itaps-parallel Question about copies

[Saurabh Tendulkar]

Hmm, I noticed that the documentation of iMeshP_getCopies says that remote 
copies are to be returned. This could be the intent for the aforementioned 
functions as well since they are in the same vein. However, even 
iMeshP_getCopies has the same statement about ghost entities (that the ghost and 
its owner are to be considered), which now doesnt agree with the statement that 
only remote copies are to be returned.

We are currently implementing that the input entity is not considered in the 
output. We do not support ghost entities yet anyway, but I think the path of 
least inconsistency would be to treat that statement about ghost entities as a 
special case (ie. in general return only remote entities, but for ghost return 
the ghost and its owner).

saurabh


Tim Tautges wrote:
> I think that the statement from the documentation implies that an entity 
> will always have at least one "copy" (maybe that word is a bit ambiguous).
> 
> - tim
> 
> Saurabh Tendulkar wrote:
>> For functions such as iMeshP_getNumCopies and iMeshP_getCopyParts, 
>> should the input entity itself be considered in the output?
>>
>> The documentation says that "if the given entity is a ghost entity, 
>> the number of copies will be two (the ghost and its owner)". Extending 
>> this to non-ghost part boundary entities would suggest that the entity 
>> itself be included. Is this understanding correct?
>>
>> Along those lines, if the entity passed is not a part boundary or 
>> ghost entity, should iMeshP_getNumCopies return 0 (no other copies) or 
>> 1 (the entity itself)?
>>
>> Thanks,
>> saurabh
>>
>