The following paper is available for circulation and comment. ================================================================ Version 3.1 (Final) 2 Aug 1991 Networking the Telecom Standards Bodies by A. M. Rutkowski * It is a classic example of the cobbler's children who had no shoes. Almost all of the scores of bodies throughout the world engaged in making telecommunication and information standards remain themselves without significant electronic internetworking capability. Although virtually every standards body has its material in machine readable form, and many have internal LANs, almost none have external access to that information; and there is no networking among the bodies. All the documents and standards are available only on paper through a few - and often user unfriendly - distribution channels. The only notable exception is the Internet Engineering Task Force (IETF) where everything is coordinated and openly accessible through the global Internet. The standards distribution problem partially arises from attempts to maintain artificially high paper-copy prices by creating a monopoly through a questionable copyright - a practice effectively impossible with open electronic inter- networks. However, almost anyone with one the new scanners available can easily provide an "optical gateway" into the network domain. If a sizeable black market in electronic copies of standards does emerge, the standards bodies will not be able to put the electronic genie back in the bottle. It is important for the bodies themselves to provide what users are demanding. Some major changes are in the wind, encour- aged by several factors. These include: rapidly changing technology and markets; participatory costs and lost expertise; new global open market norms; and increasing competition among the standards bodies. Perhaps most significant is the very recent availability of the needed internetworking capa- bilities on a scale that makes it feasible to apply to the standards bodies worldwide. This paper was prepared specially to stimulate wide robust discussion on this important subject among all the affected professional communities and standards bodies. It reviews the pertinent factors and recent developments, suggests how telcom standards internetworking could now be achieved almost immediately, and why it is in everyone's common interest to do so. * * * O THE NETWORKING TOOLS ARE NOW AVAILABLE. During the past 2 years, the information network- ing world has witnessed a revolution that is pro- foundly changing how organizations, professions, and individuals share information and collaborate in their work. The revolution is centered around the interconnection of thousands of information networks around the globe to form the open, mul- tiprotocol, cooperative meta-network called the Internet. The Internet began growing rapidly in the U.S. in the late 80's - more than doubling in size every year. Over the past two years, that exponential growth pattern was replicated around the globe. Most of telecommunications world is now con- nected; and shortly most of the remaining geo- graphical world will be. Current users exceed 3 million and are expected to reach 300 million by the mid-90s. Connectivity and interworking are very simple and inexpensive through modems, local area net- works, multiprotocol routers, and registration. Anyone can connect. The architect of the pan- Australian Internet backbone network AARNet recently described it "like going to the K-Mart." There are now scores of public initiatives and commercial ventures around the world to build and operate national and regional networks that are part of a global Internet. With access comes several simple, basic tools that include the ability: o to exchange messages with millions of users, o to search through and transfer files from thou- sands of open information hosts, o to access supercomputer resources, o to automatically propagate and receive news on hundreds of specialized subjects. Non-commercial and research users have free use of these tools because the technologies used are extraordinarily efficient, and because so many national and regional research and devel- opment initiatives worldwide all share the costs of the backbone transportation networks. New commercial ventures are now providing the same capacity at minimal cost to everyone. All of this is not the future - it is today. * * * O OTHERS ARE USING THE NETWORKING TOOLS. It is ironic that although the telecom standards bodies are not using internetworking tools, dozens of other communities ranging from high-energy physicists to primary school children are using these tools as a natural, important part of their daily environment. Indeed, I find that when almost anyone in college today visits the ITU in Geneva, they inevitably find their way to my office to TELNET (remotely log in) to their home host computer. Large active communities of diverse profes- sionals now share information and collaborate around the world. It has even spawned entirely new disciplines like collaboration theory and resource discovery. For example, those in high energy nuclear physics distribute bursts of experimental data from collision experiments for near real-time analysis and share supercomputers. This allows a physicist in Australia or China access to most of the same data and resources as if she were at the accelerator site. Molecular biologists share information on com- plex protein molecules and contribute to the mas- sive on-line data bases that automatically swap information between Europe and the US for mapping human genes . Most of the world's major library catalogues are accessible, and librarians are collaborating inter- nationally to establish standard automatic subject search capabilities. Automobile designers are sharing design ideas with their brethren in other countries. Physicians are forming speciality groups and using on-line medical references at the best dis- ease centers. The list of collaborative communities today is pretty big. But it is young people who seem to really love these tools. Educators are establishing global specialty groups - even projects to create "the global class- room." For example, a poor inner-city Hispanic primary school in Boston in internetted with a rural school in Costa Rica, allowing the children to share messages and files of video snapshots and drawings of their environment. Young stu- dents in Prague are forming discussion groups with counterparts in Australia and the US. Psy- chologists in the Soviet Union are participating in collaboration theory discussions with counter- parts in San Diego, California. And to make it even easier, the same MIT folks that brought us X-Windows are putting the fin- ishing touches on LogoExpress to allow the five- year old crowd to internetwork. Waiting for them is KidsNet - formed out of Norway to emu- late a cafi scene for children around the world to meet. * * * O APPLYING THE SAME TOOLS FOR STANDARDS MAKING. Those engaged in telecommunication- information standards making today have needs that are almost identical to most other profes- sional communities. In fact, the highly dis- tributed and autonomous architecture of the stan- dards world today (shown in the attached chart) is the very image of a distributed network. Every one of the bodies have three basic needs: o fostering collaboration (presently largely in the form of meetings and via messages.) o redistributing information (material provided to or solicited by the standards body that is subsequently redistributed, e.g., meeting doc- uments, questionnaire answers, chairpersons and rapporteurs, group lists, etc.) o distributing information (internally generated notices, news releases, or standards.) In performing these tasks, no standards body today stands alone, but is already part of a com- plex, increasingly non-hierarchical matrix of bodies where the information is constantly being transferred, compiled, and adapted among hun- dreds of different organizations. Already many of the companies and individuals that participate in standards making activities are part of the Internet, and increasingly in CCITT and IFIP groups, Internet SMTP mail address are regularly found next to people's names on the documents. If a poll were taken today, it would likely show that most companies participating in telecom standards bodies have either direct access or gateway EMail access to the Internet. Every telecommunication-information stan- dards body in the world is near enough to an existing Internet node, that with the simple addi- tion of a short local leased line and a multiproto- col router, connectivity among and with every one of those standards bodies could be attained. With just the basic tools of mail, file transfer, and remote log-in, the benefits to everyone associated with the global standards making, manufacturing, and service provisioning com- munities would be enormous - and it could be done within a month! * * * O WHY TELECOM STANDARDS BODIES SHOULD BE INTERNETWORKED. The fact that the tools exist for internally and externally internetworking the standards bodies will not by itself compel them to use the tools. However, there are many other developments that should provide strong motiva- tion. RAPIDLY CHANGING TECHNOLOGY AND MARKETS. At a recent meeting of a major standards organi- zation, a manufacturer delivered an eloquent message on the microphone. He simply said that the information-telecommunication technologies and markets were changing so fast today that his company had only about 18 months from the ini- tial feasibility of a product offering to its release, and that if a standard could not be developed within that timeframe, it couldn't be used. And even then, he noted, each additional month repre- sented major lost opportunity costs. With relatively few exceptions, the 18 month rule is the norm in today's information systems world. Even then, it is often necessary to adjust specifications to comport with constantly advanc- ing capabilities in basic technology implementa- tions in processors, memory, and transmission speeds. PARTICIPATORY COSTS AND LOST EXPERTISE. The costs of participating in standards making activi- ties have risen dramatically. These costs are not only actual cash layouts in terms of salary and travel to attend the ever growing numbers of meetings, and reviewing and writing documents. Also significant are the costs of losing expert individuals for weeks at a time around the year to meetings which often use their skills ineffi- ciently. Most companies are finding it increas- ingly difficult to support such costs; and the results are reflected in the current attendance lists of many standards meetings - where participation in too often skewed in the direction of large play- ers and particular industry sectors. Smaller companies and academic institutions - where many of the most creative and "hands-on" users of the technology abound - are effectively shut out of most of today's traditional larger stan- dards making activities. Also effectively exclud- ed are individuals from resource-limited develop- ing countries. It is very difficult to get the docu- ments which are almost exclusively distributed at the meetings. It is very difficult and exception- ally costly to get current standards in draft or even final form. Distribution is further impeded by legally unsupportable copyright assertions of many stan- dards bodies. Again the only exceptions are the IETF and new "startup" industry-user standards forums which are developing the industry's most successful standards by becoming meccas for the most innovative and enterprising individuals and companies. WIDESPREAD PIRATING. The cost and copyright concerns have already led to widespread pho- tocopying of standards. It has become almost equally easy with good, inexpensive, optical scanners and software to convert paper copies back into electronic images and place them on a server. Doing this is getting cheaper and easier by the month. There are a few locations already making available some unauthorized electronic versions of standards. If enough counter-culture people, research bod- ies, or even some "dare to sue me" commercial ventures scale up these activities, the standards bodies will lose control over the electronic stan- dards distribution process. It is eminently more sensible for the standards bodies themselves to recognize the need for robust, widespread distribution of good electronic copies of standards - and the resulting benefits to their organizations, the engineering profession and the industry. COMPETITION AMONG THE STANDARDS BODIES. One of the most significant developments in the telecom-information standards making world is the tremendous growth in the number of bodies engaged in this activity. It is, as CCITT Director Theo Irmer often repeats, "a competitive busi- ness." No one would argue that more standards are not needed in today's rapid paced, interoperating digital world. On the other hand, many of the new bodies have arisen for reasons other than just making more standards. One major factor is that no one body can do all the work in the required timeframes with the required specificity with the necessary service to local constituents. As a result, there is a layering effect where global bodies like CCITT and ISO make general - often abstract - standards or models with many options that are not capable of singular implementation. Because this lack of specificity often arises from disagreements among participants in the work, it is possible that earlier, faster, less formal electronic collabor- ation among the participants might bring about more complete standards at the global level. These standards may never have been tested to see if they actually work - even as they are adopted. (The IETF is perhaps unique in explicitly requiring extensive testing of a draft standard prior to adoption.) It is usually left to regional and/or national bodies and/or individual providers to then develop, flesh out, and test more detailed, implementable standards. Even separate conformance testing standards bodies have sprung into existence over the past few years to focus on the testing requirements alone. The quandary is that traditional standards mak- ing has tended to become a big, complex, and often inefficient business in an era where the market demands efficiency and speed. Recent new fast-track approval procedures pursued by some bodies like the CCITT and CCIR have significantly accelerated adoption timeframes. Still, significant liabilities remain because many of the big standards bodies fall prey to the tendency of all big institutions to devote large amounts of resources and energies to overhead. Such overhead includes costly activities devoted only to institutional needs - especially needless rote translations and inter/intra-body liaison paper - as well as the pursuit of standards that the ultimate potential consumers of standards don't need, perhaps don't even want, but serve the interests of a particular participant. It is the Sorcerer's Apprentice syndrome at work. Good electronic internetworking tends to promote bet- ter management, maximize horizontal commun- ication and minimize needless ritual transiting through hierarchies. Another aspect of the same problem is the dup- lication of effort that takes place among all the standards bodies simply because of a lack of knowledge that someone else is working on or even completed similar standards. Sharing information resources, including project manage- ment data, through an internet - perhaps even combined with automatic search capabilities - could save enormous monies and time, and also result in more globally uniform implementations for the same system or feature. Another major often-ignored factor is simply that the information-telecommunication industry has become very much more heterogeneous. New generations of entrepreneurs from Silicon Valley, Boston's Beltway, or France's INRIA who are interested in implementing virtual reality are not going to fit into the same institution with those who have engineered monopoly public carrier systems for fifty years. There is probably not a basis to communicate, much less work to- gether. So there are inevitably going to be mul- tiple standards making institutions, because insti- tutions are as much a home for cultures as they are for subject matter. However, in today's digital world, the subject matter significantly overlaps, and it is going to be increasingly critical to bridge these "cultural" barriers by allowing those in different institu- tional homes to collaborate. For all these reasons, there are today many standards making forums effectively in competi- tion with each other, and it is the marketplace, not the status of the institution which will largely decide which standards products are used and which are not. It is the users and the industry itself that are the benefactors because that is what the competitive marketplace is all about - max- imising the benefits to ultimate consumers of limited resources in a real world. NEW GLOBAL OPEN MARKET NORMS AND INITIATIVES. Layered on top of all other considerations are emerging new regional and global policy requirements based on antitrust and trade princi- ples and are reflected in legally binding agree- ments and law. The most notable are the draft General Agreement on Tariffs and Trade (GATT) Framework for a General Agreement on Services (GNS) currently being negotiated, and the vari- ous "open network" regimes promulgated in the European Community, the United States, Japan, and Korea. All of these developments require standards making processes be transparent and open, pro- vide prior notice, and easy access to the resulting standards and drafts. In the GATT, a widespread consensus is emerging in the direction of fair open global markets in equipment and services among the participating governments; and standards making has received close scrutiny because standards and testing requirements although generally benefi- cial, can also be used as de facto trade barriers and market impediments. In addition, the so-called Standards Summit process initiated at Fredericksburg, Virginia, in 1990 and subsequently continued as an inter- regional conference, is potentially leading toward an increasing cooperation among all standards making bodies in their own mutual interests in achieving much greater collective efficiencies. If this process is really opened up to all organiza- tions, it could provide a common institutional platform for global networking of telecom stan- dards making bodies. * * * O OVERCOMING THE OBSTACLES. Arrayed against all of these motivations toward networking are several serious obstacles - many of which are not very candidly discussed. These include copy- right, protocol wars, security fears, put it off until there is something more elegant, and local cul- tures. COPYRIGHT. Many - but by no means all - stan- dards bodies or their agents attempt to assert a copyright for their standards. The practice is usually justified purely as a pecuniary measure, but some arguments for "control" over the distri- buted copies are occasionally heard. More progressive standards bodies - and virtu- ally every standards participant and user - note instead that the test of success of a standards body is the extent of implementation of its standard, and not the income derived from the sale of documents. And although the counter argument raises the issue of income to these bodies, most of these bodies exist for and con- sume enormous participant monies - on the order of several hundred million dollars a year - for the single purpose of producing standards that are actually needed and used. Ironically, many organizations have found that the markets for paper-based and for electronic copies of standards are very distinct. In other words, the market for paper-based versions does not significantly diminish when electronic copies are made available. Despite the major broader interests at stake in standards making, it is the seeming blind pursuit of revenue obtained from the sales of paper copies of standards alone that impedes most standards bodies from electronically disseminating their standards - a position that seems badly balanced at best. Certainly the unquestionable great suc- cess of the IETF in getting its standards imple- mented around the world is in part due to the ease with which anyone can "FTP" the standards out of network servers. Raising the obstacle of copyright as a barrier to internetworking is also unfortunate because of the dubious ability under the law of most coun- tries of the world to maintain a copyright for standards. Some countries like Hungary expli- citly reject the copyright of standards, and many others reject it for anything in the nature of a national regulation or requirement. In addition, virtually all standards are very complex collective, derivative works in which the components come from a myriad of sources and authors. Many of the pieces are already in the public domain. Many of the standards bodies involved are directly or indirectly operating as agents for governmental agencies. It seems unlikely that any judicial body would uphold an assertion of copyright for most telecommunica- tion-information standards. The concern over "control" of the accuracy of standards under author's rights concepts can be handled through a combination of easy availabil- ity from authoritive servers and simple optional licensing or registration schemes. Internet com- munities already use these techniques for impor- tant reference material. For example, those who obtain a copy of a standard can be explicitly licensed to use the copy for their own purposes only; and users can optionally register to obtain subsequent versions. PROTOCOL WARS. One of the more contentious subjects already resolved by the internetworking community is the issue of competing protocols - often referred to as "religious wars." Different factions favour their own pet protocols - TCP/IP, DECNet, SNP, AppleTalk, OSI, etc - particularly factions which develop them for their own com- petitive advantage or someone else's disadvan- tage. The problems arise when the factions argue (always for the most noble reasons) for exclusive use of one protocol, or for gateways or interme- diate agents that favour one protocol or transit path over another. After dealing with these bat- tles for several years, the Internet community established the maxim of "give the user an equal choice." Thus multiprotocol routers and parallel applications are now the norm. Everyone has pretty much agreed that it is connectivity and facilitating use that is important, not protocol wars. This subject may be harder for some of the standards making community to deal with, because it is their own standards or constituents that are involved. For example, it is tempting for standards bodies that cater to public carriers to require all communications be routed through their own (generally high-priced) facilities and to promote their own standards. Thus X.400 traffic garners a lot of revenue, while SMTP mail does not. There is no reason why both should not be equally supported and allow users their choice based on convenience and cost. Most standards bodies are probably mature enough to avoid this kind of favoritism in order to promote the common good of the standards body itself, if not the engineering profession and public at large. No public carrier is likely to actually need the extra traffic to survive. The standards making community will be the ultimate benefactor of communications that provide inex- pensive collaboration, and dissemination of news and standards. SECURITY FEARS. Sometimes organizations are reluctant to connect to an open internet because of the fear of security breaches into privileged files or harm to their local networks. Although these are possible problems, there are simple steps that can be taken that provide effective remedies. An isolated public (anonymously accessible) server is but one example. Many of the largest companies and research establishments - even military facilities - are con- nected to the Internet. There is no standards body that deals with such sensitive material that it would prevent interconnection. Put it off until there is something more elegant. Another common argument against interconnection is that there are more elegant document standards and approaches on the hori- zon, and that the institution should wait until those approaches are widely implemented. Of course, there is always going to be some more elegant solution on the horizon. Users, however, generally do not want elegant solutions, they want minimal tools to get the job done. In this case, they simply want access to the elec- tronic file versions which generated the paper usually sent through the post or distributed at meetings. It is common on many servers on the Internet today for the same document to exist in several different versions: plain ASCII, native format, postscript, and compressed versions of all three. The user simply accesses the directory and trans- fers the version that suits her or him. Again, it comports with the maxim LET THE USER DECIDE. LOCAL CULTURES. Lastly, there is the issue of local computer environments that have their own favorite approach - whether it be applications, operating systems, or information agents - to fulfilling the needs of the organization. To a greater or lesser degree, this problem exists everywhere, because people and organizations tend to become familiar with their own self- learned solutions to local needs. Local cultures need not be a barrier to internet- working. Indeed, the very concept of internet- working was fashioned to accommodate the great diversity of local cultures, machines, and systems that exist. The only common element is at the point of interconnection where common proto- cols are supported by everyone in the common interest of achieving a meta-network. * * * O A MUTUAL INTEREST TO ACT NOW. The global telecom standards community stands at a fairly unique confluence of events where internetwork- ing could be greatly facilitated. The time to act is now. On an experimental basis, ITU Secretary-Gen- eral Tarjanne has recently taken an important innovative step by allowing the Digital Resource Institute project at the University of Colorado to place standards of the ITU's International Consul- tative Committees (CCITT and CCIR) on servers connected to the Internet. Usage patterns will be monitored and intelligent directory programs will be tested. The latter would allow a user to request, for example, all standards that deal with and . With internet- working, the same kind of automated searching techniques could be easily applied across all the standards organizations - a capability of enor- mous potential worldwide benefit and cost sav- ings - as McGill University's Archie system has already demonstrated together with linked servers in Finland and New Zealand. The second session of the "standards summit" of regional and international standards bodies will meet in September at Nice. This is the per- fect opportunity for an initiative were all of the many telecommunication and information stan- dards bodies, conformance testing bodies, infor- mation object registration authorities, and indus- try-user standards forums agree to cooperate in connecting to the Internet and in sharing all basic management information and standards docu- ments. The CCITT's Resolution 18 Group on working methods is meeting in late October in Geneva to lay out user information system support needs. Already there has been a significant focus on providing the CCITT standards making com- munity with significantly greater network tools; and if these tools can be make available now, the opportunity exists to integrate them into the future working methods of the body. Even as Internet connectivity goes forward, there will be new challenges. It will require, for example, more coordinated management among all the standards bodies to learn how to best hori- zontally collaborate. Experimentation will be necessary in how to structure and manage com- plex standard development projects; to encourage shared, timely goals among stategically competi- tive participants; to examine if, when, and how to standardize; and to optimise information flows through the internet architecture. At present, the global standards making archi- tecture is fragmented into many isolated camps based on history, membership, and cultures. The barriers to cooperation must be bridged, and the cherished views of institutional superiorities must be diminished. A new kind of "standards democ- racy" must emerge which compares and supports standards on their merits, and doesn't automati- cally regard one standards body as intrinsically better than another. In many cases, more effective administration and project management capabilities need to be developed within many standards bodies. And along with culture goes the necessary work- arounds to deal with the many individuals who are reluctant to use electronic information and internetworking tools. But these are welcome challenges in the face of the enormous benefits to internetworking the standards bodies. With several hundred million dollars a year being collectively invested in information-telecommunication standards activi- ties, the potential monetary savings alone are enormous - not to mention the value of develop- ing standards that are more appropriate, better, and more used. * * * O CONCLUSION. It is hard to imagine today a global community more appropriate for internet- work resource sharing and collaboration than the many telecom-information standards bodies, par- ticipants, and users at national regional and inter- national levels. Everyone associated with this community would reap significant benefits. Indeed, such a result will become almost impera- tive if high level policies envisioning open mar- kets are to be implemented in this economic sector. The internetworking tools to achieve this result are now readily available and easily implemented at negligible cost. It really is time to act now. * * * * * ------------------------ * Note: The author is counsellor to the Secre- tary-General, International Telecommunica- tion Union, Geneva, Switzerland, and a research associate at the Massachussets Institute of Technology. The views expressed are solely his personal views and are not official positions of the ITU or MIT. Thanks is given to colleagues is several different standards bodies who reviewed and provided additional ideas for this paper. He may be reached at or . 1 Some standards bodies are currently experi- menting with PSTN and PSPDN dial-in videotext bulletin boards, and messaging using X.400 and private EMail services and gateways. 2 Project descriptions and papers of the Institute are available by anonymous FTP from or by EMail to or . 3 Archie is a McGill School of Computer Science Archive Server Listing Service. It contains a central database for information about archive sites anywhere on the Internet and can be automatically searched. Additional information is available by EMail to . Phone calls are discouraged. ==============================================