Too many small files, papers need different indexing.

Shadow libraries (like scihub, Anna’s archive etc.) are the way to go - as long as scientists don’t or can’t publish Open Access which is what needs to happen.