laronde <tlaronde +AT+ polynum +dot+ com>
                      http://www.kergis.com/
Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C