[MERGE] weaves-over-hpss is just not supported, rather than an expected failure

[Martin Pool]

On Wed, Jul 16, 2008 at 12:10 AM, John Arbash Meinel
<john at arbash-meinel.com> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Martin Pool wrote:
>> Currently the test suite reports a KnownFailure for access to weaves over hpss.
>>
>> I don't think access to weaves in that way is something we intend to support in
>> future, so this patch downgrades it to just 'test not applicable', so you don't get
>> a message about it in a normal test suite run.
>>
>> I'd like to converge on having no known failures and that means having a clear
>> distinction between things we might plausibly fix and those we won't.
>> formats might not be supported with
>>
>
> I like your idea, I'm not sure how to handle it with OS issues. Specifically
> that we have symlink tests on Linux, and some Win32 globbing and FS tests on
> windows. We could make them clearly "OSFeature" or something like that.

Our documentation does say we will have a TestPlatformLimit to
distinguish them from NotApplicable, KnownFailure, and
UnavailableFeature etc.  It is not in the code yet but I'll see about
adding it.

-- 
Martin <http://launchpad.net/~mbp/>