The Case for CASE

Abstract

Arizona State University is facing the dilemma of aging systems. Most of these "legacy" systems were written 10-15 years ago and no longer adequately support the needs of the university community. Computer Information Systems has difficulty applying state and federally-mandated requirements yet alone addressing the pressing long-term requirements of the university. Too much time is spent in "putting out fires" and patching aging systems using third generation tools. Something must change to meet the needs of the 90s. During the past 18 months, ASU has performed several pilot projects using a variety of CASE tools. The KnowledgeWare ADW toolset was used to develop and implement a University Cashiering System. The Texas Instruments IEF toolset was experienced in a "Week in Dallas" session at the T.I. Corporate Headquarters to develop an Affiliate Information Management System. The Mac-based Deft tool was used to design a data dictionary and is currently being used as an instructional device to introduce data modeling. The Bachman Designer was used to develop a physical model of our IDMS database, and reverse-engineer it into a logical model of the data. This paper will describe our selection criteria for choosing CASE tools and the pilot projects used to experience them. It will compare our experiences with CASE and explore what we have learned from it. In particular, we will focus on management issues including:Can CASE help develop higher quality systems? What are the ancillary benefits of CASE? How broad must the support base be within the organization? What are the cultural/skill changes required by technical staff? What are the true costs of introducing CASE How does CASE fit in the quest for distributed computing and client-server applications?