Doesn't the exception bubble up to the caller of create_connection()? So what's the problem?
On Wed, Jan 28, 2015 at 5:19 PM, Victor Stinner <victor.stin...@gmail.com> wrote: > Hi, > > 2015-01-27 23:48 GMT+01:00 Antoine Pitrou <solip...@pitrou.net>: > > This is a breach of the API's contract. > > I agree, that's why I opened an issue and start a thread on this mailing > list. > > > Once I proposed a separate connection_failed() for that purpose. > > It was shut down, but it can still be reintroduced. > > Good idea. I wrote a patch to add a new > Protocol.connection_failed(transport, exc) method: > http://bugs.python.org/issue23333 > > What do you think of this new method? > > It is called if the transport raises an exception before > protocol.connection_made() is called. > > Currently, my patch only calls the connection_failed() in the > coroutine handling an incoming client connection for a server. We may > also change other functions creating pairs of (transport, protocol). > These methods currently only call transport.close() on error. > > Victor > -- --Guido van Rossum (python.org/~guido)
