[MPI3 Fortran] Straw vote on integers kinds

[N.M. Maclaren]

On Sep 15 2009, Jeff Squyres wrote:
>
>> Also, though it's not really related, and is clearly a trivially  
>> fixable
>> oversight, MPI_ADDRESS_KIND is a pain, because it doesn't have an  
>> associated
>> MPI type and so is not easy to transfer between processes.  Teaching
>> attributes is not simplified by that gotcha :-(
>
>FWIW, it now exists in MPI-2.2 (although is problematic in  
>heterogeneous scenarios):

Thanks.  I will update my course.

Actually, I don't see that it is much more problematic than any other
type in heterogeneous scenarios - even with purely modern twos' complement
and IEEE 754 architectures, there are LOTS more gotchas than most people
realise.  Transferring a denormalised number from a system that supports
it to one that doesn't is a common cause of chaos.

Regards,
Nick Maclaren.