I'm drawn to a lot of points in this post - Opinionated Wiki.
I'm not particularly fluent in wiki. I've always got the functional aspects and that they're pretty neat, but some of the conventions surrounding their use have always baffled me.
The link issue has always annoyed me - not wanting inline links to external resources - because it created a false divide between wikis and the web. Wikis are of the web but the web couldn't be part of the wiki? I never understood the logic.
One of my bug bears with the wiki is the concept of published. Coming from a background in design published was a temporal marker in the content lifecycle . Things led up to that point and things happened after.... but there was always a point where it was published. At that moment in time it was complete. It might be revised, edited, fixed or destroyed but at one point it existed as an artefact on it's own. The nature of the wiki destroys that point... and fedwiki maybe more so with it's forkability. Not only is there no definitive published artefact but there may be multiple versions of it. I'm not 100% sure what the fix of this is. After being part of the OERu tech team I can see the value of a wiki in the planning, development, management and deployment stages because it's one of the few technologies that can work with all those functions. I just wonder how it's possible to perhaps delineate some of those processes and functions. Master copy - as in Git might be an idea.