If you know git: a federated wiki works like that. To make a federated wiki work like a traditional wiki, you can think of there as being a "canonical copy" which you can clone, make edits, and then use a 'pull request' process for incorporating those edits back into the original. But you also don't need to have a single source of truth: you can, for example, have a group of people—say, students studying for a class—who are building their own wikis, and among each other they can copy in pages, make changes, copy changes back, and so forth, all building a web of things. In the same way that git can replicate a subversion-like workflow but also introduces the possibility of different workflows, a federated wiki can replicate a traditional wiki but also has a number of workflows it can accomplish.
Thanks! I'm aware of the abstract technical idea, but is there any evidence that Ward's wiki ever worked like that? Did he ever accept a "pull request" or are there any notable forks which do?
