Unify naming of "continuous states" and "derivatives" in API and XML

[t-sommer]

to clearly distinguish the different types of derivatives in

• fmi3GetContinuousStateDerivatives()
• fmi3GetOutputDerivatives()
• fmi3GetDirectionalDerivative()
• fmi3GetAdjointDerivative()

and

fixes #1343

[andreas-junghanns]

If you want to be correct, you have to call "states" "continuous-time states" and their derivatives "continuous-time state derivatives".
So your suggestion is now a compromise that uses longer names and still is not correct.
You even removed at least one "-time" where it makes perfect sense.

If we want to keep it short (and I would prefer short) we need to ask: is there a way to reasonably misunderstand what is ment. I see little danger of that.

[t-sommer]

If you want to be correct, you have to call "states" "continuous-time states" and their derivatives "continuous-time state derivatives".

The correct term would be "continuous w.r.t. the independent variable". But is continuous ever referring to something else in the document? If not we could just state that and use "continuous".

So your suggestion is now a compromise that uses longer names and still is not correct.

See above.

You even removed at least one "-time" where it makes perfect sense.

By accident. Fixed it.

[andreas-junghanns]

We do have discrete states. Well, discrete w.r.t. value and discrete w.r.t. time.
"Continuous states" could mean a discrete-time state with a continuous value (float/double).
To be clear, we tried to consistently name them "continuous-time states".

[t-sommer]

"Continuous states" could mean a discrete-time state with a continuous value (float/double).

The question is: do we currently or will we ever have to refer to these in the docs?

[andreas-junghanns]

"Continuous states" could mean a discrete-time state with a continuous value (float/double).

The question is: do we currently or will we ever have to refer to these in the docs?

We refer to discrete states, yes - there it is always clear what we mean.

[andreas-junghanns]

WEB design meeting: agreed with Torsten´s proposal.