Hi anteloc, congrats on some cool projects! Look forward to playing with them over the weekend.
pyldraw3 creator here btw... would love to find a forum where we (i.e. the handful of us trying to push lego tooling forward) can collaborate and communicate better.
My current main effort is deterministic model/instructions building but I've been thinking a lot about how to incorporate agents: https://github.com/hbmartin/legolization
Your pyldraw3 is a key component of both Mecha and Nova, together with the shadow library from LDCad.
Parsing LDraw, collision detection and connectivity between different parts mostly rely on pyldraw3 for validation.
About your other project, legolization, I took a look at it, and I think it's a very interesting approach, I've been thinking for a while about solvers and how to predict stability, too!
However, so far didn't find anything that would fit in, due to the high amount of parts interacting in Mecha animations.
Tried a couple of solvers, but with +1000 parts, performance degrades drastically, and a divide-conquer approach I still didn't have a chance to try.
And about finding common ground for LEGO tool makers, sure!
What is currently happening with agentic AIs quickly evolving CAD capabilities will for sure benefit LEGO and LDraw tooling.
Hi anteloc, congrats on some cool projects! Look forward to playing with them over the weekend.
pyldraw3 creator here btw... would love to find a forum where we (i.e. the handful of us trying to push lego tooling forward) can collaborate and communicate better.
My current main effort is deterministic model/instructions building but I've been thinking a lot about how to incorporate agents: https://github.com/hbmartin/legolization
Hi hmartin, thanks!
Video for Mecha below, it was not there when you posted:
https://youtu.be/P4aKXKNOBig
Your pyldraw3 is a key component of both Mecha and Nova, together with the shadow library from LDCad.
Parsing LDraw, collision detection and connectivity between different parts mostly rely on pyldraw3 for validation.
About your other project, legolization, I took a look at it, and I think it's a very interesting approach, I've been thinking for a while about solvers and how to predict stability, too!
However, so far didn't find anything that would fit in, due to the high amount of parts interacting in Mecha animations.
Tried a couple of solvers, but with +1000 parts, performance degrades drastically, and a divide-conquer approach I still didn't have a chance to try.
And about finding common ground for LEGO tool makers, sure!
What is currently happening with agentic AIs quickly evolving CAD capabilities will for sure benefit LEGO and LDraw tooling.