On 19.07.2004, at 19:36, Greg Hudson wrote:
> On Mon, 2004-07-19 at 13:11, Ben Collins-Sussman wrote:
>> This test failure disturbs me. I understand that our testsuite
>> doesn't
>> compensate for different client/server versions. I understand that if
>> we added a new test that tested for a specific bugfix or new feature
>> in
>> 1.1, that such a test would fail against a 1.0 server. That makes
>> sense. But we're talking about a very old test, update_tests.py #15.
>> And it's out of 22 tests. Why is this very old test failing? Why is
>> everything OK? Did it used to be XFAIL in 1.0?
>
> The transactionless reporter calls set_target_revision when given the
> inputs in update_tests 15, while the old reporter did not. So, with a
> new server, the client outputs "Updated to revision 3" instead of
> "Update complete". So the test case was adjusted accordingly, causing
> it to succeed with a new server and fail with an old one.
Yes, i can confirm this. I got this error message against a 1.0.5
server:
CMD: svn "up" "working_copies/update_tests-15/A/D/G" "--config-dir"
"/sw/src/svn-ssl-1.1.0-rc1-11/subversion-1.1.0-rc1/subversion/tests/
clients/cmdline/local_tmp/config" <TIME = 0.765966>
EXPECTED STDOUT:
D working_copies/update_tests-15/A/D/G
Updated to revision 3.
ACTUAL STDOUT:
D working_copies/update_tests-15/A/D/G
Update complete.
EXCEPTION: SVNLineUnequal
FAIL: update_tests.py 15: another "hudson" problem: updates that delete
So i re-ran the tests against a 1.1.0-rc1 server and everything went
fine.
Thanks for the help,
Chris.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Jul 19 21:16:38 2004