Referee 1 ******************************************* D: Referee Comments (For Editor Only) The paper is technically sound, but also readable by people not very proficient in the field of grid computing. It is definitely relevant to this special issue. The abstract, introduction, figures and bibliography are adequate, and the paper organization is good. E: Referee Comments (For Author and Editor) The paper describes an application specific grid portal built using the Globus grid infrastructure as middleware. The portal exploits the majority of the Globus system capabilities with the exception of the new Replica Management mechanism, and provides an easy to use interface to astrophysics scientists. The implementation relies on DHTML, java servlets and Java Server Pages coupled with commodity grid toolkits for accessing Globus functionalities. The approach is clean and promotes both rapid prototyping and reuse. Fundamental capabilities for a grid portal like job submission, monitoring and data transfer are provided, however the presentation does not cover sufficiently how the scheduling is done (user driven or automatic?), nor how the nformation services are exploited. The Grid CVS protocol that has been developed is, in my opinion, really valuable. F: Presentation Changes The paper is written in a clear, concise style; it is easy to read and well organized. Section 3.1, last sentence of first paragraph: ...and a module performs... should be: ...and a module that performs... A section with details on scheduling and access to Information Servers should be added. Referee 2 ******************************************* E: Referee Comments (For Author and Editor) ------------------------------ The paper describes a collaboratory based on the well known Cactus approach. The paper is generally well structured. More attention should be given to comparable work in the field which should also be referenced in order to show where the ASC stands compared to other projects. In the present form the paper is not yet ready to be published because of lack of proof reading and formatting. Some figures should be of better quality - especially figure 6. F: Presentation Changes The definition of Grids as given on page 2 is somewhat misleading since it ignores hardware resources like computers and devices. In that sense it is somehow contradicting the definition given by Foster and Kesselman. This should either be changed or explained to the reader. The contribution of figure 1 to the paper is weak. It neither clarifies a concept nor provides additional information that is not given in the text. The authors should either improve or remove it. The figure caption of figure 2 talks about an "ASC Appl. Server" which is not seen in the figure itself. No whatsoever reference is given to figure 3. It should be described in the text. Page 9, paragraph 4 "and a module ... performs". It seems that a "which" is missing. figure 5: the role of the toolkit is not sufficiently explained. It would also be helpful to see the role of the Grid/Fabric in this model/terminology page 15: As the version informations of Web browser tend to be outdated very soon, they should be replaced by a description of pre-requisites. E.g "Netscape 4.x.+" should be replaced by "webrowsers supporting Java and JavaScript". The authors could mention that the software is tested with specific versions of the browsers, if their experience shows that this is necessary. page 15: the authors state that "a remote site is brought into the collaboratory by simply implementing the Grid protocols...". Probably the term "providing Grid services" is better suited in this context than "implementing grid protocols". page 15: the authors state that the "Cactus framework ... permits a diverse group... to develop software modules.. that integrate additional physics or numerical solvers into the Cactus framework". It would be very interesting for the general reader to get more information whether and where this has been done outside the field of numerical relativity. page 16: figure 6 is especially bad in our copy of the paper. This should be considered and improved. page 16: paragraph 2 "These will also the support of using Java3D ..." does not have any meaning. Referee 3 ******************************************* E: Referee Comments (For Author and Editor) ------------------------------ The paper gives a good overview. I recommend to address also the following: 1. Is there a single ASC Server for the whole collaboration? Would different collaborations use independent servers? 2. Management and administration: 4.2 suggests ASC collaborators have to be registered at a central server. How is this done? How are the resources of the Grid made available to the collaborators, i.e. how are they authorized at the sites providing the resources? 3. A list of participating centers would be helpful.