fluencyCAD current state
https://peer.raise-uav.com/videos/watch/70e4d8a0-0722-41e0-bf9e-a53d378e143b
fluencyCAD current state
https://peer.raise-uav.com/videos/watch/70e4d8a0-0722-41e0-bf9e-a53d378e143b
fluencyCAD in the wilds
I just mirrored the fluencyCAD repository to the open section form internal so anyone who is interested have a look. Don´t expect super clean stuff though. It´s in a very early experimental state.
I´ll add more documentation as I find more time.
@scudderfish I could never find my way through #cadquery. Too weird the syntax and concepts. I forked #sdfCAD and worked on something I understood myself but in the end stayed with #solveSpace.
Basic extrusions are working again with the all new object system. I´m trying to write the code as transparent as possible. This helps me picking up the project after longer pauses and others jumping in should have a good time too :D
I´ll mirror the repo to github once its usable for some basic tasks..
So far these are the components:
Ever since I worked on #sdfCAD¹ I have waited for the moment I could actually use #sdf's (signed distance functions) for my research. I know it would come. Today is that day. 🤩
This is for my last #PhD publication I'm working on.
Pushed fluencyCAD a tiny bit forward with a basic rewrite of the sketch widget.
Less code and prettier to work with internally.
@dl4mat This property of CAD software is not to be underestimated in value. Being able to build it quickly (or at all without jumping through ridiculous hoops) is a big plus in my eyes and I have only ever used CAD software like this: :openscad: #OpenSCAD, #sdfCAD, #SolveSpace
Worked a bit on making the basic functions pretty internally. Some GUI changes visible with the new timeline and sketch and body windows changed position.
It really would be a dream to get this thin gto abasic functionality one can work with <3
#fluencycad #sdfcad #solvespace
(Not sure if this reached the verse or not)
Introducing the new structure into fluencyCAD worked. We now have a timeline, components and the fundamentals for assembly sketched out.
Lots of staring at the wall and visits to the "thought emporium" (organic) were involved :D
The design guideline is based of the compiled experience of at least 10 years in multiple CAD programs.
~fluencyCAD~
Progressed into sketch projection from and to 2D. I think this is a cleaner approach for drawing.
Next stage will be additional operations like cutting holes and timeline logic.
Intention is to have CAD geared towards maker with a focused UI with static layout so you can develp "muscle memory".
I´m pretty happy already as I completely ruled out writing a CAD software at all.
Project can be assisted by donating for tea here:
https://ko-fi.com/raise_project
#sdfcad #solvespace #fluencycad #mechanicaldesign #vtk #python
@dcz Oh lol didn't know about that SdfCAD 😅. Click the hashtags #sdfCAD and #fluencyCAD (super new). My sdfCAD is here:
@dcz Damn, that's cool! The edges look sharp, so it's ray-marched, not a mesh made with marching cubes?
EDIT: Ah, ray-marched and you want to go directly to gcode, interesting! [1]
I firmly believe #CAD benefits a lot from #sdf|s and I love to see more projects like #sdfCAD and #fluencyCAD arise!
fluencyCAD
I added mid point constrains, hotizontal, vertical and point on line to test the underlying system.
I´ll look at selecting elements on the generated meshes first before attempting snapping of lines and intelligent drawing funtions etc.
Tiny steps it seems but there were quite a few designdecisions, regarding the internal structure, to be solved so that nothing falls apart when sthg is deleted etc.
Spent some more time on fluencyCAD(wt).
Current focus is to test different workflows and how the data is handled.
Next step would be to have a basic working cad for single bodys with included solve space solver.
sdfCAD is doing all the difficult and time intensive stuff underneath :)
@raise_project Definitely a lot of potential! I mostly use #SolveSpace currently due to the constraint-driven design, but I always miss #sdfCAD's powerful smoothing/chamfering and procedural powers. SolveSpace's STEP exports are also often broken (or impossible), so forced meshes are not a big problem for me.