The Internet Protocol Journal - Volume 5, Number 2

Book Review

The Elements of Networking Style

The Elements of Networking Style , by M. A. Padlipsky, originally published by Prentice-Hall, 1985, ISBN 0132681110; now available from iUniverse, 2000, ISBN 0595088791.

Sometime in the autumn of 1986, I read Padlipsky on a flight from Boston to San Francisco, and about 15 minutes into it I began to get enraged. A few minutes later, I was snickering. By the time the attendants came around with profferings of alleged comestibles, I was laughing aloud, and a gentleman sitting near the window was grateful that there was a vacant seat between us.

Padlipsky brought together several strands that managed to result in the perfect chord for me over 15 years ago. I reread this slim volume (made up of a Foreword, 11 chapters (each a separate arrow from Padlipsky's quiver) and three appendixes (made up of half a dozen darts of various lengths and a sheaf of cartoons and slogans) several months ago, and have concluded that it is as acerbic and as important now as it was 15 years ago.

The instruments Padlipsky employs are a sharp wit (and a deep admiration for François Marie Arouet), a sincere detestation for the ISO Reference Model, a deep knowledge of the Advanced Research Projects Agency Network (ARPANET)/Internet, and wide reading in classic science fiction.

Arouet is better known by his pen name, Voltaire. He was a social rebel, a political agitator, and an acerbic satirist comparable to Swift. Isaiah Berlin, in a lecture published in Salmagundi 27 [1974], remarks:
"Voltaire is the central figure of the Enlightenment, because he accepted its basic principles and used all his incomparable wit and energy and literary skill and brilliant malice to propagate the principles and spread havoc in the enemy's camp. Ridicule kills more surely than savage indignation..."
Padlipsky is pungent and sharp and witty ... and knowledgeable. His critiques of X.25, of the International Organization for Standardization (ISO) seven-layer cake, and of the standards process in general, are still relevant.

In the early 1970s, the CCITT (now the ITU), made up of PTTs and monolithic telcos, fixed upon a putative standard for a network interface protocol, X.25. First approved in 1976, and revised in 1977, 1980, 1984, 1988, and 1992, X.25 was unsatisfactory in its original form and remains less than effective.

One of the greatest drawbacks is that it is basically a store-and-forward mechanism, meaning that it has an intrinsic delay and (as noted by Sangoma Technologies) this delay is typically 0.6 seconds. It also requires a great deal of buffering space.

Padlipsky's "Critique of X.25" (Mitre Corporation Report, M82-50, September 1982; RFC 874 12 August 1983) is revised as Chapter 9 in The Elements of Networking Style. Padlipsky has restored, however, his original title: "Low Standards."

Flush with the failure of X.25, the Consultative Committee for International Telegraph and Telephone (CCITT) moved ahead.

In 1977, the British Standards Institute proposed to ISO that an architecture was needed to define the communications infrastructure. To me, this, as with International Federation for Information Processing (IFIP), CCITT, and similar efforts, shows how "the road to hell is paved with good intentions." Because X.25 was unsatisfactory, the IFIP Working Group was set up in the hope that that the technological community could forestall the highly political arena of ISO. (It didn't.)

ISO set up a technical committee [ISO/TC 97/SC 16]. The next year (1978), ISO published its "Provisional Model of Open Systems Architecture" [ISO/TC 97/SC 16 N 34]. This was labeled a "Reference Model," and referred to as the Open Systems Interconnection Reference Model (OSIRM or ISORM—pronounced "eye-sorm"—by Padlipsky).

In general, it was based on work done by Mike Canepa's group at Honeywell Information Systems, which came up with a seven-layered architecture, which itself owed a great deal to IBM's proprietary Systems Network Architecture (SNA). SNA had been announced in 1974, and its seven layers do not correspond exactly to OSI/ISORM's. TC 97/SC 16 turned over proposal development to the American National Standards Institute (ANSI), to which Canepa and his technical lead, Charlie Bachman, presented their layered model.

This, in turn, was the only proposal presented to the ISO subcommittee at a meeting in Washington in March 1978. It was accepted and published immediately. A "refined" version of the ANSI submission to ISO appeared in June 1979. This published version is nearly identical to Honeywell's of 1977.

Rage and Ridicule
While he eschews the history I've outlined here, Padlipsky is enraged by the standards process and its results. As Dave Walden and Alex McKenzie (both then at BBN, both now retired) pointed out in 1979, both virtual circuit and datagram services are valuable. "An international standard would do well to support both." [IEEE Computer, September 1979].

The 1977–1979 models were such that extant host-host protocols did not fit ISORM. ISO was trying to construct a set of geometric figures that would be a "tidy model." The ARPANET workers, of whom Padlipsky was one, were interested in getting things to actually work. They were into pushing bits around the system.

The irascible Padlipsky has described the OSI system as two high rises with parking garages. The two high-rises are seven-story buildings; the parking garages are the three-story X.25 structures.

John Quarterman once pointed out:
"OSI specified before implementation. So specification took forever and implementation never happened, except for bits and pieces. In addition, heavy government backing (by the EC, now the EU, and various national governments) led some OSI participants to attempt to substitute official authority for technical capability. OSI and TCP/IP started at about the same time (1977). OSI wandered off into the weeds and TCP/IP won the race. Those governments that backed OSI bet on the wrong horse."
TCP/IP had clearly "won the race" by the early 1980s; it took till 1994 for the U.S. government to recognize the de facto standard by rescinding its Federal Information Processing Standards (FIPS). At that time, too, the Defense Data Network (DDN) was made up of IP router nets, not X.25-based nets.

In a totally different vein, there's Chapter 11: "An Architecture for Secure Packet-Switched Networks" (based on a presentation to the Third Berkeley Workshop on Distributed Data Management and Networking, August 1978). Here, Padlipsky suggests per-host processes. It was a really good notion.

Padlipsky's rants—and many of the chapters are just that—precede Quarterman's remarks by nearly a decade. But they are worth reading (and rereading).

I'm glad The Elements of Networking Style is available again.

—Peter H. Salus