Francesco Casella

> I tried to check if those are already covered by the unit checking tests mentioned above, but the [Jenkins server](https://test.openmodelica.org/jenkins) seems to have disappeared behind a login wall?! Is...

@bilderbuchi the problem with literal multipliers is well-known. I raised it in the MAP-Lang group several years ago, but so far I failed to get a consensus on a viable...

@perost, there are many tests for unit checking in [testsuite/simulation/modelica/unitcheck](https://github.com/OpenModelica/OpenModelica/tree/master/testsuite/simulation/modelica/unitcheck), but I suspect they are not really being run. If you look, e.g., at [UnitCheck1.mos](https://github.com/OpenModelica/OpenModelica/blob/master/testsuite/simulation/modelica/unitcheck/UnitCheck1.mos), there are several weird things...

> We seem to have some old debug facility there, that doesn't seem to be used for anything at the moment. Probably that cref was added to signify that some...

So it looks like we are running 20 unit checking tests (plus another 19 with the OF, maybe we could get rid of them now), but we are not catching...

> Correct, there's a unit named jhagemann that's always added to the unit table for some reason. I don't know why, and there's no explanation in the code. Exegi monumentum...

The idea is that the built-in functions should have units associated with inputs (and fail unit checking the argument has a unit that is different from what is expected) and...

> > * sin(x), cos(x), tan(x) should have "rad" as input unit, and return an output with unit "1" > > * asin(x), acos(x), atan(x) should have input with unit...

This is an interesting issue. I'll tell you how I understand it through an example. It is perfectly legitimate to write a model like this ```modelica partial model M parameter...

Sounds good, let's see if we can do that in the next release