I still have an issues with seemingly random Applications (WECCCOMP_fault and Wave_Markers/RM3). When running the WECCCOMP_fault case locally, I see:

• wecSim command + R2023b --> passes
• wecSim command + R2023a --> passes
• TestWECCCOMP + R2023b --> fails
• TestWECCCOMP + R2023a --> fails

The WECCCOMP_fault test fails when using both the set-up from wecSimAppTest and a simple set-up:

t = TestWECCCOMP();
results = t.run();

Copying the set-up commands from wecSimAppTest is more robust though. It seems to keep the TestWECCCOMP.m file on the path and usable when changing directories.

Update: the error catching added to wecSim.m to allow MoorDyn 2.0 to work is what is causing the misleading error messages. I'll look at updating this so that the the error itself is thrown after MoorDyn is closed, giving useful debugging information

Now that I can see what errors are actually occurring, I've identified that the failing cases are related in that they all use the wave marker. There is a bug in the wave marker mask initialization command that only occurs when WEC-Sim is run in a function (like the tests). I will work on a fix shortly.

Of the tests that are currently failing:

• Wave Marker and WECCCOMP will pass on both R2020b and Latest once Bug fixes for v6.0 WEC-Sim#1136 is merged into dev
• MoorDyn and Paraview RM3 with MoorDyn will continue to fail on both R2020b and Latest because we currently do not have a .so file for MoorDyn 2.0, which is required for Linux. WEC-Sim Applications Tests are currently only running on Linux.

@akeeste I think we should remove the MoorDyn tests from the CI until they are running properly.

I'm going to merge this and then update the tests directly.