Constraint-based geometry (CAD) sketcher for Blender
github.com
github.com
real cad tools like fusion360, autocad etc are parametric. they represent a shape in the form of constraints. The underlying representation of the shape is different in a way that facilitates modeling something that you are going to create a real world counterpart from.
Yes there are open source cad programs but nothing I've seen comes close to rivaling professional tools the way blender does.
Serious modern parametric CAD (Autodesk Inventor, Solidworks, etc.) does constructive solid geometry well, but is not well suited to animation. Making a face via CSG is terrible. Although it's gotten better; you can now do sculpted shapes in CAD, because people design cars in them. That has both sides of the hard problem - the smooth shapes must look good, and the bolt holes for the stamping dies must line up precisely.
I've been keeping up with the latest versions, but even with all the bug fixes that go in I am more likely than not to have it segfault or lose its GL context or something else that causes it to crash. This isn't even hyperbole, I literally get 2 or 3 crashes at minimum per session, and when it happens it's rarely reproducible. I simply save after every couple of changes I make because I've just come to expect it to happen.
When it does work, I frequently run into problems like the infamous topological-naming-issue that, while I understand it is a difficult software problem to fix, I do not run into in the other CAD programs I've used. The only way to avoid these kinds of issues is to understand the underlying cause and then adjust your workflow so that you don't run into them.
The only reason I use it is because there's not really any other FOSS that fits the same niches. I did learn about SolveSpace and this blender plugin today though, so I'll probably take a look.
Edit, found this article with a little explanation: http://mikeestee.com/2012/03/open-source-parametric-cad/
I wonder if there is currently a concentrated open-source effort in this direction, like a greenfield project for today's needs (including capabilities toward design optimization, subtractive and additive manufacturing, while addressing any shortcomings in OpenCascade).
I don’t really see a need to focus on subtractive and additive manufacturing in the kernel, although some hints in the UI for a particular program would be useful. Analysis and optimization predates 3D cad, in a way so it would be interesting to see what could be done from scratch.
I didn't expect much from Blender, owing to the shitty UI offered by other widely-used open-source software like GIMP and Audacity. But I was pleasantly surprised by Blender, and I'm excited to learn it.
But the CAD space is dominated by pay-only rental-style solutions, and even most of those seem to be Web-based (Autodesk) or Windows-only (SolidWorks).
So maybe one day GIMP and Audacity will get there.
And at some point you have to upgrade because the file formats change every year.
I always recommend Solidedge these days because it’s the only remaining capable option that has a simple monthly/yearly license.
it can also be used in blender, embedded into other applications. design cad as code, use git version control for your objects & designs. vs-code has plugins, or use the openscad
for many hackernews oriented persons this approach will be better than using a mouse.
Those things are really the exception in CAD. It would be masochistic to use OpenSCAD for the things CAD is more commonly used for (consumer products for example).
This is pretty cool and looks already very usable for basic things. So far I've usually worked with MeasureIt and lots of patience, but I'll definitely try this. If you're wondering how this works, here are some basic tutorials:
But in the end Blender is mesh based. So for round models you need a lot of vertices or else it will be inaccurate (subdivision modeling doesn't make it more accurate, it does only make it smoother).
I'm wondering to what degree tessellation is configurable, e.g. if I want polygon-dense output to have nice curvatures for additive manufacturing, or polygon-light output for rapid workflow and speed if I have complicated geometry.