40 Years Of RFCs
The first "Request For Comments" was released 40 years ago on the 7. April, 1969. These written discussions of the "Network Working Group" became the basis for the open internet standards.
40 years ago on the 7. April 1969, the first Request For Comments was released. In RFC 1, Steve Crocker described the software architecture of the emerging ARPANET. The original RFCs were actual contributions to discussions within a "Network Working Group" of manageable size (RFC 2 is a direct answer to RFC 1). As the group was still working on the ARPANET setup, the first RFCs were exchanged on paper by post.
Over time, the RFCs became the standard documents of the ARPANET and of its later successor, the internet. The network's growth was reflected in the number of RFCs released. After a phase of intensive discussion during the development of the ARPANET, there was a period of calm in which the network itself required less attention. In the late 1970s, however, there was a technological surge, and since the introduction of TCP/IP, the network and RFCs have developed at the same rate. The 2,555th RFC, for instance, discussed the first 30 years of RFC history. Almost 3,000 texts have been added in the ten years since.
The early phase of intensive discussions was followed by a quiet period of several years, until the expansion of the internet led to a sharp increase in the number of RFCs
It is not only the number of RFCs that has increased, but also the length of the individual standards. The definition of the Post Office Protocol (POP3) from 1985, for instance, comprises a lean 24 pages any reasonably talented programmer can quite rapidly implement in servers and clients. In comparison the admittedly more powerful IMAP (Internet Message Access Protocol) has been the subject of 183 RFCs with numerous pages, dating back to 1994. This is not only due to the growing performance of modern protocols and the internet's transformation from a research project to a global communications network. Political and economic interests have also become increasingly important factors in technological decisions. This becomes obvious, for example, from the growing average number of authors per RFC, and from the fact that the majority of authors no longer work at universities, but in commercial enterprises.
RFCs have increased not only in number but also in length.
Incidentally, more text doesn't always mean more clarity. The flame wars about protocol details between the programmers of IMAP servers and the RFC group are legendary.
Jon Postel, the legend
Another legend is Jon Postel, who was an RFC contributor from 1970 and held the post of rfc editor until he died in 1998. Vint Cerf – another internet oldtimer – of course released his obituary in form of an RFC.
Curios
Such non-technical RFCs still continue to appear. The most well-known of them are probably the April fool's day RFCs which have been released regularly for the past 20 years. These texts are often satirical takes on current technological developments, for example the BLOAT protocol of 2002, which tries to replace IP addresses as well as UDP and TCP ports with XML. Surprisingly, there was no April fool's day RFC in 2006 of all years, when a particularly high number of RFCs were released. Could it be the RFC editors were so busy they forgot to release an April fool's prank? The April 2006 list of RFCs does not appear to contain an obvious candidate.
Another curio is RFC 31, which appears to be about a year older than RFC 1, having allegedly been released in February 1968. Both the data in RFC 30 and RFC 32 and the RFC's content, however, rather indicate a release in 1970. 1968 was probably a number scramble caused in 1998, when Dave Bachmann edited the online version.
The shortest text released is the now inconsequential RFC 18, the longest is 4949, which is a glossary of security terms. Those who are interested in this type of statistics can find further online samples on the pages of RFC author Jari Arkko.
Open standards
From the beginning, RFCs were an open platform for discussions. RFC 3 already dealt with the form and distribution of the notes and explicitly states that the Network Working Group is open to new members. Once the network was able to offer the required services, the next logical step was to collect the RFCs online. The ARPANET (and its successor, the internet) was, therefore, based on open standards right from the start. The initially rather coincidental RFC concept was one of the most important factors that contributed to the success of what is now known as the internet. Other sectors of the emerging IT industry tended to regard their developments as inventions, protected them with patents and only released them at a much later stage.
RFC's are available on The H in a searchable form as part of The H's Internet Toolkit.
Next: Over 30 Years of April RFCs
- Share this article
