Referee 1 *******************************************

E: Referee Comments (For Author and Editor)  ------------------------------

The weakest part of the paper is the section about supported Grid services. 
I would hope that this section would be significantly enhanced in a final submission.

Additionally, the authors should once more look over the template of the GCE working group
to see if they have addressed many of the issues implied by this template.

Some technical problems are present in section Section 4.

"Globus IO (formerly Nexus)"

it is unclear if this means that an erlier version of SciRun used Nexus or if the author
means that GlobusIO is called formerly Nexus. Looking at the Web site the later statement is
certainly not true.

The Blazer mechanism is not described at all. Either remove it or explain what the Blazer
mechanism is.

A section about related work should be added.
Who does scirun comare to xplorer or AVS, Labview, or others  

F: Presentation Changes

Many acronyms have been introduced in the paper without ever explaining them. I recommend
that the author goes once more carefully through every capitalized word and verifies if he
has defined it upon its first use.

The quality of all Fig. is week and hopefully can be improved in the final journal
publication. Most of them are unradble.

There are minor orthographic error such as on page 5, where the the should be changed to
that the

Page 8 "C-SAFE was create by" change to was created by.

Itemized lists are always closed with "." See section 5 . missing behind different

I believe the list of reverences must be significantly enhanced. References to blazer, and
CCA should probably be augmented with references. This would make the paper more valuable.
Such references should occur typically when the project or term is used the first time in
the paper.
  	

Acckn.
	Would you have to cite the current contract numbers, I know we have to.
	

Referee 2 *******************************************


Component-Based Problem Solving Environments for Large-Scale Scientific Computing

Overview :

This paper gives a description of design and implementation of the
SCIRun PSE, BioPSE, and Uintah. Fundamental concepts in scientific
visualization are covered as well as concepts of object oriented
design and practical implementation.


Review :

The paper gives an excellent explanation of basic problems with
the visualization of enormous datasets and the practical limitations
associated with design decisions. The daunting task of implementing 
a practical, parallel, flexible system are explained well. SCIRun
is a fairly mature body of software and this paper reflects that fact.

Uintah sounds particularly interesting and the focus of a major
initiative. From the description in that paper, Unitah is in the
design and initial implementation phase. This will certainly be
and interesting body of software based on its' genealogy !


Recommendation :

Publish as is