Empirical evidence for software metrics
Someone did the leg work and found some real evidence about software metrics:
http://research.microsoft.com/en-us/news/features/nagappan-100609.aspx
The results are interesting, but not surprising if you've developed software for a while:
- "Code coverage is not indicative of usage" - Your hot spots need good code coverage, but not your cool spots.
- TDD projects take 35% more time to complete, but are 60 - 90% better in terms of "defect density".
- Confirmed that having more assertions reduces defects. However, you can't just mandate their use. Developers must have a culture of using them.
- Organisational structure matters a lot (herding your cats)
- Geographical distance doesn't matter much if your distributed team feels like they're working on the same team. Here's an interesting quote - "Most people preferred to talk to someone from their own organization 4,000 miles away rather than someone only five doors down the hall but from a different organization."
