Fix unit errors in Examples.MoistAir #4115
Conversation
|
This would help address the issues reported in #4099 |
|
I'm not sure if this is needed - and would discuss it together with #4097 |
There was a problem hiding this comment.
The examples in MSL should also be seen as examples of how people should write models.
The previous example worked for that - the new one does not.
Telling users that they cannot write such a simple test (as currently), but must introduce a number of new declaration that provides no benefit doesn't seem nice and will instead turn off users.
It also tells users that unit-checking is a chore.
If we wanted to give users something better we could do that, but then we should have public parameters, and sweep between two states with different pressures (p1A=1e5 at time=0 and p1B=2e5 at time=1) instead of a pressure slope, since that is more intuitive.
And the states should have pressure, temperature and mass-fractions for consistency reasons.
Done correctly that would automatically be unit-correct, and indicate how to sweep a medium in a parameterized way.
Are you suggesting to modify the example this way as part of this PR? (I could do that. No problem.) |
I'm saying that the current PR is to me worse than the code in master branch - it's harder to understand, having lots of parameters with non-intuitive meaning, and just indicate to users that unit-checking is a chore that reduces readability. I'm suggesting that it may be possible to get something that is easier to understand, general, and unit-correct; see #4103 and #4117 - and also that it is complicated to have discussion of seemingly identical issues spread over multiple PR. |
hubertus65
dismissed
HansOlsson’s stale review
Proposed changes agreed on for merging in MAP-LIB meeting 2023-11-14
beutlich
force-pushed
the
examples-moistair-unit-error
branch
from
b774630 to
6e81a36
Compare
No description provided.