The Internet engineering community has proposed a new communications protocol designed to help prosecutors track down people who illegally download copyrighted material from peer-to-peer Web sites.

The so-called Omniscience Protocol would be installed on all Internet-enabled devices that can be used to play protected material, including computers and MP3 players.

The protocol will work even when a user's Internet connection fails. "Since the evil-doer might try to hide his or her evil-doing by disconnecting the computer from the network, the Omniscience Protocol must be able to continue to communicate even under these circumstances," writes protocol author Scott Bradner, a senior technical consultant at Harvard University.

Gotcha!
This April Fool's joke was published last year by the Internet engineering community, which issues a phony computer network RFC every year. The RFC Editor, the publisher of Internet standards and best practices, has been publishing April 1 RFCs since 1983.

Every year, network engineers world-wide await the release of the latest April Fool's Day RFC. These documents feature satire that would make Jonathan Swift proud (if he could understand the technical references).

"The best April Fool's Day RFCs reveal truth by telling a lie," Bradner writes in the introduction of a soon-to-be-published archive of April 1 RFCs. These RFCs "play it straight but describe something that could not or should not be done."

Bradner has authored or co-authored three April 1 RFCs and been the butt of another. "There were other attempts that did not see the light of day and probably should not have," he quips.

The trick in writing a good April 1 RFC is to make it look and sound real. Many of these documents have fooled network engineers into trying to implement them.

"I take pride in the fact that about half of the many comments I received about each of the last two April 1st RFCs I've been involved in thought that I was serious," Bradner says. The Omniscience Protocol "even got Slashdotted."

The RFC Editor started its April Fool's Day tradition more than 20 years ago, when Internet pioneer Jon Postel was in charge of standards publication. Since Postel's death in 1998, the RFC Editor shop at the University of Southern California's Information Sciences Institute has carried on the beloved tradition.

The RFC Editor solicits submissions for April 1 RFCs from the IETF, which is the Internet's premier standards-setting body. Submissions are accepted until mid-March, then the six-person RFC Editor team reads them and decides which is funniest.

Wanting to believe
"It has to be something technical that you want to believe in," says Joyce Reynolds, Internet Services Manager for the RFC Editor. "It has to be written in a very tongue-in-cheek manner."

The format of April 1 RFCs is the same as regular RFCs, and publication is announced by the IETF just like any other RFC. The only clue to readers that the document is a joke is the publication date of April 1. Regular RFCs list only the month of publication.

Some years, the RFC Editor team publishes many April 1 RFCs. In the last 20-plus years, the group has never run into the situation where they didn't have a submission humorous enough to publish.

"The fun part about it is that everyone in the IETF waits to see what we're going to publish," Reynolds says. "We know they're popular by the number of hits we get on our Web site."

The April 1 RFCs remain permanently in the official archives of the RFC Editor. Sometimes network engineers think they are real documents and try to prototype them.

"We'll get e-mail saying that someone is having trouble implementing an April 1 RFC," Reynolds says. "The e-mail could come years later."

Two of the most famous April 1 RFCs were written in the 1990s by David Waitzman, a senior software engineer with BBN Technologies who says each of the documents took less than a day to write.

In RFC 1149, Waitzman describes a technique for how IP packets can be transmitted via pigeon carriers.

"The IP datagram is printed on a small scroll of paper in hexadecimal ... wrapped around one leg of the avian carrier," Waitzman writes. "Upon receipt, the duct tape is removed and the paper copy of the datagram is optically scanned into a electronically transmittable form."

Pigeon lovers
"RFC 1149 is very popular in the technical community," says Bob Braden, who serves as RFC Editor. "There are many references to the avian carriers document in other publications."

The Bergen Linux User Group in Norway actually implemented this RFC in 2001. A reference to RFC 1149 also was slipped into Cisco product literature as a joke by an IETF leader.

Waitzman followed up in 1999 with RFC 2549 entitled "IP Over Avian Carriers with Quality of Service." This document discusses the pros and cons of having ostriches, robins, hawks, penguins and other birds serve as carriers of IP packets. "There are privacy issues with stool pigeons," Waitzman jokes.

The April Fool's Day RFC tradition "reflects the classic IETF attitude of not taking itself too seriously," Waitzman writes via e-mail.

"You'd never see the ISO or IEEE organisations do humour (which they would argue is exactly as it should be)," he writes.

Braden says the best April Fool's Day RFCs are original, describe something technically relevant and feature clever writing.

"Good satire is hard to write," he says. "People tend to be heavy-handed. Subtlety is really good. You want to laugh out loud the first time you read it."

Braden's favourite is RFC 3514, which was written by security expert Steve Bellovin, a professor of computer science at Columbia University.

Bellovin's RFC describes the creation of an 'evil bit' in the header of messages that mean to do harm such as spam, viruses, denial-of-service attacks and other malicious traffic. This evil bit would let firewalls, filters and intrusion-detection systems identify and block packets with malicious intent.

Bellovin says he wrote the evil bit document on a cross-country plane ride. "I'd been using the phrase for years when speaking about firewalls and network security," Bellovin says. "I finally thought of it at the right time of the year."

Publishing an April 1 RFC is a tradition that members of the IETF community look forward to every year.

"There's a lot of anticipation" of the April 1 RFCs, says Patrik Faltstrom, a Cisco Systems engineer and member of the Internet Architecture Board. Faltstrom submitted one April Fool's Day RFC, but it wasn't published. "Why spend energy on this? Because you need to have fun."