C++ PATCH: Disallow floating-point literals in integral-constant expressions

[Marcin Dalecki]

On 2005-02-01, at 02:25, Mike Stump wrote:
>
> If users have this in their code, then this can be seen as a removal 
> of a feature by them.  In that case, we should warn about it for a 
> release, then make it an error, or put it under -fpermissive.
>
> It'll take us around 3 months to notice if this is so used.  I'm 
> hoping that other people make just know, or can run a world build 
> against their favorite large source base (100+ M lines).

Well past time shows that a source base of 100+ M *always* needed some 
C++ adjustments
if compiling it with a new G++ releases. Thus the proposed exercise 
seems a bit futile.
I favor any removal of any "dark corner" G++ specific extensions 
blindly because this
programing language is far too exhaustive in features as it is already 
and strict standard
compliance gives at least some hope that this situation will 
asymptotically settle. It's
just too easy already to use such a kind of feature "unintentionally" 
in C++ right now.