(j3.2006) (SC22WG5.3801) [ukfortran] N1761, TYPE(*), BIND(C) and arrays

[Bill Long]

Aleksandar Donev wrote:
> On Tuesday 09 December 2008 10:46, N.M. Maclaren wrote:
>   
>> In other words, N1761 has started to use the term "assumed-shape" to
>> refer to actual arguments, without specifying what it means when used
>> like that.
>>     
> I only found this one sentence where that was the case, and that 
> sentence simply needs to be fixed by replacing "object" with "dummy 
> argument". All the other appearences of "assumed-shape" in N1761 are 
> prefixed or suffixed with dummy argument. I just grepped.
> "A <<Fortran descriptor>> is an structure used by the processor to
> describe an object that is assumed-shape, assumed-rank, allocatable,
> or a data pointer."
>   

The "object" here is intentional.  Assumed-shape and assumed-rank 
objects are automatically dummy arguments; it's not really necessary to 
restate that fact. However,  C programmers can create new allocatable or 
pointer arrays that are then passed as actual arguments to a Fortran 
library routine, for example.  In this case, the associated Fortran 
descriptor is representing an actual argument, not a dummy argument.   
The C programmer can also create a descriptor for an array section that 
could be passed to a Fortran library routine that has a corresponding 
assumed-shape dummy.   Perhaps more examples would be helpful. Indeed, 
when we get to actual edits, I think healthy additions to Annex C would 
be useful. (As suggested by Reinhold.)

Cheers,
Bill

> Is this one mistaken sentence really the cause of such alarm? It is very 
> easy to find issues with something you dislike and blow them out of 
> proportion. It is not constructive. Notice how my answers to Reinhold's 
> posts are much more cooperative---at least I can understand what he is 
> asking and why. I have yet to understand exactly what you are irked 
> about.
>
> Best,
> Aleks
> _______________________________________________
> J3 mailing list
> J3 at j3-fortran.org
> http://j3-fortran.org/mailman/listinfo/j3
>   

-- 
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