I would say, that in this case, with the 50\50 false positive ratio, it is quite manageable, i.e. for every 2 analyzer messages, there is 1 interesting issue in

I would say, that in this case, with the 50\50 false positive ratio, it is quite manageable, i.e. for every 2 analyzer messages, there is 1 interesting issue in the source code.