and then it uses that chosen revision for all the data loads it does for the problem and any subproblems