Interesting information, but it doesn't support the assertion made by parent that experimentation has (often, it seems implied) been the root cause of poor bridge design implementation.