blank feedback and feedback_content

[yananchen1989]

hi author,

in the file run_eval.py,
when I run the agent (human), i find that the obs.text , feedback from get_feedback and feedback_content from obs.game_states["option_mode_info"]["feedback_content"] are always blank.

Does this make sense ?

i trace the definition of get_feedback from the file predicate.py, looks like that return obs.game_states["option_mode_info"]["feedback"] is always executed, is this right ?
i suppose at each step, after the action being executed, there should be feedback with either success or failed, to indicate whether this action is executed successfully, right ?