itaps-parallel Diagram of set-based parallel data

[Tim Tautges]

The following was written by Mark S. yesterday:

"We very much use a partition model without sets or explicit storage of 
reverse classification. Reasons for this are explained in various papers. "

When you say parts are represented explicitly, do you mean as 
classification (each entity in the part having a value denoting the part 
it's in) or a list somewhere having all the entities in a given part?

- tim

Onkar Sahni wrote:
> I do not really know what the question is but its like asking why we use
> entity-handle in the interface. In our implementation
> (classification-based), parts and partitions are represented explicitly.
> 
> - Onkar
> 
>> Why must we expose parts and partitions through the interface?  In a
>> classification-based implementation they aren't represented explicitly
>> either.  Working with mesh-based data is just as much working the
>> metadata on the mesh as it is working with the mesh itself, and sets are
>> fundamental to many kinds of metadata.
>>
>> - tim
> 
> 
> 

-- 
================================================================
"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