🚀 go-pugleaf

I've read that parallel processor implementations of PROLOG machines use some
variant of breadth first search. I was wondering if anybody has designed a
parallel implementation using DFID (Depth First Iterative Deepening). Because
it has been proven that DFID is the asymptotically optimal brute-force tree
search algorithm (asymptotically optimal in cost of solution, space, and time),
I was thinking that perhaps it may have usefulness in parallelism. 

Would it be worth while for me to try and develop an DFID-PROLOG for a single
processor, e.g., on my Apple Macintosh? Or are their some problems that would
would make such a PROLOG to large for the Mac? Is the algorithm applicable
to a parallel PROLOG?

						George Van Treeck
						Digtal Equip. Corp.

PS: If you're not familiar with the algorithm, it's description and proof
of optimality can be found in: ARTIFICAL INTELLIGENCE 27(1985) 97-109