At the moment all error codes are in one collection and some are re-used for various situations. This means to understand what error code a particular function might return, you have to look through every implementation, or assume that every call can return every error. It would be better for the interface to specify what errors can come from what calls.
At the moment all error codes are in one collection and some are re-used for various situations. This means to understand what error code a particular function might return, you have to look through every implementation, or assume that every call can return every error. It would be better for the interface to specify what errors can come from what calls.