one truck can hold many shipments (assuming that the shipments are small enough to fit in one truck or less). An occurrence of this network is shown in Figure H-26.
To represent this simple network with linked lists, we need to establish one set of pointers for each 1:N relationship. In this example, that means one set of pointers to connect CUSTOMERs with their SHIP-MENTs and another set of pointers to connect TRUCKs with their SHIPMENTs. Thus, a CUSTOMER record will contain one pointer (to the first SHIPMENT it owns); a TRUCK record will contain one pointer (to the first SHIPMENT it owns); and a SHIPMENT record will have two pointers, one for the next SHIPMENT owned by the same CUSTOMER and one for the next SHIPMENT owned by the same TRUCK. This scheme is illustrated in Figure H-27.
A simple network has at least two 1:N relationships, each of which can be represented using an index, as we explained in our discussion of trees. For example, consider the simple network shown in Figure H-25. It has two 1:N relationships, one between TRUCK and SHIPMENT and one between CUSTOMER and SHIPMENT. We can store each of these relationships in an index. Figure H-28 shows the two indexes