Dave and Nick, latest text is general, with some examples, as Bill proposed and you, Dave, also requested. Okay so? Nick, I chose the examples mainly based on what is really used in the MPI standard. Fortran, yes procedures is the combination of functions and subroutines, change is really needed, because it is wording about Fortran. The other things in Fortran are not used in MPI, but still touched by the general word "names", because future MPI versions can use them. And I tried to sort the examples in similar sequence in all languages. C:

"Programs must not declare names (identifiers), e.g., for variables, functions, constants, types, or macros, beginning with the prefix MPI_."

Fortran:

"Programs must not declare names, e.g., for variables, procedures, parameters, derived types, abstract interfaces, or modules, beginning with the prefix MPI_.

C++

"Programs must not declare names (identifiers), e.g., for variables, functions, constants, types, or macros, in the namespace MPI."

Now okay? Rolf ----- Original Message -----

From: "David Solt" <[email protected]> To: "Rolf Rabenseifner" <[email protected]> Sent: Friday, August 24, 2012 7:21:20 PM Subject: Re: Fwd: [Mpi-forum] Reserved MPI_ prefix & namespace in C and Fortran I agree with Bill's approach of making it a more general ban on MPI_ use rather than attempting to list every specific case. Dave

----- Original Message -----

From: "N.M. Maclaren" <[email protected]> To: "Main MPI Forum mailing list" <[email protected]> Sent: Friday, August 24, 2012 10:07:38 PM Subject: Re: [Mpi-forum] Reserved MPI_ prefix & namespace in C and Fortran On Aug 24 2012, Rolf Rabenseifner wrote:

...

This a MPI-2 errata based on an existing inconsistency. Therefore we do not want to rewrite the three sections.

I missed that - sorry. I am juggling just too many things :-( My response was obviously in the context of MPI 3.0.

...

The current proposal is simple:

C:

...

"Programs must not declare names (identifiers), e.g., for variables, functions, constants, macros, types, or namespaces, beginning with the prefix MPI_."

'Namespaces' is just plain wrong. The minimum change to correct that is to change "macros, types, or namespaces" to macros, or types". You could, with benefit, add "labels, tags, members," somewhere.

...

Fortran:

...

"Programs must not declare names, e.g., for variables, parameters, functions, derived types, or abstract interfaces, beginning with the prefix MPI_.

'Functions' should be 'procedures' and you could, with benefit, add "modules, COMMON blocks," after it.

Regards, Nick Maclaren.

_______________________________________________ mpi-forum mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi-forum

-- Dr. Rolf Rabenseifner . . . . . . . . . .. email [email protected] High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530 University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65832 Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseifner Nobelstr. 19, D-70550 Stuttgart, Germany . (Office: Allmandring 30)