Editing Classful network (section)

==Introduction of address classes==
Expansion of the network had to ensure compatibility with the existing address space and the IPv4 packet structure, and avoid the renumbering of the existing networks. The solution was to expand the definition of the network number field to include more bits, allowing more networks to be designated, each potentially having fewer hosts. Since all existing network numbers at the time were smaller than 64, they had only used the 6 least-significant bits of the network number field. Thus it was possible to use the most-significant bits of an address to introduce a set of address classes while preserving the existing network numbers in the first of these classes.

The new addressing architecture was introduced by {{IETF RFC|791}} in 1981 as a part of the specification of the Internet Protocol.{{Ref RFC|791}} It divided the address space into primarily three address formats, henceforth called address ''classes'', and left a fourth range reserved to be defined later.

The first class, designated as ''Class A'', contained all addresses in which the most significant bit is zero. The network number for this class is given by the next 7 bits, therefore accommodating 128 networks in total, including the zero network, and including the IP networks already allocated. A ''Class B'' network was a network in which all addresses had the two most-significant bits set to 1 and 0 respectively. For these networks, the network address was given by the next 14 bits of the address, thus leaving 16 bits for numbering host on the network for a total of {{gaps|65|536}} addresses per network. ''Class C'' was defined with the 3 high-order bits set to 1, 1, and 0, and designating the next 21 bits to number the networks, leaving each network with 256 local addresses.

The leading bit sequence ''111'' designated an at-the-time unspecified addressing mode ("''escape to extended addressing mode''"),{{Ref RFC|791|repeat=yes}} which was later subdivided as Class D (''1110'') for multicast addressing, while leaving as reserved for future use the ''1111'' block designated as Class E.{{Ref RFC|988}}

This architecture change extended the addressing capacity of the Internet but did not prevent [[IP address exhaustion]]. The problem was that many sites needed larger address blocks than a Class C network provided, and therefore they received a Class B block, which was in most cases much larger than required. Due to the rapid growth of the Internet, the pool of unassigned Class B addresses (2<sup>14</sup>, or about 16,000) was rapidly being depleted. Starting in 1993, classful networking was replaced by [[Classless Inter-Domain Routing]] (CIDR),{{Ref RFC|1518}}{{Ref RFC|1519}} in an attempt to solve this problem.