I found the post well written and informative. Though I am clueless about OCaml it feels as this would be useful for anybody working on a new server implementation in any language ecosystem as it highlights what needs to be done and potential bottlenecks.
As for the activitypub spec and the currently popular implementations it doesnt take long exposure to the fediverse to realise there are some rough edges and historical accidents (e.g mastodon being actually the defacto interpretation of the standard). Imho now that there is substantial more mindshare devoted to decentralized social it would be opportune to revisit these things and if needed revise before they get backed in.
Comments
I found the post well written and informative. Though I am clueless about OCaml it feels as this would be useful for anybody working on a new server implementation in any language ecosystem as it highlights what needs to be done and potential bottlenecks.
As for the activitypub spec and the currently popular implementations it doesnt take long exposure to the fediverse to realise there are some rough edges and historical accidents (e.g mastodon being actually the defacto interpretation of the standard). Imho now that there is substantial more mindshare devoted to decentralized social it would be opportune to revisit these things and if needed revise before they get backed in.