(j3.2006) Questions on interp F95/030 in (FORTRAN 2003 TC1 WG5 N1636)

[Van Snyder]

On Wed, 2007-11-28 at 13:34 -0500, Jim Xia wrote:
> 
> j3-bounces at j3-fortran.org wrote on 11/27/2007 01:54:26 PM:
> 
> > Hi Jim,
> > 
> > A user can supply his own specific interfaces for an intrinsic
> generic.  
> > I think the paragraph just says that the compiler does not have to
> look 
> > ahead through the rest of the declarations to be sure that the user
> has 
> > not supplied a matching specific, and thus what appears to be a 
> > reference to an intrinsic function really is that.
> 
> 
> This might well be the reason why this paragraph was written.  But
> still I see a great fallacy here: if a compiler knows that a non-
> intrinsic function will never be allowed in an initialization
> expression, then why even bother looking ahead?  The way this
> paragraph was written and inserted in the standard is odd.  It gives a
> reader an illusion that user defined functions can be used perfectly
> in initialization expressions. 

Intrinsics can be in generics, too.  The compiler doesn't have to look
forward for a generic that resulves to a specific intrinsic.

-- 
Van Snyder                    |  What fraction of Americans believe 
Van.Snyder at jpl.nasa.gov       |  Wrestling is real and NASA is fake?
Any alleged opinions are my own and have not been approved or
disapproved by JPL, CalTech, NASA, the President, or anybody else.