Le mercredi 04 août 2010 à 16:28 +0100, Paul Moore a écrit :
On 4 August 2010 13:05, Antoine Pitrou <solipsis@pitrou.net> wrote:
I'm also quite confused by the test_smtpd failures that pop up on some of the test runs that I've had absolutely no luck reproducing locally under OS X or Solaris.
It happens when running test_smtplib before test_smtpb:
Nice! How did you work that out? I'd like to learn how to diagnose this sort of thing, because it seems to come up a lot, and I'm not much use at the moment :-)
I took a quick look at test_smtpd, and the one possibly fishy thing that stood out was the monkeypatching of the socket module using test.mock_socket. Since monkeypatching typically goes wrong when several people monkeypatch the same thing, and the other test.mock_socket user is test_smtplib, and since the buildbots run tests in random order (rather than in deterministic alphabetic order), I simply tried to run test_smtplib before test_smtpd. Regards Antoine.