diff options
| author | Martin Ritchie <ritchiem@apache.org> | 2007-09-27 14:27:22 +0000 |
|---|---|---|
| committer | Martin Ritchie <ritchiem@apache.org> | 2007-09-27 14:27:22 +0000 |
| commit | 972614c8446ec8ec2a4f6c55fee4d5cc4b46e84b (patch) | |
| tree | f3b7bb28b44b9fbd4207bb0943307590a8313e64 /cpp/tests/FramingTest.cpp | |
| parent | 7286cb7d1d4f2f370c2df414cdb5150f9b658b00 (diff) | |
| download | qpid-python-972614c8446ec8ec2a4f6c55fee4d5cc4b46e84b.tar.gz | |
QPID-596 : ConnectionStartTest was broken. I've fixed it but here is the problem for those like me that like to know why:
Previously:
The setUp method created a producer connection and then sent a message
- This will result in that message being bounced as there is no consumer.
The first test should fail but the test was wrong, which caused it to pass.
There was an assert that was expecting the receive a message yet the test was recieve() == null !!!!
The second test worked because the broker was not killed between tests
This left the queue created so on the second run the message was delivered causing the test to succeed.
Now:
Fixed the InVM broker setup/teardown so the client is created first and the broker removed at the end of the test.
Also updated the asserts to be more explicit rather than having the == null or !=null put that as assertNull/NotNull.
git-svn-id: https://svn.apache.org/repos/asf/incubator/qpid/branches/M2.1@580022 13f79535-47bb-0310-9956-ffa450edef68
Diffstat (limited to 'cpp/tests/FramingTest.cpp')
0 files changed, 0 insertions, 0 deletions