(j3.2006) [POSSIBLE SPAM] Re: Coindexed scalars, etc.
Sun Jan 18 07:59:00 EST 2009
Whitlock, Stan wrote:
> Now about fields at co-indexed objects, does C(J)[p] mean "go get all of C
from p and find element J" or does it mean "go to p, get C(J), and bring it
back". I thought it meant the second.
Does it make any difference? If it means the first, the second is available as
an obvious optimization.
> So that means that C(J)%X[p] is correct, ie, just "get field X from C(I) on p".
For good reasons, we now use the term "coarray" for the object on a single
image, but the original use was for the whole set on all images. Let me call
this the "super-coarray". The model was that the super-coarray was accessed like
an array except that the indices and coindices were in separate lists. An
alternative notation that Bob considered was to put all the indices in one list
with a new symbol separating the parts, for example,
With this notation, you have to have
which corresponds to what we have now:
I think this works fine and see no need to change it. We have put restrictions
in the language to let the implementor hold the super-coarray in a very regular
way so that it can be addressed rather as if it were a big array.
More information about the J3