>> * I find it a bit safer when the error predicate is “return value != 0”. > > Can't agree.
How do you think about to reduce the probability that positive return values will accidentally be interpreted as a successful function execution. > And I have no interest to continue bike-shedding, sorry. I do not like that you prefer to put this technical detail into such a communication category. > You can't convince me regarding this. Would you still like to integrate the proposed refactoring with the use of previous failure predicates then? Regards, Markus
