According to our last call, we decided that we were ready to go forward with a reading of ticket #324 (Clarify MPI_ERRORS_ARE_FATAL scope of abort). Is that still
OK with everyone? Our other two open items are still longer term.
Is there anything else we need time for? Do we want to have some time to discuss ideas for the behavior of error handlers? I don’t know that there will be much
new to discuss with ULFM, but we can talk about it briefly to go over the new RMA / dynamic communicators text which should be drafted at that point.
Thanks,
Wesley
From: mpi-forum [mailto:mpi-forum-bounces@lists.mpi-forum.org]
On Behalf Of Schulz Martin
Sent: Wednesday, April 1, 2015 10:29 PM
To: Main mailing list
Subject: [Mpi-forum] Agenda for June Meeting
Hi all,
I added a first agenda skeleton for the June meeting to the MPI forum website at:
The details will change, but due to the nature of the meeting (where we want to ratify MPI 3.1) some things will be different than usual. In particular, we’ll start
with a plenary on Monday covering the voting rules (will send a draft for those soon) and the last changes to MPI standard (this way we’ll have extra time to make final fixes). Also, we’ll have two voting blocks, one on Wednesday and one on Thursday. Further,
we’ll have the requested discussion on MPI 4.0 on Thursday morning after the votes. Please take this into account when making travel arrangements.
This will necessarily limit the time we have for WGs and other plenary issues. We can shift between these two parts as needed, but for now, I reserved a 3/4 day
for each. WG chairs, please send me any requests for WG time, plenary time, reading requests and votes rather sooner than later, so we can plan accordingly.
Thanks!
Martin
________________________________________________________________________
Martin Schulz, schulzm@llnl.gov, http://scalability.llnl.gov/
CASC @ Lawrence Livermore National Laboratory, Livermore, USA