(j3.2006) request for interpretation

[Bill Long]

Kurt W Hirchert wrote:
> The text in subclause 12.4.4 was originally written after a thorough 
> consideration of the specific issue being now being addressed -- how 
> generics should interact with host association.  Indeed, the parts of 
> 12.4.4 not about that issue were written primarily to provide a 
> framework in which to address this specific issue.
.....

> My own inclination would be to also clarify the first sentence of 
> 16.2.3 to apply only to what 12.4.4.1 implies, but it if 16.4.1.3 is 
> fixed as I suggest, this is not absolutely necessary.
>

If I read this right, you are proposing that two specific interfaces in 
a generic can be identical if one is accessed by host association and 
the other is accessed by use association.  I would not call this a 
clarification, but a significant technical change to the rules we've 
been using for the past 15 years.   Such a change might be considered, 
in the sense that it relaxes a restriction, and hence does not make 
existing conforming codes suddenly nonconforming.  However, before 
entertaining such a change, I think there are some questions that need 
answers:

1) You say there was "thorough consideration", but what was the actual 
reason for this decision?  Was there a motivation for wanting to allow 
specifics that were indistinguishable by the 16.2.3 rules in a generic 
interface?

2) What are the costs and side effects of making the change in 
compilers?  Is the value gained worth it?

3) There is a potential land mine in the idea that an interface accessed 
by use association overrides an interface accessed by host association.  
In a submodule a procedure may access *its own* interface by host 
association.  Do we really understand all the possible side effects of 
changing 16.2.3?

My leaning is still to leave the rules in 16.2.3 as they are, and leave 
the rules in 12.4.4.1 to disambiguate on the characteristics not covered 
by 16.2.3.  This seems to be technically self consistent and unambiguous.

Cheers,
Bill


-- 
Bill Long                                   longb at cray.com
Fortran Technical Support    &              voice: 651-605-9024
Bioinformatics Software Development         fax:   651-605-9142
Cray Inc., 1340 Mendota Heights Rd., Mendota Heights, MN, 55120