[MPI3 Fortran] MPI-3, the Interoperability TR and choice buffers

[Bill Long]

N.M. Maclaren wrote:

> 
>> Why is the dope-vector and all behind not defined in a way that it
>> can handle 
>> - normal (scalar) varibles (as zero-ranked)
>> - multidimensional arrays and subarrays
>> - ***and*** ***also*** such assumed-size arrays?
> 
> That was my proposal.  Bill's CDESC would also solve this problem, and is
> probably the minimal change to the TR that would.
> 

Possibly a similarly minimal change, though one that would require 
careful wording, would be to allow an assumed-size actual to correspond 
to an assumed-rank dummy, with the rule that the extent of the last 
dimension in the descriptor be set to 1 in the caller.  This would 
require careful words about overindexing, etc.  But since assumed-rank 
is new, there should not be backwards compatibility problems. 
Basically, it boils down to a discussion of how to spell the attribute.

Cheers,
Bill


-- 
Bill Long                                           longb at cray.com
Fortran Technical Support    &                 voice: 651-605-9024
Bioinformatics Software Development            fax:   651-605-9142
Cray Inc./Cray Plaza, Suite 210/380 Jackson St./St. Paul, MN 55101