Reid,
I'm not sure why this is happening for you, but the important tidbit
from the logs you pasted was this:
1.
File
"C:\cygwin\home\Administrator\work\osaf\chandler\chandler\parcels\osaf\views\detail\detail.py",
line 1665, in SetAttributeValue
2.
assert False, "Hey - Setting 'ends' on an event without a
recurrence rule?"
3.
AssertionError: Hey - Setting 'ends' on an event without a
recurrence rule?
Python only runs assert statements in debug (or release without the -O
flag), so that's why you're only seeing it in debug. (If you ran
functional tests with do_tests.sh, and were only running 'release'
tests, it's set -O for you. I turn off the -O flag all the time, by doing
export OPTIMIZE=\
in my shell environment (not that there's a space after that backslash:
you're setting OPTIMIZE to a non-empty benign value).
From the rest of the traceback, I can see that the focus is changing at
a point where the recurrence-"Ends" text ctrl has the focus, and has
been changed, so it's trying to write the value back - there's an assert
there, though, that verifies that the item it's writing to is actually
recurring, and it's not, so it's raising the AssertionError you're
seeing. I don't know the sequence of steps that the test is going
through, so I don't know how this is happening, but maybe you can step
through the test in Wing to see.
...Bryan
Reid Ellis wrote:
Hi everyone,
With a virgin Windows XP checkout, I am seeing TestRecurringEvent
fail, but only for "release", not "debug".
I created a new install of Windows, with a user name with no spaces in
its name ("Administrator", courtesy of Parallels' "express install"
option), manually changed the OSTYPE env var to "cygwin" (default
seems to be "posix"?), and also applied the DNS changes Jared pointed
out to me earlier. The output from running the tests a few times is at:
http://rae.pastebin.ca/276732
The first run is both debug and release (default for
"tools/do_tests.sh -f", while later runs are debug (which passes
again) via "tools/do_tests.sh -f -m debug", and two "release" runs via
via "tools/do_tests.sh -f -m release".
I don't think it helps, but here's a screenshot as the final test ends
(post failure report):
http://www.flickr.com/photo_zoom.gne?id=320586463&size=o
Reid
On Dec 11, 2006, at 13:51, Reid Ellis wrote:
Excellent, thanks Jared. Now I'm failing on TestRecurringEvent. Ah well.
Perhaps we should have TestNewMail mention that URL in the case of a
failure? Is there a return value from mail.Check_DetailView() that
can be checked, allowing for a useful error message?
Reid
On Dec 11, 2006, at 12:18, Jared Rhine wrote:
Reid Ellis wrote:
Do I need to set up the Windows environment to produce a proper
fully-qualified domain name or something to pass the mail test?
Please report here if the instructions posted to this list:
http://lists.osafoundation.org/pipermail/chandler-dev/2006-November/007320.html
aren't working for you or aren't complete enough.
Thanks.
-- Jared
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Open Source Applications Foundation "chandler-dev" mailing list
http://lists.osafoundation.org/mailman/listinfo/chandler-dev
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Open Source Applications Foundation "chandler-dev" mailing list
http://lists.osafoundation.org/mailman/listinfo/chandler-dev
