🚀 go-pugleaf

I am a little confused on the usage of ports in abstract protocol
specifications.

Why is the communication between two agents split into different
ports? What is the advantage? Is this purely an abstract notation,
or does it mean that if a UA wants to connect to more than one
port on a Service Agent, then multiple connections have to
be opened. If the latter, why is this an advantage?

One advantage with ports might be do make access control easier,
if only some UA-s can access some ports which provide some kind
of privileged access. But if this is the reason for organizaing
the protocol into ports, then why has X.500 different ports
for search and retrieval?

Can the same operation occur via more than one port?