Re: [Boost-build] v2 Python toolset doesn't faithfully reproduce v1

From: Aleksey Gurtovoy (agurtovoy_at_[hidden])
Date: 2006-09-23 12:10:25

David Abrahams <dave_at_[hidden]> writes:
> Now that I look at explicit-failures-markup.xml, I'm reconsidering.
> 
> Aleksey checked in an "expected failure" mark for each and every test
                                                ^^^^^^^^^^^^^^^^^^^^^^^
> with intel-7.1-linux and intel-8.0-linux. 

Not quite, only the tests that were exposing the "supposed linking
problem" (http://thread.gmane.org/gmane.comp.lib.boost.devel/110337/focus=110396)
are marked.

> The note says:
> 
>                 The library is <a
>                 href="http://article.gmane.org/gmane.comp.lib.boost.devel/110420">known
>                 to work</a> in this configuration. The failures are
>                 due to configuration issues of the particular testing
>                 environment these tests have been run in. The
>                 regression runners and library developers are aware of
>                 the problem and plan to fix it for the next release.
> 
> I don't know which developers Aleksey meant when he said they are
> "aware of the problem and plan to fix it for the next release," 

I specifically worded it as "regression runners and library
developers" -- in the post referenced in the note Ralf indicates that
in his experience Intel 8.0 requires special setup which might have
been all that was needed to fix this -- but I meant Martin and you.

> but I think we need to determine what those failures were and what
> the planned fixes were to be, and almost certainly we need to get
> this markup out of the XML file, because it's just masking problems.

No objections from me; it _was_ meant as a temporary markup for the
current (at the moment) release cycle only.

-- 
Aleksey Gurtovoy
MetaCommunications Engineering