Make propagation final of lifted parameters in Spring #4407
Conversation
|
By the way, the discussion in MO-2748 didn't take us much further. It gives us a language to express initialization problems more clearly in Base Modelica, but doesn't provide a full specification of how full Modelica initialization should be turned into Base Modelica. |
There was a problem hiding this comment.
The thing is that I Joe Engineer would never come to the idea c is final due to a clear initialization.
@MartinOtter could you review here?
|
It seems like we are not getting the attention of @MartinOtter on this one. @casella or @AHaumer, may I kindly ask for your help to step in and drive this forward? |
There was a problem hiding this comment.
Fine with me
@tobolar I don't understand your comment.
c is a simple parameter that is propagated. It is not a good idea to let the user change Spring.spring.c to a different value than Spring.c. What do you mean by "initialization"?
Indeed, the common user would be right to wonder. If also the modification of If that helps, I'd gladly also add |
There was a problem hiding this comment.
What about the other lifted parameters of the lineForce component? Seems arbitrary to me to have final modifications at some place but not at other. I asked the same question in #2307 with @MartinOtter's reply to go for final parameter modifications consistently.
…ponent Even though the modification of lineShapeType is not parameter lifting, it would make little sense to make it non-final when all other modifications are final.
…e SpringDamperParallel
Done. Also addressed a similar issue in Once this PR is merged, it can serve as a starting point that other PR's can refer to as motivation for doing similar cleanup in other models. |
|
Well, three approvals and green CI checks ... |
beutlich
changed the title
beutlich
added
the
L: Mechanics.MultiBody
|
@maltelenz, please back-port to maint/4.1.x |
|
@Esther-Devakirubai this should be mentioned in the release notes. In general it's not going to be a problem, unless for some reason somebody changes the low-level parameters (that are now final) instead of the propagated parameters at the higher level. |
* Make propagation final of lifted parameter 'c' in Spring Fixes modelica#3707. * Make propagation final of lifted parameter 's_rel0' in Spring * Make propagation final of lifted parameters in Spring's lineForce component Even though the modification of lineShapeType is not parameter lifting, it would make little sense to make it non-final when all other modifications are final. * Make propagation of lifted parameter 'd' final in ModelicaTest example SpringDamperParallel --------- Co-authored-by: Anton Haumer <[email protected]> (cherry picked from commit a5cf0f3)
Done by #4520. |
Fixes #3707.
As there wasn't any arguments agains making the propagation of
cfinal in #3707, I suggest that we go with this solution.One could argue that there are more parameters whose propagation should be final, but I'm starting with this one which is sufficient to fix the issue.