2017-05-15 0:54 GMT+02:00 Gavin Lambert via Boost
On 12/05/2017 04:19, charleyb123 wrote:
The formal review of Niall Douglas' Outcome library starts next week (Fri-19-May to Sun-28-May).
[...]
- What is your evaluation of the documentation?
Granted std::error_code does let you transport "foreign" error codes, but surely translating recognised codes from a method's internal implementation to those of its external interface is desirable? (Though perhaps that may depend somewhat on your views of encapsulation and implementation hiding.)
I think, breaking encapsulation in case of reporting failures from functions (be it an exception or an error code) is generally expected and desired, even though nobody will say it explicitly. When I want to communicate with another subsystem, I want the mechanisms of the communications to be abstracted away from me. But when it gets wrong, I want all the information: that I was in fact using the TCP/IP for communication; even exposed to the users, because they may know how to respond. Regards, &rzej;