Answer 15.2 through 15.11 in the book. Then answer the following
questions as well.

1. Determine if the end-to-end argument would apply to
the following proposals to add "enhanced functionality" into the
network, and describe how the end-to-end argument would 
dictate that this functionality be implemented.

        a. A guaranteed delivery service that will store and
        forward all packets within the network until they are
        delivered to the destination host.

        b. An in-order delivery service which ensures that 
        packets are never reordered within the network.

        c. A multicast service, where packets are sent to a
        group of hosts identified by a group id, on a best effort
        basis.

2. Why does the Ethernet impose a maximum packet size ?

3. Why does the Ethernet impose a minimum packet size ?

4. Is the CSMA/CD algorithm used in Ethernets fair ?

5. A network connection typically requires at least three
different kinds of names: a hostname, which is translated
to an IP address, which is in turn translated into an
Ethernet address. In contrast, we use a single telephone number
to set up a telephone connection. Describe the advantages
of using DNS and ARP to map textual names to IP addresses
and to Ethernet addresses over flat addressing as in the
telephone system.

6. What properties does TCP provide over UDP ? Why and when 
would you use UDP over TCP ?

7. What are the properties of a TCP-friendly transport protocol ?

8. DNS provides a hierarchical naming scheme with a single root.
Name one advantage and one disadvantage of rooting all names
in the same common location.