I don't get it. If you put BOOST_ASSERT( false ) anywhere, then you effectively say "never exceute this code". Otherwise there is an error in the assertion itself.
Although that's a reasonable point of view, it is not supported by the current documentation, according to which executing BOOST_ASSERT(false) is well-defined and not an error. With some configurations it will invoke boost::assertion_failed(), which will pop up a message box offering to ignore, abort or break into the debugger.
As I said in my previous post, I find the current behaviour quite useful. I sometimes use assert as a convenient way of saying, "issue a warning message, but only in debug builds."
Same here. I'm using John Torjo's SMART_ASSERT and I have lines like: SMART_ASSERT(false)(a)(b)(c).msg("Surely we never reach here?"); I.e. if it does reach here I can see the values of a,b and c and can immediately see if it is a use case I never thought of, or a program bug. But instead of adding BOOST_ASSUME() I'd prefer the use of _assume() in BOOST_ASSERT be controlled by a define and off by default. It is a relatively minor optimization so could just be enabled by those who need the speed and know their code does not do ASSERT(false). Darren