OpenModelica Implementation Status

OpenModelica v.1.2.1, April 4th, 2005
Loading of whole Modelica library, both version 1.4 and version 2.?? Essentially full implementation of functions, both internal and external functions. Support for packages, import, inheritance, modifications, etc. Simulation of continuous subset of Modelica. Index reduction, solution of linear and nonlinear equations.

The following are known limitations in the current implementation:

Limitations in Executable Simulation code:

Event handling, only continuous subset of Modelica works
Limited error checking in numerical solver.
Initial values (start values, parameters) read from file only for reals
No dynamic selection of states (in index reduction)
Only possible to run with DASSL numerical solver.
initial equation can not be used. Set initial values using the start attribute instead, e.g. Real x(start=1);
All initial values must be consistent, i.e. no initial value calculation is performed. This can be done by hand and filled in into the <modelname>_init.txt file.

Unimplemented Features in Modelica:

Inner/Outer
Some excotic cases with Modifications, especially with redeclare.
Each modifier
Final element modification not checked against.
Constraining type on redeclarations
Restrictions on redeclarations (subtype etc.)
Derivatives of functions, using annotations.
Components of function type as block instances not implemented.
When clauses.
Restrictions on connections not checked.
The following operators in 3.4.2 have not been implemented:
smooth noEvent sample edge change reinit terminate mod delay cardinality
Reduction expressions,
Slice operation. Partly works but the resulting component names are sometimes messed up.
Vectorized call of functions not implemented.
Enable attribute for function output parameters ignored

Limited support for matrices and vector operations in equations. Array equality works, but not array operations such as *, +, -, /, transpose, etc. (only works if can be fully converted to scalar values.)
Array construction (3.4.4.1). Using array(... for ...) syntax. Partially implemented. Left to do:
- Constant evaluation and/or optimization. Could be implemented as function calls with index as argument. Or array can be unrolled and indexed, or if constant value can be calculated and indexed. Array should then be cached.
- Automatic type conversion of element type, e.g. (Real[]) a[5] = {(Integer)3 for ...} fails.
- Handle multiple iterators
Array constructors with iterators. Using {expr for i in ...} syntax.
Deduction of ranges (3.3.3.1). It is allowed to leave out "in range" in loops. e.g. for i loop ... end for; according to specification.
Multiple loop ranges, e.g.: for i in 1:5, j in 1:3 loop a[i,j]=i*j; end for;
Constant evaluation of record constructors for hierachical records.
Constant evaluation of record field dot ('.') operator.