From carson@siggraph.org Mon Jun 16 19:02:57 1997 Received: from siggraph.cgrg.ohio-state.edu (siggraph.cgrg.ohio-state.edu [128.146.18.100]) by dkuug.dk (8.6.12/8.6.12) with ESMTP id TAA09621 for ; Mon, 16 Jun 1997 19:02:54 +0200 Received: from study.huntleigh.com (carson@siggraph.org) by siggraph.cgrg.ohio-state.edu (8.8.5/941010.52) with SMTP id NAA12788 for ; Mon, 16 Jun 1997 13:02:39 -0400 (EDT) Message-Id: <3.0.32.19970616110444.00c76424@siggraph.cgrg.ohio-state.edu> X-Sender: carson@siggraph.cgrg.ohio-state.edu X-Mailer: Windows Eudora Pro Version 3.0 (32) Date: Mon, 16 Jun 1997 11:05:25 -0600 To: SC24@dkuug.dk From: Steve Carson Subject: HOD-C ideas on reengineering Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by siggraph.cgrg.ohio-state.edu id NAA12788 This document records the discussions that we had in Kista during HOD-C a= nd other forums regarding JTC 1 reengineering and the future of SC 24. It i= s a collection of text that is circulated so that SC 24 National Bodies may extract portions of interest for use within their own countries when preparing positions for their JTC 1 counterpart organizations. This material may be part of a future contribution from SC 24 to JTC 1 or it=92= s reengineering ad hoc. The Value of ISO/IEC Standards We believe that both JTC 1 and organizations outside of JTC 1 have appropriate roles to play in the development of standards. The subcommittees of JTC 1 are not the only place to either initiate work on new standards or to evolve existing standards. For this reason SC 24 and other subcommittees of JTC 1 have long worked with outside organizations, including those open, consensus-based Consortia and similar organizations where all individuals and organizations with an interest can freely participate on an equal basis. =20 SC24 (as well as other subcommittees) have extensive experience in transposing specifications created outside JTC 1 into International Standards. Our experience has shown that these initial submissions, without exception, fall short of the quality that the global community demands of an International Standard. These shortcomings go well beyond the obvious ones of formatting, style and spelling. The specification mu= st also have characteristics such as completeness, consistency, and correctness that allow multiple, independent, inter-operable implementati= on to be built from just the specification. Determining whether a specification has these characteristics is not a management decision that can be taken at the technical committee level, but rather a technical one that must be made at the subcommittee level. Producing high quality specifications has always taken time. This is especially true of building consensus at the working group and subcommitt= ee level. It is tempting to seek a short-cut methods of creating International Standards, such as the Fast track and PAS processes develop= ed by JTC 1. Even though these processes may provide a quicker way to get a= n initial submission than developing it from scratch within a working group= , deciding that a submitted specification has the necessary characteristics still requires review by technical experts. The main value that JTC 1 contributes to the global information technolog= y community derives from the expertise of the thousands of technical expert= s who actively contribute at the working group and subcommittee level. It i= s the talent of these contributors that has consistently transformed initia= l submissions into high quality International Standards. If lower quality specifications are allowed to become "International Standards" by paths other than the traditional multi-step process followed by JTC 1's subcommittees, then the excellent reputation that ISO and IEC standards enjoy for high quality will quickly be destroyed. Therefore, JTC 1 must maintain a cadre of world class technical experts in each field of information technology in which it wishes to create standards, and must require that these experts pass technical judgment on all submitted specifications. SC24 has recently enjoyed success working with outside organizations to transpose work that was essentially complete into International Standards with minimal change. Unfortunately, we have found that the limited scope for technical contribution has meant that only the most senior technical experts are willing or even able to be involved in what are only the fina= l, principally editorial, stages in the development of such standards. Thos= e at earlier stages in their careers must work in the early and highly innovative stages of development, at least in part so they can get the publications in refereed journals that are required for career advancemen= t. Further, a substantial number of experts, particularly in Europe, are recruited from research organizations and public institutions. The participation of these experts in standards making is tenable only if suc= h activities are relevant to the mission of these organizations. The ability of subcommittees such as SC24 to recruit and train technical experts is dependent on their ability to get their experts involved in th= e work at the creative stage where they can contribute innovative and new ideas. SC24 has experienced a significant drop in participation over the last several years due to precisely this shift of the early work into outside organizations such as Consortia. The subcommittee will not remai= n viable much longer unless it can attract and retain new experts at an ear= ly stage in their careers. If subcommittees such as SC24 disappear, JTC 1 will have no pool of experts who can do the technical and editorial work that we know is required to create International Standards of acceptable quality. Once this happens the entire future of global information technology standardization will be in jeopardy.=20 How Should JTC 1 Select Projects If we approve standards projects based only on their perceived short-term= s payoff, many of the most worthwhile projects might never have been starte= d. It is instructive to look at several specific examples. The use of International Standard character sets rather than proprietary ones took a long time to become widely implemented in products. Today their use is commonplace. The Standard Generalized Markup Language (SGML) was used mostly in only small, niche markets for many years until its use by the W= WW community in HTML catapulted it to greater attention. The Computer Graphics Metafile (CGM) was little used in 1986 when it first became an international standard, but is enjoying rapidly growing adoption in vertical markets today where there is a need for a device, system and resolution independent format. The lesson to be learned from these examples is there must be a balance between urgently needed projects with short-term payoff and those required to build markets in the longer term. The ability to stand back and take this longer term view is one value of International Standards. Another fact that we must consider is that the value from International Standard projects is not solely from the economic value of its products (the standards themselves.) The side effects of carrying out an international project are often of considerable importance. Primary amon= g these are the educational benefits. Within the SC24 computer graphics community the development of the GKS and later the PHIGS standards illustrates these benefits well. These projects led to a common world-wi= de understanding of computer graphics concepts and terminology to a far greater depth than could be achieved by other means, such as the interaction of experts at meetings of professional societies or at conferences. When a group of experts from diverse backgrounds works together over many years to create a standard, they are forced to develop= a much deeper understanding that can be achieved by more casual types of interaction. In computer graphics, the concepts of GKS and later PHIGS have been used to train the generation of technologists who have gone on = to make computer graphics an intrinsic part of all of our lives today. This training has also enabled several countries which had not previously participated in the international computer graphics market in a major way to now become world leaders in aspects of the technology. The Way We Develop Standards One of the reasons that so much early stage the development of new standards is accomplished within Consortia today is that Consortia are fr= ee to utilize innovative methods of working, including Internet based electronic communications. While JTC 1 has encouraged its sub-organizations to utilize electronic means to do their work, the key decisions still must be taken by formal ballots with time frames (3 to 4 months) that can be viewed as excessive by today=92s standards. The rece= nt experience of the VRML community has shown that consensus can be achieved in shorter times if reasonable practices are followed. If the subcommittees and working groups of JTC 1 are to be competitive as developers of future standards, the JTC 1 directives must be revised to allow subcommittees and working groups more flexibility in establishing t= he most appropriate procedures for developing consensus on a particular proj= ect.=20 A major reason that both information technology providers and users of information technology often choose a specification from a Consortium or = a single commercial organization over an ISO/IEC standard is that the forme= r specifications can usually be obtained quickly at no cost over the Internet. For this reason, International Standards are often not competitive with other specifications. To ameliorate this situation, the International Standards in the field of information technology developed = by JTC 1 should be accessible freely through the Internet. This free accessibility should extend to pre-standard documents such as Working Drafts, Committee Drafts, and Draft International Standards. Since so much standards development work is now done in partnership with organizations outside of JTC 1, it is important that both sides realize that the goals of interoperability and consensus can only be achieved by cooperating in an open manner. From the JTC 1 side this means taking a relaxed and flexible view of the specifications that might appear as normative references within a standard, as well as being cautious about demanding compatibility with established International Standards, especially ones that might not have wide marketplace acceptance, or whose features are not required for interoperability with the installed base of commercial products. From the Consortia side this means considering alternative approaches based on the experience of ISO and IEC experts, especially in the area of internationalization. Decisions on issues be ma= de on both the technical merits of a proposal, and on other considerations including interoperability. Often both sides can adopt a phased approach = to change whereby a specification is scheduled to evolve over time rather th= an making all desired changes at once. Balanced Interests In the rush to reengineer JTC1, we must not forget that ISO and IEC have the responsibility to balance the interests of a diverse set of interests in the development of International Standards. The perceived short-term interests of the producers of information technology, while just one of these interests, is the main one driving JTC 1 reengineering. While the full list of other interests is too large to discuss here, it is again instructive to look at several examples. Larger producers of information technology can make sure that their voice is heard due to the size of their staffs and the amount of money that the= y can spend. These organizations have always sought to preserve their current market share by controlling how standards are developed. Small a= nd innovative organizations, both for-profit companies and research institutes, are greatly under-represented above the subcommittee level within ISO and IEC. Yet most innovations come from these smaller organizations and it is mostly their growth that fuels the global information technology economy. It is easy to see why the larger produce= rs want to lock them out of global markets by denying them an equal right to contribute to the development of International Standards. Another minority interest is that of constituencies that are disadvantage= d in some way. ISO and IEC have a special responsibility to insure that their interests are fairly considered. Among these are a variety of smal= l constituencies with unique language or cultural requirements that product= s based on International Standards should support even though each unique requirement may not itself be economically feasible to meet. Conclusion We must strive for a global information technology standardization community where all interested parties are free to contribute and where their views are considered on an equal basis. There is a price that we mu= st pay to meet this goal and insure that no one is left behind in our rush t= o develop standards in a more timely and responsive manner. One of the prices that we must pay is insuring that technical experts at the working group and subcommittee level can still make meaningful and innovative contributions to International Standards. --------------------------------------------------------- Steve Carson phone: +1-505-521-7399 GSC Associates Inc. fax: +1-505-521-9321 5272 Redman Road e-mail: carson@siggraph.org Las Cruces, NM 88011 USA ---------------------------------------------------------