aww ty!

There are a bunch of crypto projects with varying levels of grifty-ness doing things around this. Here’s a pretty interesting one. blog.bagel.net

yep!

the way they do it is by approximating the weighting with some number of threshold shares. so the message size over the wire will increase substantially. haven’t seen it in production, so I don’t have concrete numbers but it probably depends on share precision and number of participants.

you’re going to take a big latency hit

github.com/anoma/ferveo

for real

The linked CRDT paper kind of answers your question, which is that two nodes are never intended to be fully in sync, however the chosen tree structure enables efficient discovery of diffs for local insertion or removal. You may still have items in your tree that the other side doesn’t tho.

my impression was atproto only gave you eventual consistency. so no notion of a deterministic block of transactions.

this is great. Ty!!

only problem is this has a bunch of random other people on it

🙏

yes plz