What we talk about when we're talking about "Webmentions"

Sometimes I think that the way people talk about webmentions (even me) is far too limited. The protocol itself is incredibly flexible. Why should it be that we limit ourselves to reinventing

  1. Comments sections, or
  2. Twitter?

Especially when those two things can be more elegantly done (even in a distributed manner) with less flexible solutions.

What are other fun/weird things you could do with webmentions?

His point on Javascript I'm not entirely in agreement with. For one, webmention.io is fully self-hostable so having a static site with webmention content provided at runtime isn't at all bound to someone's ability to provide a service for free (I use webmentiond for this). For another, this sounds like something a caching layer was designed to solve if anyone's site hits Scale. Add to this that runtime fetching of other people's PII is arguably superior, as this gets into a little. And finally, I don't think that displayed webmentions are something that necessarily should be machine-parsed. My content is marked up with mf2 because I want to make it easy for people to parse and consume my content. I don't own the content I'm displaying as replies; even if there's an etiquette about excerpting for comments-section-type-usage, why should I be exposing it for even-more-indirect consumption?