I came to realize that the TREE specification should be written entirely from the perspective of a client that wants to make sure it can traverse an information structure as efficiently as possible.
Servers can be implemented based on examples that we can create, and a server implementer understanding the behavior of a client.
At this moment, the spec is often written from a server’s perspective.
I came to realize that the TREE specification should be written entirely from the perspective of a client that wants to make sure it can traverse an information structure as efficiently as possible.
Servers can be implemented based on examples that we can create, and a server implementer understanding the behavior of a client.
At this moment, the spec is often written from a server’s perspective.