Page 149 -
P. 149

134     HENDERSON-SELLERS
                    SUMMARY


                    To date, the evolution of AO methodologies has been disparate with many groups worldwide
                    creating individual offerings. These vary in style and, particularly, in heritage and have a specific
                    focus, either in terms of domain, application style, or life-cycle coverage. For industry adoption, it
                    is essential that full life-cycle coverage is achieved in a “standardized” way. One way of achieving
                    some degree of standardization yet maintaining full flexibility is through the use of situational
                    method engineering, underpinned by an agreed standard metamodel. With this approach, method
                    fragments are created and stored in a repository or methodbase. For an individual application,
                    only a subset of these is then selected from the repository and a project-specific (or sometimes
                    organization-specific) methodology is constructed. Here, we have demonstrated how this might
                    work by using the OPEN approach, which already provides a significant coverage of AO method
                    fragments as well as more traditional OO and pre-OO fragments.
                      The advantages of an SME approach to methodology provision lie in the flexibility and com-
                    plete tailorability of the construction process. Only fragments suitable to the organization’s culture
                    and capability (e.g., in terms of ISO/IEC 15504 or CMM/CMMI), the skills of its developers,
                    the organizational culture, and leadership style (e.g., Constantine and Lockwood, 1994) are used.
                    Construction costs are commensurate with the costs of attempting to mold an “off-the shelf, one-
                    size-fits-all” methodology to a specific situation. Starting with an SME approach with a large
                    extant repository (such as that provided in the OPF) is clearly advantageous. Empirical studies,
                    while undertaken only in OO and not AO situations (e.g., Henderson-Sellers and Serour, 2005),
                    have shown that organizations can readily benefit from this SME approach in practice.

                    ACKNOWLEDGMENTS

                    I wish to thank Dr. Cesar Gonzalez-Perez for his useful comments on earlier drafts of this manu-
                    script. This is Contribution number 06/03 of the Centre for Object Technology Applications and
                    Research.

                    NOTES

                      1. For the present chapter, the words method and methodology will be used synonymously (e.g., Jayaratna,
                    1994).
                      2. A more recent manuscript in preparation does, in fact, acknowledge influences from object technology
                    (see also later discussion).
                      3. In the near future, this traditional, process-focused metamodel will be supplanted by the forthcoming
                    ISO/IEC 24744 standard: Software Engineering Metamodel for Development Methodologies. The new
                    metamodel supports method enactment as well as method construction and brings together process-focused
                    components of a methodology and product-focused components.
                    REFERENCES

                    Bernon, C.; Camps.,V.; Gleizes, M.P.; and Picard, G. 2005. Engineering adaptive multi-agent systems: the
                      ADELFE methodology. In B. Henderson-Sellers and P. Giorgini (eds.), Agent-Oriented Methodologies.
                      Hershey, PA: Idea Group, 172–202.
                    Bernon, C.; Gleizes, M.-P.; Picard, G.; and Glize, P. 2002. The ADELFE methodology for an intranet system
                      design. In P. Giorgini, Y. Lespérance, G. Wagner, and E. Yu (eds.), Agent-Oriented Information Systems
                      2002. Proceedings of the Fourth International Bi-Conference Workshop on Agent-Oriented Information
                      Systems, Toronto, 1–15.
   144   145   146   147   148   149   150   151   152   153   154