Re: [boost] Boost regression notification (2005-06-17)

From: Phil Richards (news_at_[hidden])
Date: 2005-06-17 13:32:00

On 2005-06-17, Douglas Gregor <dgregor_at_[hidden]> wrote:
>  Boost regression test failures
>  ------------------------------
>  Report time: 2005-06-16T15:27:18Z
[...]
>  Detailed report: 
>    http://engineering.meta-comm.com/boost-regression/CVS-HEAD/developer/issues.html
[...]
> |regex|
[...]
>    grep:  vc-7_1  vc-8_0
[...]

I don't understand this.  g++ will *not* compile the test -
and the detailed report shows that OSL2 confirms the test
fails for gcc-3.3.6-linux and gcc-3.4.4-linux (I can confirm it for
gcc-4.0.1pre-linux).

The "pass" seems to come from Martin Wille's regression run -
according to it, gcc passes for versions
gcc-2.95.3-stlport-4.5.3-linux gcc-3.2.3-linux gcc-3.3.6-linux
gcc-3.4.4-linux and gcc-4.0.0-linux.  The only one that fails is
gcc-2.95.3-linux.

Frankly, I don't believe Martin Wille's test result in this particular
case - any idea what's going on?

phil
-- 
change name before "@" to "phil" for email