I am in favor of removing JoD from the final MPI-2 document,
but I strongly suggest that we distribute it for SC'96.
The goal of SC'96 is to give a clear picture to all interested
parties, that includes ongoing work and work that we do not want all
implementation to have, but that can be added by individual implementors
for added benefit for users...
It is true that we should not create MPI-3 after MPI-2 right away
and let user to get familiar with MPI-2 features...
However, real-time MPI will continue to work on its chapter(s)
since it is not required for all MPI implementations. I invision
that MPI I/O will continue its work if it will not be approved as
part of MPI-2.
Sincerely yours,
Arkady
>> I make the following proposal. Since the JOD document is the core of
MPI-3,
>> and since it may need chapter stratification, I suggest that it be
separated
>> from MPI-2 document, and be physically attached, rather than being jointly
>> managed.
>
>I think that it would be a *very* big mistake to say anything at all about
the
>JOD with respect to MPI-3. In fact I think that even mentioning MPI-3 at this
>point should be grounds for...I dunno, but something very unpleasant! MPI-2
is
>already so large and so complex that we can't even get firm commitments to
>implement all or even most of it from the current MPI-1 implementors.
>
>The JOD is for things that have been taken *out* of MPI, and I am personally
>opposed to even including it as part of the SC'96 release. So I favor
anything
>that puts distance between it and the "real" MPI chapters.
>
>--
>Eric Salo Silicon Graphics Inc. "Do you know what the
>(415)933-2998 2011 N. Shoreline Blvd, 8U-802 last Xon said, just
>salo@sgi.com Mountain View, CA 94043-1389 before he died?"
>
>