Re: [Boost-testing] broken reporting of failure expectations

From: Aleksey Gurtovoy (agurtovoy_at_[hidden])
Date: 2005-07-08 23:38:11

David Abrahams writes:
> David Abrahams <dave_at_[hidden]> writes:
>
>>> Whatever the consensus is going to be, in the short term explicit
>>> entries are the way to go.
>>
>> Okay, I'll fix those up.
>
> Well, I did it, but I'm not really sure what I'm doing.  For one
> thing, none of the expected failures in the iterators library seem to
> show what the actual error was.  For example,
> http://tinyurl.com/936ec.

Do you mean the actual compilation error? There shouldn't be one,
since it's a "compile-fail" test.

> For another, there are a number of toolsets in the report, such as
> "gcc" and "mingw," whose version numbers are hard or impossible to
> determine. 

Yes, and these need to be renamed. As it has been numerously stated
here, having a toolset name conforming to our standard naming
conventions
(http://article.gmane.org/gmane.comp.lib.boost.testing/1321) is
essential to being able to mark up things properly.

Beman, Martin, Toon, could you please address this issue for the
toolsets you are testing with? (Please see
http://thread.gmane.org/gmane.comp.lib.boost.testing/877 for the
instructions).

> I can't really tell that the failure was what I expected.
> So I may have silenced quite a few dark green squares, but I'm not
> sure that what I did was meaningful.

Did you check it in? The markup in the CVS still uses '<toolset
name="*"/>'.

-- 
Aleksey Gurtovoy
MetaCommunications Engineering