The Sakai project in the USA, http://www.sakaiproject.org, is creating an open-specification open-source community portal famework for contribution and sharing of portlet-based tools, principally for Collaboration and Learning Environments (new acronym, CLE).
The idea of a "Sakai Alliance" has arisen in discussion with other UK members of the e-Science and Virtual Learning communities who are interested in developing and porting tools into a portlet framework. Sakai is the leading example of such a Java framework built upon the open source uPortal hosting container. It also builds upon the JSR-168 and WSRP portlet standards and the OKI OSIDs which could be extended for a variety of purposes.
Participants in other JISC VRE funded projects who are interested in portal delivery will be invited to join the Alliance as a means of informing each other. The aim of the Sakai Alliance is to ultimately allow the UK developers to play a recognisable part in the wider Sakai development community in a similar way to that in which the UK participates in the Globus Alliance.
The Terms of Reference of the Sakai Alliance will be formulated in discussion with early participants and subjected for agreement at an open meeting.
The Sakai Alliance will host an open research group platform for UK HE based on the Sakai platform. Anyone will be able to form a group to support their research group in a similar fashion to Yahoo Groups. Initially there will be light authentication of users as the demand for the service is tested, but the Alliance will develop an implementation of a distributed trust authentication architecture such as Shibboleth and integrate it into the platform so that individual Universities accredit the researchers.
If the framework proves popular, the Alliance will explore sustainability models but the partners offer to keep the basic service running for up to a minimum of 5 years. Basic service means deployment and maintenance of Sakai production release 2.0. Again if the tool proves popular, it is likely that a number of enhancements will be requested. These may be developed by the Alliance and added to the service as part of other proposals not covered here. The Alliance will seek to QA and deploy open source extensions of Sakai on request, subject to a compatible license and available resource.
The vision is an open resource for all of UK HE inter-institutional research groups which does not depend on time-limited project finance and institutional resources for sustainability and so which enhances inter-organisational research collaboration and better preserves and disseminates the research outputs.
The primary aim of this bid is to link UK related projects contributing to a portal-based Virtual Research Environment (the Sakai Alliance) to a wider community using Sakai for Digital Information Management and Virtual Learning. It will provide facilities for the growing community of VRE portals and tools developers and users to openly share experiences and software. Portal-based facilities will be provided to aid the necessary processes. It will maintain strong links with the US portal developers SEPP members and other portal activities, e.g. for TeraGrid.
The secondary aim of this bid is to produce a dissemination programme and an ongoing evaluation of Sakai uptake and usage. Comparisons will be made with other portal-based and non portal based VRE demonstrators which are also made up of reusable tools/ services. We aim to further stimulate the collaborative development and takeup of a VRE in the UK by openly sharing standards and software components.
The Web browser is one of today's most familiar research tools. HTML pages are by far the most widely used way of presenting information and services. A enormous number of Java language tools have been developed to support the Web hosting industry funded by all major vendors and the private sector.
Every discipline already has a large range of Web-based research tools, usually with a form-based interface comprising keyword searches or pull-down menus and delivering results in the form of text which the user can cut and paste into another such interface. These are however not integrated. Only a few Web interfaces are backed up by database-enabled content management systems. They are usually referred to as ``portals'' and often include personalisation and customisation capabilities, e.g. for accessibility.
There are now several agreed standards, such as Java JSR-168 and WSRP which can respectively enable the independent development of plug-in tools for portals and content aggregation. Sakai uses the first of these plus the Open Knowledge Initiative's open service interface definitionss (OSIDs, developed with funding from the Andrew Mellon Foundation) which could be extended for a variety of purposes. For Grid-based services there is the WS-I Web services stack which has strong Java support but is language-agnostic enabling remote research services to be accessed from lightweight clients such as Web browsers and programming libraries. This is being augmented to create WSRF, the Web Services Resource Framework which is currently being debated in the Grid community.
The interest in Sakai is shared by the Virtual Learning and Digital Information communities of the JISC portfolio, and indeed the proposers have been involved in discussing commonality between the three ``pillars'' and are signed up to the Sakai Educational Partners' Programme, SEPP. There is clear interest in cementing a common and re-usable set of services, see e.g. http://www.grids.ac.uk/ETF/public/WebServices/classes.html and many of these are already addressed by OKI. Whilst the JISC circular 05/04 refers to a Virtual Research Environment, and thus focuses on tools specific to e-Research, discussions over the last few months have indicated that a ``bigger picture'' can be realised. We therefore believe that adopting a state of the art 2nd generation Web portal environment (such as Sakai) and adding a re-usable set of tools which can be customised for various end user communties is a desirable and cost-effective way of proceeding.
The project will also aim to forge close links with the OMII, the UK Open Grid Midlleware Initiative Institute based at University of Southampton. We will also work closely with e-Science Centres within the UK, e.g. the National e-Science Centre (Edinburgh) and their equivalents overseas. We will have especially strong links with the Sakai and TeraGrid OGCE portal developers and members of the Sakai Educational Partners' Programme (please see their letter of support in Appendix A). This will ensure full participation of both tools and services developers and end users. As part of the Sakai Alliance, a usage evaluation activity will grow as the requirements of e-Research grow and become more expressive. By its very nature then this project aims to assist the creation and invention of tools and services as well as assisting and guiding others with the use of the Sakai VRE interface.
Two of the developers, Rob Allan and Rob Crouchley, received JISC funding to evaluate Sakai in mid 2004. A draft report showing work undertaken so far in this evaluation is available at http://tardis.dl.ac.uk/ReDReSS/sakai.pdf. It focuses on:
As part of this evaluation, we have recently put up a simplified Sakai demonstrator at http://redress.lancs.ac.uk:8080/sakai-uPortal (username=Guest, passwd=eResearch). This will allow you to view the proposed Sakai Alliance worksite.
The proposed work packages of the project are described below. Coordination of the Sakai Alliance is the principal activity, but the others are desirable to facilitate the community process of contributing open-source, open-standard tools to a VRE framework. The project overview is as follows:
There has been a lot of interest in using the Sakai portal framework as a front end to a suite of tools and services comprising a Virtual Research Environment. We know of several independent demonstrator and tools deployment bids to JISC to enhance and work with this framework. Other developers of portal-based tools for virtual learning and digital information management may be willing to port their tools into the Sakai framework to make their services available.
We will work together with them and other projects to help achieve integration and inter-operability of independent project outcomes and to ensure common goals are met in an efficient and effective way.
We will work closely with JISC programme managers to ensure that this binding project meets the requirements of the VRE programme and meshed with JISC goals in other areas.
We will aim to ensure inter-operability and mutual benefit of integrating VRE and IE tools.
We will aim to ensure inter-operability and mutual benefit of integrating VRE and VLE tools.
We will be working closely with the Sakai developers and SEPP. We will provide feedback from UK projects and act as a point of contact for bug reports and other comments.