On Wed, 5 Jul 2006 15:26:41 +0100, "Reece Dunn" <msclrhd@hotmail.com> wrote:

Gennaro Prota wrote:

...

On Tue, 04 Jul 2006 15:12:43 +0200, Gennaro Prota <gennaro_prota@yahoo.com> wrote:

...

[something...]

I always wonder how a lack of answers should be taken: (a) nobody disagrees (b) nobody agrees (c) nobody cares.

How about: (d) they are busy, so haven't had chance to look at it yet; (e) they haven't noticed it because, since last they checked, there they had 50+ new e-mail from Boost only!

That could be :) (you might mitigate that problem if using the NNTP interface, perhaps)

...

Hint: (c) is impossible, as *I* care, and for that reason I'll take (a) by default ;)

As for the original question, I would personally prefer:

c) # define BOOST_ASSERT(expr) assert("BOOST_LIBRARIES:" && (expr))

but with "BOOST_LIBRARIES:" something more meaningful/readible, like "Boost assertion failure: ".

I used all-uppercase just to make the message stand out more. "Assertion failure" will likely appear anyway (though the exact message is implementation dependent) so I just added some text to specify the "source" of the assertion. -- [ Gennaro Prota, C++ developer for hire ] [ resume: available on request ]