Re: [Boost-testing] [reports] Bogus unexpected pass results.

From: Aleksey Gurtovoy (agurtovoy_at_[hidden])
Date: 2005-08-04 15:01:06

Rene Rivera writes:
> Aleksey Gurtovoy wrote:
>> Rene Rivera writes:
>> 
>>>The following:
>>>
>>><http://engineering.meta-comm.com/boost-regression/CVS-RC_1_33_0/developer/static_assert_release.html>
>>>http://tinyurl.com/a8hjc
>>>
>>><http://engineering.meta-comm.com/boost-regression/CVS-RC_1_33_0/developer/range_release.html>
>>>http://tinyurl.com/8r3tv
>>>
>>>Show unexpected pass results for CW-8.3. But I removed those expected 
>>>failures from the markup a long time ago. What's going on?
>> 
>> 
>> Well, the note says:
>> 
>>     This test case used to fail in the reference ("last-known-good")
>>     release.
>> 
>> ... LKG release being 1.32.0.
>
> Yes. But those are explicitly fixed in the current code. So it seems
> incorrect to label them "unexpected pass" because they used to fail,
> but now they are fixed.

I agree that these should be somehow distinguished from "real"
unexpected failures. I've put it on the TODO list.

> Should I be explicitly marking those tests some way 
> to tell the reports that it's an "expected pass"?

You could make them go away by removing the corresponding failures
from $BOOST_ROOT/libs/expected_results.xml.

-- 
Aleksey Gurtovoy
MetaCommunications Engineering