1.7 How does the framework provide value for money?
The core of the framework has a small set of deliverables:
  • A list of services identified within the framework, and for each of these
    • The definition of the scope and purpose of the service
    • List of applicable standards and specifications
    • Information on open-source toolkits developed by the programme
    • List of prior work of value, such as projects and case studies, or demonstration projects using the service
    • Information on shared services offered on an ASP basis, for example relevant JISC IE services
  • General guidance on creating, exposing, and consuming services
Note that the current version of the framework only contains:
  • A list of services identified within the framework, and for each of these
    • The definition of the scope and purpose of the service
    • List of applicable standards and specifications
Also, note that we are interested in producing toolkits for the framework rather than complete solutions. This is because we want to encourage a diverse range of software for institutions to choose from, both open-source and commercial.
Therefore, development activity for the framework will be primarily concentrated on providing libraries that enable developers (again, from commercial as well as open- source communities) to create applications that provide and consume services defined within the framework.
Demonstrator projects will be aimed at testing out the service definitions and toolkits in actual use.
By developing these libraries and toolkits and providing them free-of-charge we substantially reduce the cost of creating software based on services defined in the framework, and reduce the cost of entry to a service-oriented architecture for UK institutions.
We also don’t want to build large pieces of software such as content management systems, as these are very complex pieces of software (and there are plenty of commercial and open-source solutions available).
It may be the case that there are some parts of the framework for which there is demand from the JISC community but no available solutions from vendors or the open-source community, in which case it may be necessary to build an open-source reference implementation to fill the gap. This is not an ideal approach, as it tends to lead to problems of sustainability.
Overall, then, the strategy is to create small, tightly scoped pieces of development work that will have maximum impact on the sector. There will be no “white elephants”.