Proxy exercises feature two algorithms to decide for an exercise instance presented to a learner: best-match and random. With the best-match option, however, there is a relevant performance problem: In order to determine the best-matching exercise, we consider the working time of previous attempts. This makes starting a proxy exercise in our production environment really slow and should be fixed.
Proxy exercises feature two algorithms to decide for an exercise instance presented to a learner: best-match and random. With the best-match option, however, there is a relevant performance problem: In order to determine the best-matching exercise, we consider the working time of previous attempts. This makes starting a proxy exercise in our production environment really slow and should be fixed.