Because it is on the idea stage without funding. My guess is that it will stay that way. That said, it is probably feasible to do some parts of it, similar to

Because it is on the idea stage without funding. My guess is that it will stay that way. That said, it is probably feasible to do some parts of it, similar to the 3x3 project Ruby has. I wouldn't be surprised if numpy glue is a lot more tractable for optimization than Rails, but the interaction with C libs would pose some serious issues.