This came up in #830 regarding the implementation of storm::utility::isLess.
Might not always be correct: isLess(0.9, 1) with precision 0.2 yields False because of 0.9 < 1 - 0.2
Using value1 - precision < value2 as suggested in #768 would yield True because of 0.9-0.2 < 1. This might have some other problems though, for example isLess(0, 0.1) with precision 0.2 yields True.
Originally posted by @volkm in #830 (comment)
However, fixing this causes a plethora of tests to fail as we've likely worked around the buggy behavior preivously.