Rotational: default values are now guesses + fix tests

[ven-k]

• Defaults in rotational lib are guesses
• Passes du0 and u0 for tests (in Rotational, Blocks)
• Fixes the linear analysis tests and docs (related to DC Motor example)

Checklist

• [x] Appropriate tests were added
• [x] Any code changes were done in a way that does not break public API
• [x] All documentation related to code changes were updated
• [x] The new code follows the contributor guidelines, in particular the SciML Style Guide and COLPRAC.
• [x] Any new documentation only uses public API

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 54.46%. Comparing base (13e4e93) to head (3a7ccfb). Report is 8 commits behind head on main.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #274 +/- ## ========================================== + Coverage 0.00% 54.46% +54.46% ========================================== Files 35 46 +11 Lines 1626 1634 +8 ========================================== + Hits 0 890 +890 + Misses 1626 744 -882 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[ven-k]

@ChrisRackauckas this is ready for review.

[ven-k]

Downgrade CI doesn't seem to be happy yet; but I will recheck up to which versions I can bump them to (as a follow up). I don't want to increase the compats to latest available versions; but keep them just above the minimum required ones.

[ven-k]

The failing Downgrade CI test is fixed by:

• https://github.com/SciML/ModelingToolkitStandardLibrary.jl/pull/276

[baggepinnen]

Considering how much of the user-facing documentation and tests had to be changed in this PR, I'd say this is a massively breaking change that requires a new major version?

---

Related, that the user now has to provide initial condition for every variable in every subsystem will make a lot of workflows significantly more unpleasant. For example, dragging components into a GUI and simulating will not be possible without clicking every single component to select the initial condition? Is this really how we want it to behave?

A simple option that uses guesses as defaults if no default is available would be useful to not slow down experimentation and exploration to an absolute grind.

---

I noticed that defaults were changed to guesses for rotational + two blocks from Blocks. The change makes some sense for the rotational components but not for the blocks. Was there a particular reasons these two blocks and only these were changed?

[ven-k]

I'd say this is a massively breaking change that requires a new major version?

This sounds good. Even the Units PR will be unblocked with this PR. So I'd say we can have that in too. (And probably guesses in other sub-libraries as well - with dedicated PRs for each libraries)

Was there a particular reasons these two blocks and only these were changed?

The idea is - incrementally change all the sub-libraries; tend the ones with broken tests first.

The analysis-point tests were failing as the rotational components now have guesses instead of defaults - fixed with op - while there, I changed these two as they were involved.

I've reverted the changes to Integrator and Derivative blocks.

[baggepinnen]

This was released in a minor version but it should have been a major version bump 😔

[ChrisRackauckas]

It probably should have.