This paper identifies and discusses the persistent problems and development practices of information systems development (ISD). A critical examination and comparison of past times ' 'traditional' and present-day 'web-based' development shows that contemporary ISD can be seen as an accentuated evolutionrather than a revolution -of well-known challenges and solutions. On this basis, (1) diversity; (2) knowledge; and (3) structure are identified as inherent and interrelated problems, while the practices for coping with these three challenges are (a) organization and specialization; (b) constant verbal communication and negotiation; and (c) pragmatic application of certain development methods and methodical concepts. We conclude that more research on the occurrence and interaction of problems and practices at, and between, different contextual levels (e.g. the business environment, company, project, team and individual levels) is needed to understand and assess (the gap between) 'observed practice' and 'good practice' across the many types of Web and non-Web ISD projects conducted today. We outline a possible research agenda to investigate these issues.
Keywords: information systems development, empirical studies of ISD practice
I N T R O D U C T I O NInformation and communication technologies are rapidly evolving. Hardware processing and storage capacities have been increasing at tremendous rates and computer-based devices and applications are spreading into nearly every human life sphere. The tools, techniques and processes used for producing information systems (IS) are undergoing profound changes as well. High-level languages and program generators have replaced machine languages. Similarly, ad hoc approaches to information systems development (ISD) have been replaced by development methods supported by powerful CASE tools, and new systems are now rarely built from scratch, but based on standard software packages or assembled from module libraries.Unsurprisingly, these changes are accompanied by frequent claims about their fundamental and paradigmatic nature. Already in the mid-1960s, early third-generation languages such as K Kautz et al.