BaseModelica should be a proper subset of Modelica

[christoff-buerger]

One very important concern for us (Dassault Systèmes, Dymola), is that BaseModelica is a proper subset of Modelica.

If I remember right, that is currently not the case for a few exceptions. We believe, that there are no strong reasons -- particularly no fundamental -- for such.

This issue is to collect where BaseModelica violates Modelica syntax or semantic and propose remedies.

We hope, that we can avoid splitting the language apart into two incompatible specifications, that existing Modelica tools will be capable to process BaseModelica models and a Modelica user can understand BaseModelica models out-of-the-box without any expert knowledge about exceptional rules.

[bilderbuchi]

Related: #3548