Hi Nick, On Fri, Feb 11, 2011 at 9:13 PM, Nick Mathewson <[email protected]> wrote:
> Well, something's certainly wrong here. I'm not sure whether to call > it a bug in the documentation or a bug in the code. My inclination is > to call the code correct and change the doxygen comment from > > "This function does not close the socket" > > to > > "This function closes the socket if you created the > evhttp_bound_socket using evhttp_bind_socket_with_handle or > evhttp_accept_socket_with_handle. This function frees the listener > you provided if you created the socket using evhttp_bind_listener." > > My rationale is that if we change the code, we risk introducing socket > leaks to programs that previously didn't have them, whereas any > programs that followed the documentation in this case would not get > any buggier than they were before. I could be wrong, though. Anybody > want to talk me out of it? > > Well, I agree with you(as this approach looks more flexible to me). -- WBR, Constantine
