[Boost-testing] Regressions: Are all this errors real?

From: Roland Schwarz (roland.schwarz_at_[hidden])
Date: 2007-01-13 08:56:26

I tried to get an overview of the outstanding regression failures. When 
looking closer some of the problems shown on:

http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/issues.html

don't look "real" to me.

Let me give examples:

date_time:
==========

Last run on 2006-12-20.
I also have an outstanding bug for Boost.Thread on this compiler. I 
already contacted the regression runner, and got the response that he 
intends to work on the issue "this weekend".
Is this really a compiler that belongs to the release set? Is it 
important enough, given the fact that seemingly none noticed that 
regressions had stalled for so long?

regex:
======

The reported test is passing without problems on my local setup. Anthony 
Williams, the regression runner told me that he is getting strange 
results with regression.py while invoking test manually works perfectly. 
He told me about this behavior in a related case. Could it be that this 
test is affected by the same problem, and does not exhibit a real bug?

test:
=====
When visiting the output of this page, one can see:
no errors detected

EXIT STATUS: 0

So what does this mean? Where is the error?

Roland