coverity#735354 SQL_ODBC_SQL_CONFORMANCE "same on both sides"

[Lionel Elie Mamane]

On Thu, May 15, 2014 at 09:39:54AM +0200, Lionel Elie Mamane wrote:

> And the problem with *all* these functions is that they return a
> sal_Bool, and should thus be normalised to sal_True or sal_False. But
> they are not, and can return any non-zero value (or maybe it is always
> "1"? I think that's how C++ coerces a bool to an integer type) rather
> than sal_True (which is -1).

That's wrong. sal_True is 1, so the C++ bool-to-sal_Bool conversion
does the right thing.

-- 
Lionel