INTRODUCTION TO TECHNICAL STANDARDS
Philip Doty

There has been a proliferation of technical standards of all kinds in the past several decades for all sorts of products, e.g., nails, tires, and computer hardware and software. Why do standards matter generally? Among the reasons you will see cited are:

• Safety
  
  
• Efficacy
  
  
• Efficiency
  
  
• Interoperability or compatibility
  
  
• Mass production
  
  
• Quality control
  
  
• "Guarding" capital investments

.The National Research council, in their 1994 report Realizing the Information Future: The Internet and Beyond, note that standards are "the conventions that permit the successful and harmonious implementation of interoperable networks and services" (p. 70). Among the major contextual factors in the development of IT technical standards are considerations linked to competition and trade in the increasingly global marketplace. Since most industrialized countries have standards setting bodies that meet in international bodies, international protocols such as the General Agreement on Tariffs Talks (GATT) are places where national and international competition are often clearly evident.

In most countries, the government takes an active role in establishing standards, especially for communication technologies. An archetypal example is MITI (the Ministry of International Trade and Industry) in Japan. Additionally, there are a growing number of public/private partnerships in major industrialized countries for setting technical standards. There is very little centralized governmental control of technical standards in the United States. Instead, we largely rely upon standards set by the private sector, especially through professional and industry associations. Such standards are market-based and self-imposed.

Setting Standards

There are two kinds of standards in temporal terms:

1. ex post facto, after a product is developed
2. anticipatory, before product development.

Tanenbaum (1996, p. 40, Figure 1-20) cites the work of David Clark at MIT, describing the timing of the standards setting process.

The Congressional Office of Technology Assessment report On Global Standards: building Blocks for the Future (OTA, 1992, p. 6, Figure 1-1 and p. 106, Figure A-1) presents us a 3 x 3 matrix (http://www.wws.princeton.edu/cgi-bin/byteserv.prl/~ota/disk1/1992/9220/922003.PDF).

This figure is a 3 x 3 matrix, with the mechanism for achieving the standard as the rows (de facto, regulatory, and voluntary consensus process) and the type of goals as the columns (control, productivity/quality, and process/interoperability). See the Executive Summary or the report as a whole for a more discursive description of these concepts.

In Realizing the Information Future (1994, pp. 70-74), the National Research Council (NRC) discusses several important elements of standards of interest to us. First they specify some of the important standards efforts relevant to the National Information Infrastructure: the Internet standards formulated by the Internet Engineering Task Force (IETF), the Open Systems Interconnection (OSI) network protocols, and standards for local area networks (LAN's) and metropolitan area networks (WANs) such as Ethernet and Token Ring formulated by committees led by the Institute for Electrical and Electronics Engineers (IEEE). We will discuss specific network standards later in the course when we address the TCP/IP (Transmission Control Protocol/Internet Protocol) protocol suite.

The NRC also identifies factors that complicate the setting of standards for digital networks. These factors also complicate other sorts of standards setting initiatives and include:

1. The fact that much of the functionality enabled by the Internet runs off the network on the user's end-node equipment, eroding the easy distinction we often assume between the network and users' systems. Such systems are often called customer premises equipment (CPE).
   
   
2. There is no government-derived mandate for setting standards.
   
   
3. The Internet relies on the bottom-up approach, as exemplified by the process of Requests for Comments (RFCs). Such an approach is hard to translate into an integrated, long-term strategy.
   
   
4. At the same time, a more integrated top-down approach seems inimical to the existing Internet culture and has proven to be unsuccessful thus far.
   
   
5. The power of strong commercial forces to influence the standards setting process may result in private, for-profit products' becoming de facto standards. While there are benefits to such an outcome, it has many negative effects, e.g., the reduction of competition and innovation, monopoly pricing strategies, and implicit extortion in the marketplace.