>> [Comp Sci Dept]
>> [R J Botting]
>> [New Bibliographic Items]
|| [Site Search]
|| [Bibliography Search]
Wed Sep 14 14:51:25 PDT 2005
- Harry Goldstein
- Who killed the Virtual Case File
- IEEE Spectrum Magazine (Sep 2005)pp24-35
- =HISTORY VCF Trilogy FBI DISASTER FAILURE RISKS
- Demonstrates that bad management and bad process can lead to a lot of money being spent for nothing.
- Changing goals in midstream
- Tightening the schedule
- No process or control
- No "system architecture"
- Changing bosses in midstream.
- Muzzling and ignoring critics.
- Giving very detailed requirements (position and color of buttons) but no vision or goals.
- Changed requirements flood in after first demo.
- Cost plus contract.
- NIH: no COTS bought in
- It may be cheaper to spend time writing good requirements using math and prototypes rather than hiring lawyers to wrangle over the ambiguities.
- Philip E Ross
- The Exterminators
- IEEE Spectrum Magazine (Sep 2005)pp36-
- =ADVERT Praxis MATH Z Spark PROTOTYPES V&V TOOLS TESTING MONDEX
- Company demonstrates that by getting very clear and reliable requirements expressed mathematically and using a reliable language to write the code you can reduce the number of faults to 1 for 25KLoC on project with 200KLoC.
- Robert N Charette
- Why Software Fails
- IEEE Spectrum Magazine (Sep 2005)pp-
- =ADVERT RISK MANAGEMENT DISASTERS COSTS
- p44. Hall of Shame 1992..2005
- Side boxes describe case studies
- Rounds up the usual
- Bad goals
- Bad estimates of needed resources
- Bad systems Requirement
- Poor project control and/or planning
- Bad communication
- Commercial pressures
- Susan Hassler
- Learning from Software Failure
- IEEE Spectrum Magazine (Sep 2005)p8
- =EDITORIAL COSTS RISKS FAILURE DISASTERS VCF
- Good software can transform a business: Dell Walmart.
- Notes that a future disaster may be when the USA digitizes and automates medical records.