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

E: Referee Comments (For Author and Editor)

My principal complaint is that the paper as submitted is very unfinished. 
While clearly relevant, the quality of the final paper cannot really be
judged on basis of this version. 

The paper describes the development kit for Grid-based computational
science portals, and as such fits well within the topic range of this
special issue of CCP&E.  The paper describes various elements of the 
development kit, often in great detail. In this sense the paper is 
somwhere between a manual (for which it has too little detail) and
a paper describing the concepts (where the level of detail becomes
confusing).

Sadly, the bibliographic references in the body of the paper were
empty, which doesn't help in reviewing.

In section 2 (Overview of the Grid Portal Development Kit), the author
briefly mentions cookies and says that they will be discussed in more
detail in section 2.1.  I can't find any mention of cookies in there.

Section 4 appears to be (partly?) missing (See below).

F: Presentation Changes

The paper doesn't use the right style for typesetting and the figures
are of unsatisfactory quality.

Either there's something wrong with section numbering, or some pages
are missing.  Namely, there is no section 4, section 3 (GPDK as a
portal development environment) is followed immediately by section 5
(Related work).

The quality of the figures is poor.

I spotted several typos in section 2.3.2 (Job Submission) and 2.3.3
(File Transfer): "GSI enhance", "idf", "succesfully", "coommand".

More references are needed - e.g. for ANT

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

D: Referee Comments (For Editor Only) ------------------------------------

    This is a very good paper, well written.  Technical ideas are good and timely, with
    enough information given so the reader can understand what was done. I particularly
    appreciate that he cited other related work, and in general references are
    sufficient (better than average).

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

F: Presentation Changes

    (minor) font size in figures should be larger
    (minor) should give web address of GPDK home page?

Referee 3 *******************************************

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

This paper describes a development toolkit for building grid portals for
applications. The toolkit is built on top of Globus middleware using Java.

The paper is restricted to discussing features of Java that the toolkit
used. For example, most of the meaty Section 2 is devoted to describing
server pages, servlets and Java beans. While such a discussion is certainly
necessary, it should not be the major focus of this paper. Instead, the
paper should bring out interesting grid-related issues in the development
of the toolkit. Likewise, it's not clear what the reader is supposed to
glean from Section 3. The section does not show techniques by which the
toolkit can be used by others; it does not show how the toolkit was used in
example; it does not contain any evaluation metrics for the toolkit or
portals developed using it. Perhaps the missing Section 4 contained all of
these points, but since I did not receive a Section 4 (and couldn't tell if
a Section 4 existed, because pages were unnumbered), I can't evaluate the
work presented here.

In general, when presenting Related Work (Section 5), the aim is to compare
and contrast the work in the paper with others' work in the same area. In
this paper, Related Work is used to describe invocations of the toolkit. In
other words, Section 5 is more a Deployment section than a Related Work
section. No related work is presented in the paper.

From my reading of the paper, I understand that developing a grid toolkit
is an important task. I believe that the toolkit developed by the author is
probably a useful one. However, the presentation of the toolkit in this
paper leaves much to be desired. 

F: Presentation Changes

There are several suggestions for improving the presentation in this paper:
* Fill the empty citations. Currently, they are all "[]", which is
  extremely shoddy.
* Improve the grammar. There are several split infinitives ("to better
  utilise"), run-on sentences (page 2, middle para, 2nd sent.), missing
  hyphens ("... for building customised application-specific portals..." -
  hyphen added by me), disagreement in numbers ("All other resources...
  forms the third tier"), missing commas ("e.g.," - comma added by me),
  incorrect usage ("is comprised of"), etc.
* Remove informalisms. Phrases such as "tweaking of the web pages", "at the
  heart of", "on the fly", etc. should be replaced by appropriate technical
  phrases that describe the concept better.
* Correct spelling and capitalisation errors. "coomand", "idf", "cpu", etc.
* Insert missing components. Blank sites, missing Figure 1, missing Section
  4, Related Work, etc.
* Define terms before usage. "ANT", "idf", etc.