[Mpi3-rma] Updated MPI-3 RMA proposal 1

Torsten Hoefler htor at illinois.edu
Sun Jun 20 15:20:56 CDT 2010

Hi Pavan,
> Are you planning to have an actual vote on proposal 1 without a straw  
> vote? What if the working group missed some detail and the Forum catches  
> it and rejects the entire proposal?
I didn't say that the proposal is ready for a vote at the Forum. I don't
think it is and I believe we need several full-Forum discussion sessions
with straw votes.

> I'm aware that we initially planned to group items that will likely not  
> be contended into proposal 1 -- I was in the group that made that  
> decision. However, we are dropping even related items on the grounds  
> that it might be contentious. For example, splitting flush/flushall and  
> allflushall. I'm arguing that this doesn't make sense anymore, so we  
> need some changes here.
We're not dropping it, we're just not forcing it all into one. The issue
is that we have two people who seems vehemently opposed. If we lump it
into one proposal 1 (the name is really bad :), then we don't give them
the chance to argue against it separately. If they wanted to rally
against it (and we had several rallies of this kind, just remember the
const proposal or the dist graph stuff), then they would need to kill
proposal 1 as a whole including all non-controversial things. I don't
think this would be fair. 

I don't understand why separating allflushall does not make sense. Could
you elaborate? I think flush(all) can stand by itself and allfushall is
an additional functionality to express collective semantics. Allflushall
indeed depends on flush(all) (which is not contentious), but not vice
versa. So splitting the black sheep seems to make sense :). Splitting
it up doesn't mean dropping it! I am willing to drive it forward and
can prepare a separate document. I personally think allflushall makes a
lot of sense but I'm editing the proposal on behalf of the group.

I also agree that it's not very useful to group proposals by political
issues. We could split proposal 1 up into purely technical parts (but
that would probably not change much :).

A strategic discussion about this would be helpful. 

See you tomorrow,

 bash$ :(){ :|:&};: --------------------- http://www.unixer.de/ -----
Torsten Hoefler         | Modeling and Simulation Lead
Blue Waters Directorate | University of Illinois
1205 W Clark Street     | Urbana, IL, 61801
NCSA Building           | +01 (217) 244-7736

More information about the mpiwg-rma mailing list