next up previous contents
Next: Sakai SEPP Developers' Conference Up: SAKAI EVALUATION EXERCISE (A Previous: Content Aggregation: WSRP, RSS,   Contents

Portals and Portlets 2003

This section discusses the background to the international workshop held in summer 2003 to discuss these ideas and various issues concerning the deployment of portals, sharing of services and user perceptions. It focusses in particular on the impact of the JSR-168 portlet standard on portal development and the ability to share underlying tools and services. The workshop Portals and Portlets 2003 was held at the National e-Science Centre, Edinburgh, 14-17th July 2003.

The agenda and presentations from the workshop, including the majority of presentation materials, are available on line at the NeSC Web site Links to portal projects are included in the full report [28].

In spring of 2003 it was felt timely to organise an international workshop focussing on portals and portlets for e-Science. This had been discussed over the previous year among members of the GGF Grid Computing Environments Research Group. With strong interest from Rob Allan and Mark Baker in the UK, Jason Novotny and Michael Russell in Germany, Massimo Cafaro in Italy and Mary Thomas, Charles Severance and Dennis Gannon in the USA, it was eventually decided to hold a 4-day workshop in the summer of 2003 hosted at the National e-Science Institute in Edinburgh.

The workshop did indeed prove timely for several reasons. The GridLab group had recently finished the project's new GridSphere Java portlet framework and the 4th day of the workshop acted as its first major tutorial. Discussions had started between groups in the UK e-Science and JISC portals programmes with a recognition that the two groups had much to learn from each other. Indeed Chris Awre was able to bring in a number of speakers from the JISC Information Environment community. Finally, the Java JSR-168 and WSRP standards had been proposed to W3C for ratification, which actually happened in the week following the workshop.

There have been major changes in the UK e-Science programme since the workshop and even closer links have formed with JISC. The easy delivery of access to both Grid resources and information services to end users involved in multi-disciplinary research and training is more important than ever. There have recently been a number of discussions to consider how active UK groups can best collaborate to consolidate and extend best practice and functionality of existing portals. Some suggestions are provided in the workshop conclusions as follows.

It is clear that there is a lot of established expertise and momentum in the UK to develop Web-based portals for a variety of purposes. We have established strong links and potential collaborations bridging the UK, USA and other European developers and now also bridging the e-Science and JISC communities. It is important to continue this work and lead identified areas which will be taken via the Global Grid Forum research and working groups as input into the definition of standards leading to software sharing.

There are a number of UK groups already actively developing re-usable resource and Grid-based portals and portlet services. There are actually a large number of other groups developing informational, institutional, e-Learning and awareness and training portals. In many cases the frameworks being developed could be shared and the underlying portlets and services could be re-used if an appropriate architecture and standards were adopted. This implies a portlet framework plus a message-based service approach rather than a methods-based approach (advice from Geoffrey Fox, 26/2/04).

Based on our experiences and outcomes of the NeSC workshop we recommended the following:

  1. Portal services development should be recognised as a strength of the UK middleware initiatives, e.g. as noted by Fox and Walker their UK e-Science Gap Analysis [34];
  2. The active UK groups should collaborate. These include developers from the three JISC pillars: support for research, teaching and learning and the Common Information Environment plus the e-Science and Particle Physics Grid user communities;
  3. UK developers should continue to work with the GGF Grid Computing Environments research group and the American Open GCE and Sakai projects plus other relevant international fora;
  4. Developers should save effort by sharing services and methodologies and customising the existing Web-based presentation layers for delivery to all end-user projects;
  5. The UK should be active in defining, classifying and developing portal services for input into the OGSA space between (possibly changing) infrastructure and application layers;
  6. A range of toolkits (thin clients, portals, scripting languages, GUIs etc.) should be developed to extend and simplify access to Grid resources and information systems leading to the eventual emergence of one or more interfaces to a Virtual Research Environment.

It seemed that there is a good possibility of linking future activities in the areas we have described more closely with developments in other sectors, including institutional and learning and teaching arenas. There did appear to be a convergence in technology and benefits from sharing some of the tools being developed. To this end we looked to JISC, the Joint Information Systems Committee, to fund a small number of evaluation studies to examine issues of re-usability and inter working of components in the emerging portlet frameworks. Conclusions are noted in this report of the Sakai Evaluation Exercise.

next up previous contents
Next: Sakai SEPP Developers' Conference Up: SAKAI EVALUATION EXERCISE (A Previous: Content Aggregation: WSRP, RSS,   Contents
Rob Allan 2005-05-09