[OpenSCAD] F5 works, but F6 gives: ERROR: CGAL error in CGAL_Nef_polyhedron3():

Parkinbot rudolf at parkinbot.com
Thu Jan 18 18:39:13 EST 2018

Carsten I reread your post and you are right. An additional parameter for
polyhedron properly defaulted for novices will not close the door to hell.
Ok, it will break some code, but the fix would be straight forward.
My point is that added complexity like introducing a new, somehow "safer"
polyhedron primitive (say it will solve the "pink" challenge (CCW/CW
orientation) and check for manifoldness (without the Boolean operation
workaround) will not seal all traps a novice can and will encounter.
Wouldn't it just throw yet another, hopefully more reliable CGAL error (no
cache flush required any more) or even THE error, polyhedron users are all
waiting for? The only thing, I really opted for was: Make a reliable warning
and not an error out of it, to not prevent people from outputing
non-manifold STLs if they want.  

An error will indicate, but not solve the self-intersection and singularity
challenges and trap the people there. While the documentation is quite
explicit on the pink challenge, it happens from time to time that people
sign in, who got trapped there. Is this a weakness of the language? Yes it
can be fixed. OpenSCAD could check it and automatically switch to F12 mode.
But then a novice might not know, why the output (partly) turned pink. So
let's also output a warning or even an error? But most errors and warnings
are hard to read or neglected ...  

Sent from: http://forum.openscad.org/

More information about the Discuss mailing list