Erroneous equation count with arrays

[massimo ceraolo]

The following model:

model errorSize
Real x[:](start={ 0.2, 0.4, 0.6});
equation
  der(x[1:3])+x[1:3]=fill(1/(time+0.1),3);
end errorSize;

gives rise to the following result when checking:

Check of errorSize completed successfully.
Class errorSize has 3 equation(s) and 1 variable(s).
1 of these are trivial equation(s).

and the model obviously won't run.
AFAIK this is valid modelica

If I repeat checking with -d=newInst I get, instead:

[1] 16:20:30 Translation Error
[Es1: 2:1-2:34]: Failed to deduce dimension 1 of x due to missing binding equation.

Tested with OM v1.13.0-dev-380-g8b177b9 (WIN 64-bit)

[Per Östlund]

The specification doesn't say anything about how to deduce unknown dimensions in non-functions, see ​this ticket, so this is not strictly valid Modelica. However, we already deduce unknown dimensions from binding equations, since that's the de facto standard. Since Dymola also allows using the start attribute for this I added the same behaviour for the new frontend in 70d5639f. Note that the example model still doesn't simulate with the new frontend due to other unrelated issues though.

[Francesco Casella]

Rescheduled to 1.14.0 after 1.13.0 releasee

[Francesco Casella]

Releasing 1.14.0 which is stable and has many improvements w.r.t. 1.13.2. This issue is rescheduled to 1.16.0

[massimo ceraolo]

This model now simulates with the NF