Talk:GridServer/ClientDocs

Why should a client need to first get "metadata" about an asset request and then get the real data? So the asset information is distributed. Why should the client have to be involved directly in what seems to be a matter of chasing proxies to the real information? Why can't real proxies that mucking with finding where things really are or the inner keys or whatever be done transparently to the client regardless of how distributed the information is or may be?