Where Does Scientific Data Go to Die?

That's a very interesting idea, and it would be relatively easy for them to implement. It does help solve the problem of making the data available over time, which is a big step. But unless they built a canonical data format (i.e. a proper database) underneath it, you still end up with difficulty making use of the data.

Neal (author)

Someething like a github repo for every arXiv submission - shouldn't be that hard to do.