You’re right, I was confusing Sequence data type with LSeq CRDT. So my question stands, but I’m asking it about the SAFE data types Sequence and FileTree rather than the underlying CRDTs.
My understanding is the the Sequence data type is loaded/saved as a single blob, and that each replica has a complete copy. I could be wrong about that, I tried to clarify How SAFE Sequence CRDT Data Type is implemented, so this is the basis of my understanding and from there, I thought you planned to implement FileTree in a similar way at least to start? I don’t have a picture of how the TreeCRDT fits into this, so maybe I just need to understand your plans for implementation (e.g. block diagram stuff) better.