This is the high-level pull-request for the Whack-A-Mole Performance Feedback, that contains all the work of @mrvilix06 and @tcaruchet internship.
Before we perform a final merge, we will need to fix a number of items.
We need to correct the authorship of commit f709f10ff69be7b271c2fbb528282690e0d12ece
We need to squash most commits and split it like this: isolated bugfixes/features as separate commits, performancemanager+feedback+scene changes as a single separate commit (unless).
We need to properly document the functions of PerformanceManager.
We may want to open pull requests for important bugfixes separately and land them first, before the big new feature(s).
The branch introduce a regression on the countdowns, forcing us to use a STOP() at the end of the treatment programs. This should not be necessary. The regression needs to be tracked down. It most likely has something to do with changes to PatternPlayer.cs and/or PlayerPanel.cs.
We need to remember to revert the changes we made to Motorspace Small.
We observed that sometimes moles don't respect waiting times or countdowns, we need to track down what is causing this misbehavior, and whether it can be reproduced on main branch. We created a "Debug" test pattern on the AVALAB machine we can use to reproduce this behavior.
We should consider using some kind of sensible average for max/min speed, like fx averaging the 5 highest max speeds, rather than using a single datapoint to indicate the max.
Some experiment todos:
✅ We need to verify the transparency of the checkmarks in the task feedback corresponds to action feedback.
✅ We need to adopt the Medium motorspace for the study.
We need to verify the algorithm used for calculating distance judgements for actions - is it only the "ideal trajectory" or does it also use the accumulated distance that operation used? (bastian will verify)
✅ We may want to implement a "None" out of bounds indicator (simply not show any arrows at all). This can be used as a training/initial warmup to the first part of the study.
Speed up out-of-boundary arrow animation speed. (bastian needs to make tweaks)
Need to check whether its a side-effect or intentional that the pitch increases for each shot using operation-feedback.
This is the high-level pull-request for the Whack-A-Mole Performance Feedback, that contains all the work of @mrvilix06 and @tcaruchet internship.
Before we perform a final merge, we will need to fix a number of items.
STOP()at the end of the treatment programs. This should not be necessary. The regression needs to be tracked down. It most likely has something to do with changes toPatternPlayer.csand/orPlayerPanel.cs.Some experiment todos: