App.h Private Members about fpsci

[bboudaoud-nv]

The variable m_targetDistance creates some issues here, since it should likely be moved to Experiment.h but is used in the JumpingEntity::create() call in App.cpp.

[bboudaoud-nv]

There are a number of variables in Experiment.h that can also be removed/protected. These include:

• m_app (move to private)
• mResultFileName (remove)
• m_lastMotionChangeAt (remove, doesn't seem to be used)

[bboudaoud-nv]

A recent commit fixes most of these issues, but leaves behind the targetDistance as a public field to the App so we can keep the JumpingEntity::create() in the same place and w/ the same signature.

[bboudaoud-nv]

On targetDistance here, we should think carefully about how we want to deal with this parameter. If we are passing scene control to the developer via a config file, perhaps we want to include this in a config as well. Currently the call chain for producing targets is a bit complicated, the Experiment calls the App's spawn[X]Target() method using parameters it has been passed from the PsychHelper's Param structure, which is created from the Experiment's config.

I may create a separate issue or link this to issue #11 in an attempt to better understand where the "right" place to spawn targets is. With our new structure the App doesn't necessarily seem like it should control much about spawned target location.

[bboudaoud-nv]

targetDistance was moved out of App.h as of a recent commit. By moving this value into the private member Experiment::m_targetDistance we have paved the way for inclusion of this value in the experiment config.

If we choose to move forward with adding the true (i.e. not projected) target distance to the experiment config we should think about what the right level at which to do this is (i.e. target parameter or high-level parameter). Since the scene is specified at the high level, I assume this would make sense (keeps target parameters modular from scene controls).

[bboudaoud-nv]

A recent commit cleans up the last few commented out/unused variables from the initial commit. Closing this issue in favor of moving all discussion regarding whether the app or experiment should own spawning targets to issue #11.