improper XFail on svnadmin_tests.py:inconsistent_headers()

[kf...@collab.net]

In subversion/tests/clients/cmdline/svnadmin_test.py, the test
inconsistent_headers() is run as XFail.  But from looking at the test,
this is a misuse of XFail.

The test just makes sure that loading a dumpfile fails when that
dumpfile has a wrong 'Content-length'.  Obviously, the load should
fail, but this does not mean the test should be run XFail!  Subversion
is expected to behave a certain way (i.e., to error), and it does
behave that way.  Therefore, the test should test for the failure, and
simply succeed when the failure occurs as expected.

Before I fix this, are there any subtleties I'm missing?

r12644 is the place to look for more context, by the way.

-Karl

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: improper XFail on svnadmin_tests.py:inconsistent_headers()

[Erik Huelsmann <eh...@gmail.com>]

On 5/31/05, kfogel@collab.net <kf...@collab.net> wrote:
> In subversion/tests/clients/cmdline/svnadmin_test.py, the test
> inconsistent_headers() is run as XFail.  But from looking at the test,
> this is a misuse of XFail.
> 
> The test just makes sure that loading a dumpfile fails when that
> dumpfile has a wrong 'Content-length'.  Obviously, the load should
> fail, but this does not mean the test should be run XFail!  Subversion
> is expected to behave a certain way (i.e., to error), and it does
> behave that way.  Therefore, the test should test for the failure, and
> simply succeed when the failure occurs as expected.
> 
> Before I fix this, are there any subtleties I'm missing?

Nope. Subversion should error and, yes, therefore the test should pass.


bye,

Erik.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org