This posting, from Guido van Rossum, is a great example of language design in action: person A proposes feature X, person B says, "No, we can do that with Y, and a whole lot more besides," and eventually a lightbult comes on.

Unfortunately, it's also an example of Parkinson's Other Law (quoted in the title of this post). Neither adapters and generic functions will win Python thousands of new users, or ensure its longevity; coming up with a credible web application framework to compete with Ruby On Rails might. At this point, I'd even be willing to settle for language design discussions that are driven by the needs of persistence frameworks, asynchronous computation, and testing...