Greg Bronevetsky wrote:
> At 04:35 PM 2/19/2008, Josh Hursey wrote:
>
>> They are available on the Wiki at the link below:
>> http://svn.mpi-forum.org/trac/mpi-forum-web/wiki/Meeting-15-Feb-2008
>>
>
> I've up just written up and uploaded the initial proposal for
> Transactional Messages. Give it a look. BTW, I don't know how to use
> this Wiki, so I've just uploaded my .txt file as a page attachment.
> If there's a cleaner way to do this, someone please let me know.
>
> Greg Bronevetsky
> Post-Doctoral Researcher
> 1028 Building 451
> Lawrence Livermore National Lab
> (925) 424-5756
> bronevetsky1_at_[hidden]
> _______________________________________________
> Mpi3-ft mailing list
> Mpi3-ft_at_[hidden]
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft
>
I've read the Transactional Messages proposal and I am a ittle confused
here. Is there a reason why we believe that message faults themselves
should be handled by the application layer instead of the MPI library?
Using the latter model allows one to reduce the error conditions
perculated up to the user to revolve around loss of the actual
connection to a process (or the actual process itself).
--td
|
![[Flag]](../../../../../images/flag-us.gif){kind=small}