Subject: Re: [geometry] Compute Normal in 3D
From: Adam Wulkiewicz (adam.wulkiewicz_at_[hidden])
Date: 2014-01-31 07:33:56
Tomislav Maric wrote:
> Hi everyone,
> I've been doing some work on the 3D calculations. You haven't heard
> from me for a while, as I had to switch and work on another project.
> I did my best to follow the design guidelines that I could find in BG,
> and based on that I have a AABBbox, PolygonSoup, Halfspace,
> LineSegment, Tetrahedron, TetrahedralDecomposition and similar things
> + intersection (test) algorithms up and running in 3D.
> Right now, the code is tied to a C++ Computational Fluid Dynamics
> framework OpenFOAM - currently only by the Point type and vector
> operations (cross/dot product). As soon as I finish the method I'm
> working on, I will have time to detach the 3D geometry into a separate
> I would really like to work on it until it fits the standards of BG,
> but I would probably need input regarding the design from your side.
> If I don't get interrupted again by something else, I estimate I will
> be able to start porting all I have in a few weeks time. At that point
> you guys can see if what I managed to do is worth an effort from your
> side, or not.
That's great. In fact we have a raw version of guidelines for developers
(thanks to Mateusz), it mainly describes formatting the code but maybe
you'll find it interesting:
I'm curious, why did you add AABBbox and LineSegment? BG already has
n-dimensional box and segment. Is it just a naming thing?
Btw, intersects()/disjoints() works for n-dimensional pairs Point/Box,
Segment/Box and Linestring/Box (last two released in 1.55). In case
you'd like to check it out.
Geometry list run by mateusz at loskot.net