Here is some suggested cosmetics. All corrections have original text
 with ---- and suggested new text that follows marked with ++++.
 Comments:
 1) and->or in front of Khoros is meant to make this list looking more casual 
 2) corrections in "II. Project Goals" are trying to impose better connectivity
    between WWVM, VSL and HPFCL. Not clear if that's the best way but 
    I think -2- should refer to -1- and -3- should refer to -1- and -2-
 3) Corrections in "III. Technical Summary" try to clarify a little bit better
    what WWVM is and how it evolves within VSL. One major suggested change
    is to say that we publish an early VSL on day one - that's in the spirit
    of previous exchange with Jim re what do we bring to the (Internet) table
    on day one etc.
 4) I guess we lost the phrase on "consumers being also producers". I heard
    it lately from an NSF officer and it is indeed enforced by webtools
    (via editors, managers etc.) so perhaps we restore it somehow somewhere?
 5) Perhaps we should say http://webtools available in March, not April?
    I guess this letter will be reviewed in March and so an April demo
    might sound as vaporware in March.
 - ----------------------------------------------------------------------------
 Title Page
 
 Webwork: Integrated Programming Environment Tools
 	for National and Grand Challenges
 
 Syracuse University -- NPAC
 Geoffrey Fox
 	3154432163 FAX 3154434741
 	gcf@npac.syr.edu
 	http://www.npac.syr.edu,http://www.infomall.org
 
 Syracuse University -- NPAC
 Wojtek Furmanski
 	3154433891
 
 Boston University -- Computer Science Department
 Marina Chen
 	617.353-8919 FAX 617.353-6457
 	mcchen@cs.bu.edu
 	http://cs-www.bu.edu
 
 Cooperating Systems Corporation
 James Cowie
 	(603)464-5799  Fax:(603)464-5799
 	cowie@cooperate.com
 	http://www.cooperate.com
 Submitted to Computer Science Challenges and Problem Solving
 Environments
 
 3 year Proposal -- First year budget $370K
 
 ABSTRACT
 
 Challenging-scale problems consistently demand solutions that fuse
 geographically distributed and heterogeneous data, personnel, expertise,
 and resources.  For example, national health care problems require
 collaboration among experts of fields as diverse as medical
 informatics, public policy, robotics, and high performance computing to
 solve problems ranging from telemedicine to cost management and quality
 control.  Indeed, many National Challenges include Grand Challenges as
 subcomponents.
 
 We can pose the Integrated Challenge as the solution of metaproblems
 hosted on world-wide metacomputers linking all three aspects: simulation,
 information processing, and collaboration.  We suggest a
 hybrid approach to Integrated Challenges that combines
 World-wide Web technologies with the current portable scalable software
 systems developed by the HPCC Initiative for Engineering and Scientific
 simulations.  
 
 The resultant system will support collaborative rapid
 prototyping, applied to programming environments at different
 scales ranging from support for systems such as ELLPACK (parallel
 - -------numerics PSE), MATLAB (uniprocessor PSE), and Khoros (signal
 +++++++numerics PSE), MATLAB (uniprocessor PSE), or Khoros (signal
 processing/visualization PSE) in the simulation arena to complex
 National Challenges including education, manufacturing and healthcare.
 
 Roughly we consider the world-wide metacomputer as a set of future
 generation WWW servers, each of which (individual workstation, 
 cluster, SMP, or MPP) is implemented internally (in the
 closely-coupled homogenenous environment) using appropriate HPCC
 technologies (e.g. HPF, MPI, PVM, Fortran-M, pC++, parallel Oracle,etc.). Then we link these servers together using generalized WWW
 technologies to allow executable program components to be published as
 services, and so create a distributed problem solving environment for
 prototyping large scale software.
 
 We therefore propose WebWork, an extensible world-wide virtual machine
 (WWVM) model for heterogeneous and distributed high performance
 computation. In our three year plan, we propose to initially develop
 the basic software engineering tools in a bootstrap fashion -- our
 first Webwork tools will accelerate the distributed development of
 further capabilities. Then we will prototype the computational services
 and test on two applications -- chosen as one Grand and one National
 challenge for which prototypes have already been developed using
 conventional methods and for which metacomputers and HPCC are
 appropriate approaches.
 
 
 
 I.  Web and Scalable Software Background
 
 This project combines a basis in scalable HPCC software standards such 
 as MPI and HPF with the pervasive power of agent based distributed
 computing of the evolving World Wide Web.  WebWork employs an innovative
 publication-based model of distributed high performance computing,
 transforming stateless document servers into general-purpose repositories 
 for computational expertise, and integrating Web technologies with HPCC 
 semantics.  By building atop a broad base of pervasive application 
 technology, WebWork will contribute to the "mainstreaming" of HPCC as 
 discussed in the January 1995 HPCC Workshop in Pasadena.
 
 II.  Project Goals
 
 In the WebWork Project, we will design 
 
 -    -1- a Web-based world-wide virtual machine architecture (WWVM), 
        defining coarse-grain HPC coordination in terms of publication 
        of new computational services and computational service requests, 
 -    -2- a Web-based, object-oriented virtual software laboratory (VSL), 
        built to be shared with the virtual community of Web software 
        developers, offering interactive forms-based testing and derivation
 - -------capabilities for new computational Web tools, and
 +++++++capabilities for new computational Web tools that contribute to the
 +++++++evolving WWVM, and
 
 -    -3- a prototype problem-solving environment for high
        performance computing based on High Performance Fortran Coordination
        Language (HPFCL) illustrating how we use the VSL to pose challenge-scale
 - -------problems and coordinate their collaborative solutions over the Web.
 +++++++problems and coordinate their collaborative solutions over the Web
 +++++++within the VVWM paradigm.
        This will illustrate how a hybrid technology (compilers and
        interpreters, MPI and HTTP) can be supported by single semantics and a
 - -------uniform user interface.  Our HPFCL implementation will be consistent
 - -------with emerging VRML Web standards for distributed physical object
 - -------representation.
 +++++++uniform and yet open user interface. This will be achieved by building
 +++++++HPFCL consistent with both HPCC and the emerging Web standards such as  
 +++++++VRML for distributed physical object representation.
 
 III. Technical Summary
 
 1.  World-Wide Virtual Machine (WWVM) and Virtual Software Laboratory (VSL)
 
 The World-Wide Virtual Machine starts from the existing baseline of Web
 services (servers that speak HTTP, plus client-side tools that speak
 HTML and MIME), and inherits that model's extensibility (the Common 
 Gateway and Common Client Interfaces, CGI and CCI).   From this basis,
 - ----we will construct a Virtual Software Laboratory for extending the WWVM's
 - ----repetoire.++++we will design core HPC coordination mechanisms and we will develop
 ++++a Virtual Software Laboratory for continuously extending this core WWVM's
 ++++repertoire.
 The VSL will provide unified text- and graphics-based interfaces 
 for composing WWVM primitive services with online engines that translate 
 requests for computational services into invocations of primitives, 
 to support the WWVM's innovative publication-based model of distributed 
 computation.
 
 The VSL will also support the publication of problem specifications as 
 WWVM primitives.  Like published services, published problems range from 
 conceptual descriptions to application source code and executable problem
 specifications.  Clients of WWVM sites will browse the published space of 
 unsolved problems to find subproblems that match the resources
 (computation and expertise) they can contribute.  As clients check out
 subproblems,  the WWVM notes their migration to remote locations;  when
 they are solved and returned, the WWVM reintegrates them into the 
 developing solution.
 
 VSL tools will also help developers debug their WWVM-based applications, 
 advertise their status to the world, and establish hyperlinks to peer efforts.
 As improved solutions to the more traditional challenges of single-site, 
 homogeneous high-performance computing emerge, VSL tools will help assimilate 
 them into the WWVM, as primitives supporting new WebWork services.  We will 
 - ----publish the Virtual Software Laboratory on the Web, packaging the growing 
 ++++publish an early version of the VSL on the Web on day one 
 ++++and then succesively refine it in the course of the project, by
 ++++packaging the growing
 space of WWVM primitives into a CGI-extended server, for other developers to 
 adopt and improve upon.  The resulting virtual community of WebWork software