Paragraph Level Atomicity

Presently a wiki page is constructed out of individual items (roughly speaking numbered paragraphs), each with an id, and a plugin that renders the json for display.

# For now

For now we can leverage this structure using IPFS to create a Wiki Page Package - that is a folder that contains an indexed list of multihash entries to individual paragraph items in IPFS.

As long as we have some sort of versioning for the package, so that it can refer to code that would reconstruct the original json properly, and a way of updating which wiki page package is used - we should be good to go.

This feels like writing a Fuse file system specifically for wiki.

Software would then reconstitute the wiki page json from the individual items. This could be the server or a virtual file system of some kind. Perhaps we could use Vinyl or an extended version of Vinyl, or a package.json based format.