Disposable information systems: the future of software maintenance?

✍ Scribed by Voas, Jeffrey M.

• Publisher: John Wiley and Sons
• Year: 1999
• Tongue: English
• Weight: 59 KB
• Volume: 11
• Category: Article
• ISSN: 1040-550X
• DOI: 10.1002/(sici)1096-908x(199903/04)11:2<143::aid-smr186>3.0.co;2-v

✦ Synopsis

News from the Information Technology (IT) community is mixed. Project backlogs, high employment and personnel shortages are claimed on one side; project cancellations, unproductive use of personnel time and continued poor quality software are claimed on the other. Which side is component-based software on? The drive to use components such as COTS components comes from claims of instant productivity gains, shorter time to market, lower cost, management mandates and a 'parts' philosophy. On the other side are difficulties in composing systems from unreliable components, the 'ility' factors, the traditional white-box software processes, short-changed testing, software component support limits and software malleability. If components such as COTS are used a lot, then finger-pointing grows more common as we move toward black-box component integration. Some of the defensive options include component certification, strengthened testing, and adding wrappers and middleware. All add to maintenance cost. If we use components such as COTS, our ability to maintain the resulting composite systems is reduced to the level of 'unmaintainable'. This leads us to think in terms of disposable information systems. Overall, I observe as we move toward component-based software, quality will be even more important for maintainability.