Subject:
our revised paper
From:
"Elias Houstis" <enh@cs.purdue.edu>
Date:
Mon, 3 Dec 2001 17:16:02 -0600
To:
"Geoffrey Fox" <gcf@indiana.edu>
CC:
jrr@cs.purdue.edu, "Naren Ramakrishnan" <ramakris@cs.vt.edu>

Geoffrey hi,
I am attaching our revised paper. 
We have undertaken seriously to revise the paper to address the concerns
of the reviewers. To ensure that the article is widely appreciated as
describing a methodology, we have added paras in both Sec 1 and Sec 2
outlining the "grid context" in which MyPYTHIA is to be used. This
addresses the concerns raised by reviewer #2. With this background, it
will be clear that MyPYTHIA provides tools to support data-driven inquiry
of simulations conducted elswehere on the grid. Reviewer #1's concern is
whether such a mode of inquiry is already happening on the grid. To some
extent, this is true in particular domains (again, we have clarified
this) but in other cases, our system can help usefully harness and catalog
large runs of simulations. Moreover, our system is not merely an
architecture; a running implementation is available on the web for which
we have provided an URL.

Finally, reviewer #3 is looking for computational services, whereas
MyPYTHIA is more an informational service about how computations should be
done. As such, it is applicable in forming an abstraction of how grid
services should be intelligently chosen. We have addressed this reviewer's
concern by saying that "MyPYTHIA's contribution is in the modeling of the
recommendation problem." This sets it apart from existing grid services.

Ultimately a "grid computing environment" (the focus of the special
issue) should be qualified by the asbtraction it provides of underyling
grid lower-level functionality. MyPYTHIA provides such an abstraction in
the form of recommending software services. The context in which the
recommended software is actually executed is important but orthogonal to
MyPYTHIA's design philosophy.

Thanks
Elias