Add importance weighting to rules for more robust "fails"
Instead of a binary critical or not critical flag for rules, it would be nice to have a weighting system for rules (e.g. "each violation of this rule is worth 3, but each violation of this important rule is worth 8").
Then perhaps we could assign a threshold for whether our analysis failed based on point totals, or some percentage formula.
Robert Leahey
shared this idea
with the technical debt support of v2017 that let’s assign a cost-to-fix and a severity estimation for each issue
-
Robert Leahey, Rob Smyth, the next version we currently prepare will handle hat in a customizable and advanced way, please come back to us to private beta test if you want (support@ndepend.com)
-
Rob Smyth
commented
Yes. This would also help measure code health improvement on a legacy code base over a long period. Fail flags are good for day to day use but a weighting gives a trackable metric.