Missleading Class-Subclass hierachy in SIDimensionalUnits

[markdoerr]

MassUnit, TemperatureUnit, LengthUnit, etc. are currently modelled as subclasses of SIDemsinalUnit.

This is not correct, as there exist non-SI MassUnits, LengthUnits, TemperatureUnits ....

Shouldn't the subclasses be better called "MassSIUnit", "TemperatureSIUnit", "LengthSIUnit" ?

[jesper-friis]

Agree, there exist non-SI MassUnits, LengthUnits, TemperatureUnits ....

However, the "SI" in the label of SIDimensionalUnit does not mean that it is the class of dimensional SI units (which we would have called DimensionalSIUnit). Instead it refer to dimensional units who's physical dimensionality is described according to SI.

[markdoerr]

Thanks, @jesper-friis , for the fast explanation.
Since this subtle difference, you described above, might not clear / obvious to many scientists, it should be worth an design explanation in the annotation of the SIDimensionalUnit class, e.g. like:
All derived subclasses of SIDemensionalIUnit are based on physical dimensionalities as described in the [SI specification](https://www.bipm.org/documents/20126/41483022/SI-Brochure-9-EN.pdf)