Freecad 0.18 daily-build and Realthunders topo-naming improvement and assembly3

[realthunder]

I am very aware of the licensing issue, which is why the SolveSpace solver backend exists as an independent Python extension. I even wrote a solver backend based on sympy and scipy, which is in fact a very rough python translation of the SolveSpace solver. It is very slow comparing to SolveSpace. If someone is motivated enough, he can further refine this, maybe using this for better speed.

[easyw-fc]

I am very aware of the licensing issue, which is why the SolveSpace solver backend exists as an independent Python extension.

These are good news! This is like DXF for FreeCAD
https://www.freecadweb.org/wiki/FreeCAD_and_DXF_Import

Legal Notice
To be able to import DXF-Files you have to manually install several files. Why this? These libraries are published with another license than FreeCAD, so the developers of FreeCAD would not have them being bundled with the FreeCAD-Application.

The SolveSpace can then be installed after FreeCAD instead of been bundled with it, and Assembly3 can just be easily used

[triplus]

First things need to mature. After i guess it won't be all that hard to decide what makes the most sense.

[sgrogan]

I am very aware of the licensing issue, which is why the SolveSpace solver backend exists as an independent Python extension. I even wrote a solver backend based on sympy and scipy, which is in fact a very rough python translation of the SolveSpace solver. It is very slow comparing to SolveSpace. If someone is motivated enough, he can further refine this, maybe using this for better speed.

Thanks for this!

I'm a fan of your whole development model. Documentation, distribution, answering questions on the forum, and yet you find time to code.

What is your opinion of including your work in master? Can it be done in stages? Or is it all too inter-related? I ask the question from "ignorance" I'm not qualified to judge.

[realthunder]

What is your opinion of including your work in master? Can it be done in stages? Or is it all too inter-related? I ask the question from "ignorance" I'm not qualified to judge.

I originally planned to submit PR in stages, but now it has been stretched for too long. There are bug fixes everywhere. It is going to be very hard to divide the patches. On the other hand, I always try my best to maintain backward compatibility for both the end user and the developer. So if that statement is proved to be true, there is no need to divide the patches.

[easyw-fc]

I think that FC0.18 should be used to switch to Py3 Qt5 and A3 and the switch should be done as soon as possible.
(obviously when the developers think it is ready for it)

Then the user testing audience will be enlarged ...

FC0.18 is already a 'daily' release, so instability and weird behavior are in user's count, as they were in FC0.17 PDN daily cycle...