T h e m e Issue on IT S t a n d a r d s Guest Editor's Introduction
Even if we disregard social, moral and religious rules for the moment, standards--in a very general s e n s e - - h a v e been with us for a long time. About 5,000 years ago the first alphabets emerged, enabling completely new forms of communication and information storage. Some 2,500 years later, the first national, coin-based currency, invented by the Lydians, established the basis for easier inter-regional and even international trading. The industrial revolution in the 18th century and, more so, the advent of the railroad in the 19th century resulted in a need for technical standards, which was once more reinforced w h e n mass production generated a demand for interchangeable parts. In parallel, the invention of the electric telegraph in 1837 triggered the development of standards in the field of electrical communication technology. In 1865 the International Telegraph U n i o n - - t o become the International Telecommunication Union (ITU) in 1932--was founded by twenty states. The other major international standards setting body, the International Organization for Standardization (ISO), was established in 1947. These days, a web of Standards Developing Organisations (SDOs) at the global, regional, and national level produce what is commonly referred to as"de-jure"standards--although n o n e of their standards have any regulatory power per se. Likewise, a plethora of industry fora and consortia (a recent survey found more than 260) produce so-called"de-facto"standards almost by the week. As a result, there exists an almost impenetrable maze of what is generally called"standards,'ranging from company specific rules, over regional and national regulations, up to globally accepted standards. Moreover, one may distinguish between different types of standards: there are voluntary, regulatory, de jure, de facto, pro-active, reactive, public, industry, and proprietary standards; this list is by no means exhaustive. Indeed, it pretty much looks as if Andrew Tanenbaum was right: "The nice thing about standards is that there are so m a n y to choose f r o m ' . ...........
But--what exactly establishes a"standard"? The seal of approval from one of the"official"standards setting bodies? The degree of usage of a sysKnowledge, Technology, & Policy, Summer 2001,Vol. 14, No. 2, pp. 4-7.
T h e m e I s s u e o n IT S t a n d a r d s
5
tem or a product? Do consortia actually issue"standards"? And are the Internet RFCs published in the STD series standards? Ask any three people and the odds are that they will come up with at least four different opinions. As does the literature. For instance, Webster's N e w Universal Unabridged Dictionary defines a standard as "7~n authoritative principle or rule that usu-
ally implies a model or pattern for guidance, by comparison with which the quantity, excellence, correctness etc. of other things may be determined." The Oxford English Dictionary says that a standard is "The authorized exemplar of a unit of measure or weight; e.g. a measuring rod of unit length; a vessel of unit capacity, preserved in the custody of public officers as a permanent evidence of the legally prescribed magnitude of the unit." According to ISO, "Standards are documented agreements containing technical specifications or other precise criteria to be used consistently as rules, guidelines, or definitions of characteristics, to ensure that materials, products, processes and services are fit for their purpose."These definitions already hint at a major dilemma in the theory of standardisation: there is no generally agreed definition of what constitutes a standard, and the definitions that do exist cannot really be meaningfully applied to the IT domain. A definition I once came up with read: '~4 publicly available definitive speci-
fication of procedures, rules and requirements, issued by a legitimated and recognised authority through voluntary consensus building observing due process, that establishes the baseline of a common understanding of what a given system or service should offer." This restricts the scope of what is colloquially referred to as a standard in three ways: firstly, it includes only base standards (the "baseline"); as opposed to functional standards or profiles, which rather more address implementation interoperability issues. Secondly, it limits the sources from which a standard may emerge to"recognised authorities.'In particular, this excludes specifications issued by largely self-styled industry fora. Finally, as standards are said to be established'through voluntary consensus building', this definition also excludes legislation from being seen as standards. Thus, the sources from w h i c h s t a n d a r d s m a y e m e r g e are limited to recognised national, regional or international standards setting bodies. In any case, standardisation represents an interface between technical, economic, organisational and social considerations. Standards are not rooted in purely technical deliberations, but also result from a process of social interactions b e t w e e n the stakeholders and, ultimately, individuals. And that's precisely what makes the whole topic so interesting. Today, (the h y p e s u r r o u n d i n g ) e - c o m m e r c e , m - c o m m e r c e , a n d globalisation in general call for even more standards.Yet, there appears to be a shift of focus--from technical to process. Companies more and more realise (believe?) that they need standards for business processes to enable both intra- and inter-company transactions, most notably for B-2-B electronic commerce.Vendors of e.g. ERP systems already offer support for the adaptation of customers'business processes to the capabilities of their software. Personally, I have my doubts w h e t h e r this is the right way to go; after
6
Knowledge, Technology, & Policy ! Summer 2001
all, business processes are a crucial means for a company to distinguish itself from its competitors, but this is just a view from academia and quite possibly far removed from harsh business realities... Be that as it may, firms look to standards setting for several reasons which are typically related to their own economic well-being. Thus, it is hardly surprising that the economics of standards is the by far best researched field--a fact also reflected in the papers of this special issue. Yet, there are other questions surrounding standards. Do they really hamper progress and stand in the way of technical innovation? A very popular perception, but is it accurate? Should we leave it to the market alone--and its hype--to decide about winning technologies? Technology studies tell us that this would be unwise, but are they correct? And, do we actually need users in the committees? Most people would say"yes'--rightly so? Unfortunately, the papers of this special issue will hardly come up with definite answers to these questions. They will, however, provide you with additional insight into some of the most interesting issues surrounding standards and standardisation The Papers
This Special Issue comprises three full papers; by Egyedi, West & Dedrick, and Gaynor & Bradner. In addition, four short papers are included. Discussing a considerable variety of aspects, the work presented in these papers will (hopefully) contribute significantly to a broader understanding of the very diverse aspects that make up the probl6matique of"IT standards and standardization." We begin with a rather"light-hearted" paper by King-Tim Mak and Arkalgud Ramaprasad (U. of Illinois). In their paper"An Interpretation of the IS/IT Standard Game, Circa 2001"they discuss past, present, and future of IT standardization, and argue that standardization of business processes will become increasingly important. The next two papers look at different aspects of what is easily the most popular domain within standards research (in terms of number of publications, that is)--economics. The paper by Bernd Reitwiesner and StefanVolkert, (U. of Augsburg, D), entitled"Standardization and the Competition on the Market for ERP-Systems" presents a microeconomic model to evaluate the pros and cons of standard software vs framework technology. Mark Gaynor's and Scott Bradner's (Boston U.; Harvard U., US) paper "Using Real Options to Value Modularity in Standards" proposes a new prescriptive model of technology standardization under uncertainty and shows how its value is quantifiable using the theory of real options. Through case studies, the next two papers discuss the importance and the impact of standards on such diverse things as software projects and chipcards, respectively. Roy Rada's (U. of Maryland, US) and John S. Craparo's (American International Higher Education Corporation) paper on the problem of"Stan-
T h e m e I s s u e o n IT S t a n d a r d s
7
dardizing Management of Software Engineering Projects" presents case studies of corporate standardization across software engineering projects. In the paper "Banking Chipcards in the Netherlands--One or two Syst e m s ? ' H e n k deVries (Erasmus U., NL) discusses whether or not the decision by one Dutch bank to introduce a second chipcard system in the Netherlands was justified. Last--but most definitely not least--the final two papers approach the topic from the perspective of Technology Studies. At least in my view, this approach deserves much more attention; so maybe these papers can contribute to its wider appreciation. Joel West and Jason Dedrick (U. of California, US) present a view of how a new standard emerged to successfully compete with an established one. Their study o n " O p e n Source Standardization: The Rise of Linux in the Network Era"examines the rise of the Linux operating system. "Strategies for de facto Compatibility: Standardization, proprietary and open source approaches to Java"by Tineke Egyedi (Delft U. of Technology, NL), describes the standardization strategy Sun Microsystems deployed for their Java middleware platform. Finally, I would like to thank the reviewers. Without them, a meaningful evaluation and selection of papers would have been next to impossible (well, it would have to be based on my personal preferences, which pretty much contradicts"meaningful"). Thus, my thanks go to Knut Blind, Heide Coenen, Ole Hanseth, Wolfgang K6nig, Ken Krechmer, Rob Procter, Tim Schoechle, Mostafa Sherif, Michael Spring, Klaus Turowski, AlexanderVoss, Michael Wallbaum, Tim Weitzel, and Robin Williams. Kai Jakobs, Guest Editor