just a joke, if somebody wants animate...

and finally I got it solved at very high accuracy. without any modification. nice assembly!

Hi All, my 2 cents on constraint direction: an assembly is usually built one constraint at a time, if the direction "aligned" matches the desired result fine, otherwise simply click on flip button. "none" direction is really useful when starting small assemblies or even bigger as...

Hi Turro, analyzing my problematic test case, if found a basic systematics error in solver. We have big mistake regarding planesCoincident constraint. Perhaps others too. Alignment of a planes should be only done by adjusting the axes. But the refPoint of a planesCoincident constraint is attracted ...

Hi Manuel, since in my branch I fully removed "none" direction forcing me (and the user) to select the wanted direction It was fast to me rebuild all constraints. I also set the flip direction button in the toolbar to make easier the switch to the other direction I think it should work eve...

Thanks to this I found an interesting behavior: During trials, I left solutiontoparts only at the end of final solution, ok sometimes I got the asm solved but angled part was not actually moved..... I found that when accuracy improves, the normalize() fails as vector axis is too small so the rotati...

Hi @Turro75, @anuelkrause, @wandererfan, i found a severe bug in rotation code of angledPlanes dependency. The rotation angle had been converted from radians to degrees twice. I think most of our problems with rotations were caused by this bug. The bug was already inside old master branch v.0.0.1 H...

In the property You can find on most constraints (where it make sense) the property Direction, which can be aligned, opposed or none. this is exactly what You need. none means use the closest to starting position, maybe auto would be more clear I agree that having a clickable button to flip directio...

I sent a PR regarding partial stage 3 implementation to turrodof3 branch, objs fully constrained on each other but disconnected to tempfixed parts are fused and then moved/solved together. it seems working quite well, accuracy level has been increased to 0.00001mm Hi Turro, i tested your new versio...

@kbwbe: Hi Klaus, I hope that I don't bother you with my not quite professional behaviour. It is sad that I wasn't able to reproduce malfunction of color-import on my system and today remembered a integer/float division issue with Python2/3 Turro75 stumbled across some weeks ago. So my fix in the P...

I sent a PR regarding partial stage 3 implementation to turrodof3 branch, objs fully constrained on each other but disconnected to tempfixed parts are fused and then moved/solved together. it seems working quite well, accuracy level has been increased to 0.00001mm Hi Turro, i tested your new versio...