2011/6/14 Jayesh Salvi <jayeshsa...@gmail.com> > > > > I mean 'hazardous' because there might be a confusion between the CGAL > and > > pythonOCC (BRep) data models. > > > > Reading your github repos, I see that the Sphere class inherits from > > Polyhedron. But a sphere is *not* a polyhedron, whereas its tesselation > is. > > Polyhderons have flat faces and straight edges. > > > > According to me, 'simple' API means: easy to learn, enabling few lines of > > well readable code to achieve expectations. It must be semantically > explicit > > and unambiguous (which is not the case if you make the Sphere and > Cylinder > > classes inherit from PolyHedron). > > Oh I see what you are saying. This answers some questions I had. > > IIUC, you are saying that the name Polyhedron implies only topology, > which sphere and cylinder are topology+geometry. Is that correct? >
For the difference between geometry ans topology, I suggest the reading of Roman's blog: http://opencascade.blogspot.com/2009/02/topology-and-geometry-in-open-cascade.html http://opencascade.blogspot.com/2009/02/topology-and-geometry-in-open-cascade_09.html > I didn't think about that while naming the base class Polyhedron. I > think I should rename it to "Primitive". Does that sound ok? > These are 'BRep Primitives', and the OCC module that creates such primitives is named 'BRepPrimAPI'. > > Moreover could you tell me, is there a way for me to get only the > topology of sphere/cylinder? As you said, they are tesselations, so > OpenCASCADE does its best to generate as smooth a surface as possible > to match the real curvature. But this results in lot of polygons. What > option do I have to limit or control the number of polygons in the > tesselations? > With OCC/pythonOCC, you only handle topology/geometry, never tesselations unless you explicitely call BRepMesh::Mesh. The 'deflection' parameter of this function enables modifying the quality of the number of triangles generated. However, the meshing algorithm does not allow to generate very small meshes. > > Thanks for clarifying this. > > Thomas
_______________________________________________ Pythonocc-users mailing list Pythonocc-users@gna.org https://mail.gna.org/listinfo/pythonocc-users
