Repository navigation
McCabe Complexity #264
Description
Activity
I would recommend simply applying your own judgement. If you read a method that you think is too complex, for whatever reason, please open an issue to have it refactored. McCabe complexity has been shown to have no predictive power over a simple count of the number of source lines of code, and it's almost a truism that more code => more bugs. As far as I can tell from the literature, there is no empirical backing for the arbitrary limit of 10. Using McCabe complexity screening in this fashion just causes meaningless churn without actually improving the codebase.
Our collective subjective judgements about complexity are the actual ground truth here: if one of us reads a method and think it's too complex, then it's too complex. If no one who reads it thinks it's too complex, it's not too complex, no matter what McCabe says. And if no one reads it, then it doesn't matter how complex it is. :-)
This issue on another project puts it better than I can:
the output is:
The
sixstuff isn't our problem. But the other functions definitely look too complex.This is low priority but I thought it was interesting. This might be a good issue for someone to tackle if they want to become acquainted with the codebase.