It sucks reading all the Mastodon vs Bluesky takes because the whole *point* of decentralized social media was that it didn’t have to be a zero-sum game. Yes Mastodon is more nerdy and yes Bluesky is more “normal”, but if they all used ActivityPub then everyone could win!
But instead of contributing to the open social web, Bluesky decided to fork the ecosystem with their own proprietary protocol :/ And now we’re headed for the same siloed world as before
@rileytestut There is a difference with the two:
ActivityPub is built as 1-to-1, i.e. content goes to your followers primarily via follower lists. This is a reason why some are hostile to search here because connections are point-to-piont.
Bluesky's protocol is one-to-many and everything is public. It prioritizes sending content to a big-data-gathering indexer where everything is searchable.
This simple difference makes a huge change in how they can be implemented and scaled.
@Ciantic I understand they have different strengths, but proprietary protocols always have strengths over their open equivalents for certain use cases; it’s whether those differences are worth splitting the ecosystem vs slowly improving the standards (e.g. improving HTML5 to support games)
I personally think if Bluesky was conceived today, they’d just adopt ActivityPub; instead we’re seeing the sunk-cost fallacy play out because they already had been working on AT Protocol (and it may win)
@rileytestut I agree that everything Bluesky wants could have been done with little extensions or improvements on ActivityPub: Consent to search, index, sign messages, etc. Even the handle formats could have been redecided because they are not part of ActivityPub.
Currently AP doesn't require you to sign messages (only HTTP header signatures are required). In Bluesky it's a requirement because you have to trust middle servers like BGS to provide the messages on behalf of the original server.
Riley Testut 
Jari Pennanen