Makes sense. I'm just going through that spreadsheet still that no one seems to be updating. I think we need to start thinking about indexing and metadata and all that. And whatever happened to that P2P filesystem direction? Sadly, I don't have that much time at the moment to propose something. But at a minimum I might spend time setting up indexing of the URLs we've got. So in case like this, it's a simple centralized lookup that tells us exactly what copies of a given URL we have.
Indexing sounds good, Borislav. Not only of URLs, but if we could have a master index of files, by server, and path, that could be useful. Why not write a ticket to that effect, and assign it to yourself?