Maybe this is an opportune time to bring up a related concern. It's crucial that any system for disseminating scientific work be reliable and persist indefinitely. I don't like the current system, but I wonder how this will be achieved otherwise.
I agree. I don't want papers to dissapear indefinitely due to an IT mistake, and I'm sure there are many researchers that don't save previous papers locally. That, and being able to access cited works is also important.
