I'm not entirely sold. Application data... Let's say there's a Great Open Database of microblogging. This is somehow distributed across multiple systems, and

I'm not entirely sold. Application data... Let's say there's a Great Open Database of microblogging. This is somehow distributed across multiple systems, and sized so that it can handle the load, and provides a query language to make that database useful... Shoot, let's spin up a couple clusters of CouchDB and go from there and build a microblogging platform on top of that? Does it solve the problem? It's trivial for anyone to spin up their own clone of a CouchDB - just replicate it. You want your own uncenersoable copy - its yours. And yes, there are numerous problems with this that as you pull the thread get to the point of "at some point, you're going to have to trust some central authority" That might be where you're getting to the "PGP for the modern world" - though I'm going to contend that every solution has a similar problem of getting the average person to treat their keys securely... and that means having one copy and getting ransomeware or having key stealing malware has become much, much more problematic than it is in today's world. I believe that that social / computer literacy problem needs to be solved for the average person before one can reasonably move forward with PGP for the modern world.