Apparently I missed the part in the STL spec where values need to be non-negative and non-zero.
Right now the code I’ve written to generate a model reminds me a great deal of the first program I ever wrote. I was 8, typing on a Coco 3, and I wrote a project that was basically a giant pile of prints that animated a rocket taking off, viewed from a portal (the screen) so you saw the tip, the capsule, the rocket body, and the exhaust scroll by.
In “rip.py”, I should give that a better name, I have an endless series of for loops generating different planes of triangles to build different box structures matched up to the mesh of mountains I read in another set of for loops.
I’ve not been concerned about code quality up to this point, just about all the work is done in one 400 line function that does contain a handful of local functions. It’s been fine, I’m learning what work in this space looks like and finding useful points to consolidate and improve work.
What I am starting to worry about is translation spaces. I’ve done a think where all points are written by a single function that translates points from the integer x,y index of the data source to a uniform space, meters, that are then scaled again for output. The generated data is unitless, but gets used as millimeters by my printer.
Now I’m doing work in real space (output mm) as well, and I need to mix measurements between the two spaces, integer index land, scaled by the data (meters per input pixel) and the output (millimeters in 3d space).
Certainly this has been handled countless times, but it’s time for me to learn how.
http://www.solidterrainmodeling.com/ :- Solid Terrain Modeling uses patented technology to produce geospatially accurate, physical 3D terrain models that you can touch and hold. Enjoy some of our work, our clients’ and partners’ work, and other interesting videos as we continue to explore our GIS universe.
Eric shared this with me.
Anybody you know?
Woops, sorry shapeways, I’ll try harder.