Stefan Jaud
Stefan Jaud
See #334 as well.
Great issue description! A note to the developer: for *first-*, *second-*, and so on *-orderpolynomial*, a common function can be called by all with zeros at the corresponding locations.
Well that's just how the calculation is organized. Improvements are definitely possible. But I would argue that this currently has a low priority.
Please have a look [here](https://github.com/IFCRail/IFC-Rail-Unit-Test-Reference-Code/tree/master/alignment_testset/IFC-WithGeneratedGeometry). I believe that the example files checked in in our repository are not up-to-date. Task: if there are differences, *all* alignment unit test might need...
Expand `RepresentationConverter::convertIfcGeometricRepresentationItem` appropriately for `IfcCsgPrimitive3D` and call `SolidModelConverter::convertIfcCsgPrimitive3D`. Should work then.
Since we're renaming already, I'd call these variables `OIP_DOCUMENTATION_...`, not mentioning doxygen (it may change in the future).
See also: https://github.com/tumcms/Open-Infra-Platform/blob/1c104206852fae7d4a6ab5480344212ae38c5392/UserInterface/MainWindow.cpp#L608-L625
No, not yet. This changed quite a lot with IFC4x3_RC2, so it would need at least a check if not a rework.
This definitely looks interesting. I see multiple separate issues that can be addressed: - handling of I profiles: it seems that the code disregards the bottom/top flange inner point -...