Abstract
Software engineering attempts to produce systems that are ??good systems? in terms of reliability, ease of maintenance etc. We take a broader definition of a good system as any general system that produces benefits that exceed initial expectations or intended scope or initial investment. There appear to be common characteristics that tie together such systems. These are hypothesized to include functional ??goodness?, good infrastructure, reliability, connect-ability, versatility and benefits that overflow/overwhelm the system??s scope or initial investment. A case study approach involving four examples of what are regarded as ??good systems? and four examples of what are regarded as ??bad systems? fully supports this hypothesis. But support for the converse hypothesis, a bad system not having these characteristics was only 68.7%. The implications of these findings are discussed.
Keywords