Basically you want to be able to return a diff.<p>I half remember that this is possible over http for updating a resource (e.g., you looked at example.com/a-page an hour ago and now you want to look at it again; your browser can just download the bits that have changed). But I don't remember the details. Maybe it's part of WebDAV?<p>Then you only need to hint to the browser that a certain link might be diffable against the current page (or another page that the browser might have cached). That means only one new attribute, one new http header, and no new url structure required.<p>Adding new mandatory structure to urls icks me out for some reason. It doesn't feel very web-like.<p>Edit: I think that something like your idea would be helpful, but I don't think your idea would be the best way of doing it.
I've been playing around with this idea for a while, and given this week's context, I figured I might as well throw it out there. My main questions for HN are: (a) is this a good idea, (b) do you see any big pitfalls I'm overlooking, and (c) how might we get this to happen?