[exasperaited]

OpenSCAD isn't a CAD tool, IMO. It does nothing at all to aid your design. It barely even helps describe it in any abstract way.

It is a useful tool for programmatically describing either very simple or heavily geometric objects. For everything else it's the wrong tool.

[voxleone]

I'd put it this way: CAD = Computer-Aided Design, i.e., software that assists in the creation, modification, analysis, or optimization of a design.

OpenSCAD clearly qualifies because it creates precise 2D/3D models; supports parametric modification; enables reproducible designs; produces manufacturable geometry (STL, DXF, etc.). The fact that it is code-based does not disqualify it. By the same logic, CadQuery, SolidPython, or FeatureScript would also not be CAD - which is not a position held in engineering practice.

[exasperaited]

I am not disqualifying it because it’s code-CAD. (I think it’s a bad language design but that is a total aside)

CadQuery, Build123D, Featurescript etc. all have something OpenSCAD does not have: they all have facilities to operate on generated geometry such as edges, faces and vertices. So they have generalisable operations, and they can output pure geometry representations as STEP/IGES. Featurescript presumably also allows access to sketch constraints. (Build123D is beginning to add those I think.)

These are the things I think are necessary to “aid” design, because they are the things that begin to abstract away the mathematics. OpenSCAD never does. So it “aids” design little more than any pixel graphics library aids 3D software developers.

And it does not, in fact, create precise 3D models! It produces mesh representations of them. So it cannot, by definition, produce a precise representation of a cylinder, sphere, cone or curve.

The 3D models from OpenSCAD are not highly manufacturable at all, except with additive production techniques. STLs do not make good inputs for a CNC (nobody will tell you that is good engineering practice).

[WillAdams]

STLs for subtractive CNC may not be good engineering practice, but it is quite workable as MeshCAM has shown:

https://www.cnccookbook.com/meshcam-great-ease-of-use-in-a-3...

[exasperaited]

I am pleased MeshCAM exists, I am pretty sure I will end up using it!

But I think we agree that mesh formats don't make good engineering interchange outside additive manufacturing.

Personally I think they often make a bad format even in additive manufacturing. They are the right format for medical/dental, and yes, slicers make internal STL-ish mesh representations of STEP etc. before slicing, but they are much harder to modify precisely than STEP.

[WillAdams]

If you have to modify an STL, you're doing something wrong --- go back to the original/source file.

[exasperaited]

Well this is true, but it is STL files that people send around everywhere and publish in the 3D printing world, especially.

People will reasonably choose not to publish their original designs but can only publish STL because of OpenSCAD, Blender and because of the free tier of commercial CAD.

And having the OpenSCAD source doesn't help that much; you can modify it but you are still stuck with STL. OK so FreeCAD can reinterpret some — not all — OpenSCAD operations in a bRep context. But it's not a great way to work; it doesn't really help much with interchange.

I absolutely have had to modify STL files for people because that is the only interchange format the designer made available to them, or even could make available because they used the wrong tool for the job (like Fusion's free tier). At this point there are many STL-folk-song designs out there, CC-NC, with bad geometry, non-centred exports, etc.

Thankfully there are ways, in core FreeCAD and in a third party workbench (MeshRemodel), to more easily extract and reconstruct some key geometry from a mesh, such that parametric modifications can be made. (Whereas with OpenSCAD you're basically stuck with cut-out-and-roughly-refill).