[OpenSCAD] Discuss manifoldness, co-incident faces edges etc

nop head nop.head at gmail.com
Thu Nov 14 14:06:43 EST 2019


A figure of 8 is only a self intersecting mesh if you create it by sweeping
a 2D shape. If you create it by unioning two rings the mesh will not
contain any self intersections. To avoid the problem with sweep you can
sweep sections that don't self intersect and union them. The problem of
repairing a completely general a self intersection mesh is the exactly the
same complexity as union because that is what union does. It computes the
boundary of two intersecting meshes, removing the internal faces.

The fact that 3MF says the positive fill rule should be used creates a sort
of Boolean XOR and is exactly the problem I had when printing somebody
else's design. Skeinforge used the positive fill rule but Cura at the time
printed the union.

>"there is no explicit topology but it is preserved."contradicts itself
because you cannot preserve something that does not exist.

No it doesn't because any manifold has some topology. If the manifold can
be stored in an STL file and reconstructed properly it still has the same
topology. That will be true as long as there are no shared edges. So
although the topology is not explicitly stored it can be reconstructed from
a polygon soup. I.e. you can pair up the edges and make a data structure
that explicitly connects them.



On Thu, 14 Nov 2019 at 18:43, Jordan Brown <openscad at jordan.maileater.net>
wrote:

> On 11/14/2019 7:32 AM, arnholm at arnholm.org wrote:
>
> I think self-intersections are almost always a sign of something else gone
> wrong, it is typically a body that overlaps itself and and not a result of
> a boolean operation. A self-overlapping volume is physically impossible.
> There are lots of ways to use bodies that are not 2-manifold, but a self
> overlapping volume seems like an obvious mistake.
>
>
> Is it?
>
> Consider the 2D case.  I'm drawing with a paintbrush, and so I'm drawing
> areas, not lines.  I draw a figure-8.  It's self-intersecting, but it's
> clearly not impossible.  The two parts of the stroke are unioned.
>
> If we view the 3D equivalent as painting in 3D with a 3D brush, the result
> doesn't seem impossible at all.
>
> If we were drawing with voxels, I suspect that the question wouldn't even
> arise.
> _______________________________________________
> OpenSCAD mailing list
> Discuss at lists.openscad.org
> http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openscad.org/pipermail/discuss_lists.openscad.org/attachments/20191114/0d9bbaef/attachment.html>


More information about the Discuss mailing list