itaps-parallel [Fwd: Re: Subcommittee for Parallel Interface Definition: Global IDs]

[Tim Tautges]

[Forwarding for archival]

-------- Original Message --------
Subject: Re: Subcommittee for Parallel Interface Definition: Global IDs
Date: Fri, 19 Oct 2007 11:41:11 -0600
From: Vitus Leung <vjleung at sandia.gov>
To: Onkar Sahni <osahni at scorec.rpi.edu>
CC: Devine, Karen D. <kddevin at sandia.gov>,   Mark Shephard 
<shephard at scorec.rpi.edu>,   Kenneth Jansen <kjansen at scorec.rpi.edu>, 
Tim Tautges <tautges at mcs.anl.gov>,   Carl Ollivier-Gooch 
<cfog at mech.ubc.ca>,   Lori A. Diachin <diachin2 at llnl.gov>,   Jason 
Kraftcheck <kraftche at cae.wisc.edu>,   Knupp, Patrick <pknupp at sandia.gov>
References: <C338DFBB.B5CD%kddevin at sandia.gov> 
<49549.128.113.131.145.1192482310.squirrel at www.scorec.rpi.edu>

Onkar,

Overall, the functions look good.  However, I wonder if moving some of
the functionality from the iProcParts level into the iPart level and
removing the iProcParts level would simplify things.  Similarly, I
wonder if the iPartMesh level should simply be the iMesh level.  At a
lower level, I found the use of both zero-based and one-based indexing
in Subsection 2.1 to be a little confusing.  Also, Subsection 3.2.1 is
titled "Return Number of Parts on Processor" when the first two of the
three functions actually return number of parts on all processors.  For
iMesh compatibility, the functions in Subsection 3.4 should take the
mesh instance as well as the set handle, and second-order adjacencies
should be provided above the iMesh level.

Best regards,
Vitus

On Mon, 2007-10-15 at 17:05 -0400, Onkar Sahni wrote:
> Hello all,
> 
> I have attached interface functions related to partition model. The
> document first provides basic terminology and then provides interface
> functions in three levels:
> 
> 1) Processor level (for multi-parts per processor).
> 2) Part level (part Id, processor rank etc.).
> 3) Part mesh level (entity level part information).
> 
> - Onkar
> 
> >
> > Honored Parallel Interface Subcommittee members:
> >
> > Please note that Step 0 of the Parallel Interface draft is due today.
> >
> > Step 0:  draft what functions are needed in each
> >          category; due Oct 15.
> > Step 1:  agree on what functions are needed in each
> >          category; due Oct. 18.
> > Step 2:  formalize the syntax for those functions;
> >          due Oct 25.
> > Step 3:  review and revise output from Steps 1 and 2,
> >          and write first draft; due Nov. 16.
> >
> > Global Queries:    Argonne (Tim)
> > Global IDs:        Sandia  (Vitus, Karen)
> > Partition Model:   RPI     (Mark, Ken, Onkar)
> > Partitioned Mesh:  UBC     (Carl)
> >
> > So we should circulate our draft of functions today and have a lively
> > discussion for the rest of the week.
> >
> > Toward that end, the Global IDs function list is attached (GlobalIDs.txt).
> > Please respond to all with your comments.
> >
> > The original email of tasks (draftzero_assignments.txt) is also attached
> > in
> > case you already deleted the email of assignments.  Let me know if you
> > need
> > more info.
> >
> > Karen
> >
> >
> >




-- 
================================================================
"You will keep in perfect peace him whose mind is
  steadfast, because he trusts in you."               Isaiah 26:3

             Tim Tautges            Argonne National Laboratory
         (tautges at mcs.anl.gov)      (telecommuting from UW-Madison)
         phone: (608) 263-8485      1500 Engineering Dr.
           fax: (608) 263-4499      Madison, WI 53706