Interesting People mailing list archives
IP: "DeJure ENUM, F.500, and Lessons Learned" -- by Tony Rutkowski
From: David Farber <dave () farber net>
Date: Thu, 12 Jul 2001 20:43:46 -0400
DeJure ENUM, F.500, and Lessons Learned by Anthony M. Rutkowski* For those folks who have been around the telecom and computer information fields prior to the past seven years or so, the current formal ENUM directions are taking some familiar turns. Indeed, the parallels of players, institutions, and schema are striking similar to one of the most costly and complete failures in information networking history - that revolved around "The Directory." The Directory was a central feature of a monumental standards making effort initially begun in the early 1980s as part of a kind of parallel universe to the Internet known as Open Systems Interconnection or OSI. During the 1980s, up until the mid-90s when it disappeared off the world radar screen like a plummeting rocket into oblivion, OSI consumed tens of thousands of meeting days and millions of pages of paper in a dozen different government-approved standards bodies around the world. OSI was supposed to define by government fiat the way all telecom and computer networking was to have occurred from the most trivial circuit connection to entire national network architectures and service provisioning arrangements. THE ORIGINS OF F.500 Today, telecommunication and information systems standards typically take the form of specifications or protocols developed in diverse forums that are taken by entrepreneurs and implemented in the marketplace in form of products and services. Some products, services and providers are successful; many are unsuccessful. The marketplace ultimately determines the winners and losers. In some cases, however, industry and governmental institutions attempt to manipulate the marketplace by giving some products and services special status. In the more extreme cases, they actually attempt to establish new regulated monopoly or government provisioning schemas that are imposed on the marketplace. In the old monopoly-oriented provisioning universe of telecoms, their international organization, the ITU, developed the specifications and protocols in various "series" of recommendations. For example, modem standards are part of the familiar V-Series specifications for data communication devices via the telephone network. In many instances in the past, the protocol specifications were complemented with F-Series standards intended to formally established a particular uniform global provisioning schema. Here the representatives of monopoly telecoms operators worldwide sat around a table and specified what users would and would not receive as "public services" in minute detail. In many cases, these F-Series provisions were in turn complemented by D-Series accounting and tariffing schema that specified the basis for charging users for services, and how the revenue was to be tabulated and split up among the monopoly providers. At one time, the actual amount of the fees paid in gold francs for the service was denominated. Thus, the "golden tree" was not only a windfall, but really golden! Generally, countries turned around and adopted F-Series and D-Series standards as law and were rigorously enforced worldwide. These provisions became the basis for maintaining global monopoly regimes for telecoms. For example, until comparatively recently private parties could not own or obtain telecommunication circuits and turn around and share the available capacity with others. In other words, Internets were not allowed. As telecoms players in the late 70s/early 80s envisioned themselves providing monopoly integrated telecom and information systems and services on a global scale, an extensive set of F-Series recommendations were adopted for Data Communications Networks. At that time, most telecom players viewed the Internet as a major threat to their business interests, and sought to impede its development and use through a variety of devises, including mandated OSI standards with F-Series recommendations. Even in countries like the U.S. where there was a breaking away from monopoly telecom provisioning, and where some agencies were supporting Internet development, other government agencies were enamored with the more regulated telecom approaches and jumped on the pro-OSI, anti-Internet bandwagon. In this vein, ITU-T Recommendation F.500 was developed in the 1980s to specify how the X.500 technical specification was to be manifested as a service in the marketplace. X.500 AND F.500 At the pinnacle of the OSI world was a highly structured, top-down hierarchical naming tree. It's purpose and functions were very similar to the Internet Domain Name System (DNS) - "name-address mapping." This structure was denominated "The Directory" and given the designation X.500. Like DNS, the X.500 specification, as noted in its opening section, was "not intended to be a general-purpose data base system" but responsive to "queries" to allow "mapping" of "user-friendly" names to addresses. It even specified that changes to The Directory "do not affect OSI network operation." However, it differed from DNS is several very significant ways. The purported "user-friendly" names were actually complex, lengthy, and cumbersome to use. The work was done through global standards committees entirely separated from any market or user feedback and based on what the standards participants thought desirable for users. The entire naming schema didn't allow for user choice or personnas, but created addressing formats that were imposed on end users - you got what was handed to you. And lastly, this was all implemented through a single Golden Tree with a single ITU root and country domains managed by rigid administrative structures substantially catering to governmental and PTT telecom monopoly regimes. Everything was confined to national borders, and transnational administration domains were not allowed. The last feature for achieving a Golden Tree is where F.500 came into play. F.500 specified a single global schema for implementing the provisioning of X.500 services. This was placed in the hands of ITU Member State Administrations who would themselves or through authorized parties whom they regulated, populate the X.500 golden tree and "make available for subscribers or users at the discretion of the Administration offering the service." As was common with F-Series recommendations, F.500 specified service requirements and features, user facilities, and the operational requirements and quality of service features that to be provided to the public. It even enjoyed its own complementary D.37 recommendation establishing revenue accounting and settlements among national F.500 providers. This was all said to be necessary to protect the public. SO WHAT HAPPENED? There is a kind of perverse side of "Internet Time." It's known as Internet Memory Loss. Much of the work described here was done in the early 80s, and all the international provisions were formally adopted in 1988. National implementation work occurred in the 90s. It all would have been mandated into place had it not been for two major developments. One was the breakup of the U.S. ATT monopoly combined with the FCC's Computer II and III rules that became manifested outside the U.S. through a variety of telecom monopoly busting developments. This included the European Commission, the GATT, and even the ITU (under Secretary-General Butler). The other was the dramatic growth and adoption rate of Internet technologies within private companies. The private marketplace - albeit confined to large corporations - began to subvert the public monopoly cartels and their so-called DeJure (i.e., by law) OSI strategies. These dramatic paradigm shifts were still substantially playing out in the early 90s - even in the U.S. Although contrary to FCC policies, the F.500 Golden Tree found substantial favor among the dominant telecom providers, the Department of Commerce, State Department, and some parts of DOD and GSA. They met for many months under the aegis of Dept of State sponsored advisory committees and worked through favored U.S. standards body umbrellas during 1991-92. The entire schema and copious regulations for administering the Golden Tree were developed and put into place for the US domain. The ANSI Secretariat won the bid for the national public registry, and the Dept of Commerce assigned itself the registry for governmental bodies. Even as late as 1993-94 as public use of the Internet began to scale, some major U.S. telcos were insisting that the only legitimate schema was F.500 and dissuaded use of the Internet domain name system as an inferior "DeFacto" service. F.500 lost dramatically in the marketplace not only because of the growing popularity of the Internet, but also because Internet DNS users could be masters of their own domains. They could craft themselves names that were relatively simple and meaningful; that were theirs. This was substantially enhanced by the lack of administrative rigidity, quick availability, relatively low cost, and the transformation of names into a public marketplace by pioneering entrepreneurs. Instant Messenger and similar business-driven, user-oriented naming schema also enjoyed great success. Amazingly, these F.500 Golden Tree provisions have never been rescinded. They are still the official domain schema of the ITU and in most countries, including the U.S. As of a recent tally, only a few hundred customers ever registered US domains. The story, however, isn't entirely negative. Pieces of the X.500 standard itself were extracted and a "X.500 Lite" was created and given the name Lightweight Directory Access Protocol or LDAP. It has been re-engineered by a number of bodies and vendors and become autonomously used on a widespread basis in the marketplace. Similarly, the piece dealing with security known as X.509 has become the cornerstone of Internet ECommerce and communications security. Here again, the success of the technology rested in the bottom-up adaptation, implementation, and use by commercial vendors in the marketplace. ENUM Deja Vu One of humankind's more persistent adages is that "history repeats itself." The admonition has a substantial basis in this arena. Last year a handful of standards participants quickly developed a X.500 like directory schema using telephone numbers and called ENUM. It was based in part on another ITU specification, E.164, that was married to an elegant if somewhat unnecessarily encumbering mandated pointer schema. However, it didn't stop at just being a protocol. They bundled it with a F.500 provisioning schema with a single DeJure root under the ITU. It was touted as the Golden Tree capable of mapping to all user communication applications. The proponents specified a rigid and complex administrative schema around country domains where users got whatever number was doled out to them. It was completely devoid of user demand or marketplace approaches. (Seem familiar?) Virtually the same repertoire of F.500 standards activities, advisory committees, and parties have been spun up over the past year. Copious detailed administrative provisioning specifications in an ever expanding number of standards groups are being created and beginning to disgorge pages of requirements documents. The same refrains of regulating to protect the public are offered. Long-standing FCC policies are being ignored. Official "ENUM roots" are proffered and "DeJure ENUM" blessing of governmental agencies is being sought. This juggernaut continues to go forward today - even though it makes little long-term sense even for dominant telcos today. Where this will end up remains to be seen. However, if history and user choices in marketplaces are any indicator, the DeJure ENUM effort will end up exactly where the DeJure F.500 effort did. On the other hand, to the extent that the technology and the underlying protocols of ENUM are slimmed down, further developed and offered in a lightweight fashion in an entrepreneurial marketplace, ENUM Lite may have a successful future. --- * These are personal views of the author and not necessarily those of any entity with which he is associated. I would also like to thank Dave Farber for the many years of facilitating discussion on important issues, among his many other activities. For archives see: http://www.interesting-people.org/
Current thread:
- IP: "DeJure ENUM, F.500, and Lessons Learned" -- by Tony Rutkowski David Farber (Jul 12)