RIP (Routing Information Protocol) is a true vintage class among the routing protocols that are still in common use. Its ancestry goes back to the ARPANET routing algorithm that was designed in 1969 [MFR78]. In the mid-1970s, a similar protocol was implemented for the Xerox PARC Universal Protocol (PUP) under the name Gateway Information Protocol (GWINFO). Later, GWINFO was included in the Xerox Network System (XNS), but its name was changed to Routing Information Protocol. However, the most important milestone in the RIP history was the release of the 4.2BSD Unix in August 1983. This operating system included the routed daemon, essentially a generalisation of the Xerox RIP capable of handling IP addresses and networks. The routed daemon then become a de facto standard for intra-AS routing in the emerging Internet. It was not until 1988 that RIP became a proposed standard in IETF [RFC1058].

The second version of RIP [RFC1723] added support for classless IPv4 addresses (CIDR) along with other significant improvements. The strongest point of the RIP protocol is its simplicity, which makes it easy to implement and configure. As a result, interoperable implementations have long been available for virtually all IPv6 routing platforms and so RIPng used to be a popular choice, especially in mixed environments. Although this niche is gradually being taken over by the more advanced intra- AS routing protocols (IS-IS and OSPFv3), RIPng does still have some appeal, especially for small networks.

On the other hand, the most serious weakness of RIP, and one that is unlikely to go away, is its hard limit on path length: only 16! RIP considers destinations that are 16 or more hops away to be unreachable. This may seem very restrictive but there are good reasons for keeping the maximum allowable path length low.