Re: alternate simplified proposal #1 (fwd)

William C. Saphir (wcs@nas.nasa.gov)
Mon, 10 Jun 1996 14:34:23 -0700

On Jun 10, 12:23pm, Richard Frost wrote:
> Subject: Re: alternate simplified proposal #1 (fwd)
>
> In a few discussion groups on Friday, it was proposed that the
> "info" structure being proposed in mpi-dynamic become an explicit
> MPI opaque object (MPI_INFO).
>
> If this occurs, then a general MPI mechanism for constructing and
> manipulating {key, value} pairs could be used to specify info (e.g.,
> "hints") about the execution environment.
>
> I suppose this would go in the miscellany chapter. Any volunteers
> to write it?

I am going to be working with some MPI2 IO folks
and whomever else wants to participate to come up
with a proposal for the next meeting that merges
the proposals in the Dynamic and IO chapters. We should
be able to come up with something in the next
week or two. There are significant differences, so
this has to be done carefully.

I think the IO folks have hit upon the most important
requirement, which is that "hints do not change
semantics". I believe we may already have violated this
concept in the dynamic chapter, but by some houdini magic
we might still be able to claim we haven't.
Actually, that discussion can take place within the individual
chapters, and is orthogonal to the issues of syntax
that we need to resolve between the dynamic and IO chapters.

Bill