Metrics%20

1
Metrics Verifying Software

• Paul E. Black
• National Institute of Standards and Technology
• http//www.nist.gov/
• paul.black_at_nist.gov

2
Candidate Software

• Secure SCADA Reference Implementation
• Supervisory Control And Data Acquisition systems
  measure and control infrastructure for
  electricity, oil, gas, water, etc.
• Rigorously specified and thoroughly tested
• Medical Infusion Pump
• Many models from many universities, etc.
• An archetype of FDA approval process
• Open source software Apache, Linux, etc.
• Change revision history available

3
Reminders

• Be willing to set aside your favorite method,
  tool, approach,
• To a man with a hammer, every problem looks like
  a nail.
• Parable of the street light
• This will take lots of good, hard work
• Generally applicable tools take 3x longer
• Integrated tools take 3x longer again
• Brooks, Mythical Man Month
• Little of this is considered publishable

4
Useful Metrics

• A measure is an objective attribute
• A metric is higher-level or more subjective
• Measures are bases for metrics faults/LOC
  approximates quality
• Careful version tracking is essential
• Reports (problems or certificates) should be
  traceable to particular entities (specifications,
  version, module, )

5
Society has 3 options

• Learn how to make software that works
• Limit size or authority of software
• Accept failing software