Fix typo in IdealStartingState Javadoc (#1007)

[codydg]

I believe I've fixed the typo.

If there could be more clarification added, I'd be interested. I find the term "Ideal" a bit vague. Does it mean the "expected" starting state? That would make sense to me since I'm using it for on-the-fly generation, but I'm not sure if there are other use cases of the class.

[mjansen4857]

It is basically an expected starting state, but not required. It's basically only used for GUI trajectories to pre-generate the "ideal" trajectory that will be followed if the robot's starting state matches. Otherwise, it will generate a new trajectory to match its actual state. The IdealStartingState is not needed for on-the-fly trajectories, since those will need to be generated anyways.

If you want to add a note about that, feel free.

[mjansen4857]

Also, any changes to the javadoc probably need to be made to c++ and python docs as well.

[codydg]

The IdealStartingState is not needed for on-the-fly trajectories, since those will need to be generated anyways.

When I construct the PathPlannerPath, doesn't it make the time-based trajectory at that moment? I may have a core misunderstanding of how the generation works, but I'll give an example:

Generate a Path from <0, 0> to <1, 0> to <1, 10> (forwards 1, then left 10 units)
If I tell it my initial velocity is 10/s, I'd expect a different trajectory than if I said it's zero.

If I tell it null, it has already generated the path so I'm not sure how it would account for the 'sudden finding' at the start that I'm actually moving at 10/s.

Also, any changes to the javadoc probably need to be made to c++ and python docs as well.

I did check the C++ documentation which just says "Create a new ideal starting state", meaning no typo was there.
I had forgotten about Python but I just checked and it had the same typo as Java so I've now pushed a fix.

[mjansen4857]

The trajectory does not get generated until a path following command starts.