[MPI3 Fortran] Agenda for MPI3 Fortran Working group next week

[N.M. Maclaren]

>Jeff and I have worked out a list of items we need to discuss for next  
>week.  The goal is to finalize things so that we can begin to put the  
>API down in LaTex documents.  Some of this issues are pretty much  
>resolved but we should go over them one more time.

Here are some unasked for comments :-)

>1. Use of intent(IN/OUT/INOUT) for parameters.
>
>These requires careful thought in some circumstances.  For example,  
>should the Send/Recv and Isend/Irecv data buffers all be INOUT to keep  
>the compiler for doing surprising things?

In that case, yes, clearly.  I agree about the careful thought.

>4. What should the name of the new module be.  We had discussed not  
>changing the name, i.e.,
>
>      USE MPI
>
>However, users will have to explicitly change their code to use the  
>new module so I don't see how we can get away with using the old  
>module name.

I fully agree.  MPI3?

>5. Review what integer kind should be used for things like tags.   
>We've previously decided to use default integer kinds except for  
>possibly for counts, in which case we would specify a specific kind,  
>MPI_COUNT.

Are there any tokens representing opaque types left?  Other than counts,
those are the only ones that I can think of a strong reason for adding
the complication of a special KIND.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QH, England.
Email:  nmm1 at cam.ac.uk
Tel.:  +44 1223 334761    Fax:  +44 1223 334679