> It looks like (without digging through code) it's so they can parse the file-structure and generate the URIs to forward requests. I haven't looked at the cod

> It looks like (without digging through code) it's so they can parse the file-structure and generate the URIs to forward requests. I haven't looked at the code either. But the hosting seems to be handled in JavaScript. There are no required clients (though there does seem to be an optional one); this is all completed in the browser. So you're not "uploading" content, you're providing the content to the client-side application so it can be shared through that same client-side solution.