6

Click here to load reader

Why users must co-operate internationally on standardization

Embed Size (px)

Citation preview

Page 1: Why users must co-operate internationally on standardization

57

Why Users Must Co-operate Internationally on Standardization Ray WALKER

Information Technology Users" Standards Association 1TUSA, Centre Point, 103 New-Oxford St., London WCIA IDU, UK

It is estimated that the availability and implementation of appropriate standards could reduce the IT costs of the average user by some 20% but only a small proportion have recognized that it is necessary for them to become actively involved if these savings are to be achieved.

The traditional standards making process is so slow and inefficient that most cost conscious users are reluctant to participate in it. This situation seems to be widely accepted but the procedural changes needed to improve standards making productivity and reduce development times dramatically seem obvious. General Motors and Boeing are demonstrating con- clusively the influence that large users can exert once the commercial incentive has become apparent.

Most suppliers also see commercial advantage in the devel- opment of standards but it is wrong to assume the objectives and priorities are the same. Users and PTI"s, for example, probably share a common interest in the standards which facilitate communication between users but all PTT's may not share the same enthusiasm for Addressing Standards which would allow users to change their carriers "at the flip of a parameter".

Commercial considerations of importance to suppliers, but of little interest to users, may reduce the value of functional profiles and conformance testing in ensuring that systems will interwork.

The paper selects a number of areas in which the absence of standards are significantly increasing the costs of users or impairing their organizational efficiency. It proposes practical steps which could by initiated by users to improve the situa- tion.

Kt~vwords: IT Standards, User involvement, OSI, ITUSA, MAP, TOP, ISO, ISDN, ISPBX, UNIX, National Protectionism, Standard Interface, Digital Private Network Signalling, Software Portability, Standards Making Process.

North-Holland Computer Standards & Interfaces 7 (1988) 57 62

1. Introduction

Almost without exception IT suppliers are IT users so why should it be necessary to distinguish between suppliers and users? Is it even possible to make a distinction?

It is increasingly common for the lack of user involvement in IT standards making to be noticed and commented on with concern. John Rankine, IBM, Director of Standards and Data Security, writing on the subject of integrated services digital networks has said i

"we are faced with a task that cannot be left only to professional standardisers, network providers and equipment manufacturers. For the proper foundations to be laid users, who will have to live with whatever structures finally emerge, must be prominent among today's architects of ISDN".

It is estimated that the availability and imple- mentation of appropriate standards could reduce the IT costs of the average user by as much as 20% so why are so few of them actively involved in the standards making process?

In distinguishing between the IT user and the IT supplier it can only be said that the prime interest of the IT user is the use of information

1 Computer Standards" & Interfaces" 5, 57-58.

R.S. Walker joined the Alliance, sub- sequently the Sun Alliance and London Insurance Group in 1945 and worked in a number of capacities in- cluding Manager Computer Planning from 1974-1980 which involved re- sponsibility for systems and program- ming for a large IBM installation (3031, 3032 and 3033 with about 100 systems and programming staff). He was manager Research and Planning with responsibility for corporate plan- ning until 1984. Now he is consultant

and Secretary General to the Information Technology Users' Standards Association (ITUSA).

Mr. Walker is Chairman of OIS/14, the BSI Committee responsible for data interchange at the application level and representative of ISO at UNECE meetings on Trade Data Interchange.

0920-5489/88/$3.50 © 1988, Elsevier Science Publishers B.V. (North-Holland)

Page 2: Why users must co-operate internationally on standardization

58 R. Walker / Why Users Must Co-operate Internationally

technology for the improved conduct of his affairs whereas the prime interest of the IT supplier is the sale of IT products for profit.

The nature of their interest in standards is very different.

2. The Suppliers' View of Standards

The technical content of a standard can be a major factor in the profitability of an IT supplier. It can make the difference between gaining a substantial market share and going out of busi- ness.

A supplier tooled up to produce a product which happens to be in accordance with a new International Standard is in a very fortunate posi- tion, particularly if that standard is demanded by the majority of procurers.

His less fortunate competitor tooled to produce a product which is not in compliance with the standard may be forced to re-design and re-tool at heavy cost. It is perhaps this factor above all others which means that suppliers cannot afford not to be involved in the standards making pro- cess. It is also the reason that there are so many options in standards which are not directly related to the need to provide variety of service.

On the subject of standard interfaces suppliers are ambivalent. Some, usually smaller suppliers, see them as essential for the marketing of their products which are used in conjunction with the products of other manufacturers. Standards may be less attractive to larger suppliers able to create their own environments and not dependent for their sales on the ability of users to interconnect products from different suppliers to create a workable system.

Unfortunately there is also another factor which has traditionally influenced the standards making process, national protectionism.

By developing national standards to which pub- lic authorities in that country procure they are able to give an advantage to their own suppliers. Large and efficient manufacturers in other coun- tries can only compete in that market by losing the benefit of scale of production and incurring the entry costs of creating a specific product. National standards tend to reduce the disad- vantage of the small supplier in relation to the large international supplier.

3. The Users' View of Standards

The IT user requires standards so that he has wide choice in the source of supply of hardware and software, can migrate gradually or incremen- tally as the demands of his business dictate and can communicate easily with other information technology users. His interest is in the availability of hardware, software and data interchange services which he can mix and match. He is con- cerned with the quality of the products and services but not usually with the technical standards which are embodied in them.

Products designed and built for world markets usually provide better value for money and wider choice of source of supply because the design and tooling costs can be spread over a substantially larger number of units of production. Whilst spe- cialised local requirements may favour local sup- pliers this benefit is at the expense of local users unless they really do have a unique requirement, which is unlikely.

This paper therefore concentrates on standards making through the eyes of a user and selects a few examples of users being disadvantaged by the absence of standards.

4. Need for a Standard Interface for IT Devices

The most obvious user requirement is for a standard interface for information technology de- vices. It came as a surprise to many information technology managers to find that open systems interconnection applies only to connections be- tween intelligent systems. It does not cover a standard interface between printers and computers or standard control protocols between devices.

One large supplier made it quite clear that they did not consider a standard interface to be in their interests "because it would only let the others in". This is probably a very sound commercial view for that supplier but it is not one which is in the interests of users.

Until an international standard exists which users can quote in procurement specifications basic and unneccessary incompatibilities between de- vices will continue to exist. Not just incompatibili- ties which are inevitable in consequence of differences in function or speed of device but incompatibilities which are either deliberately in- troduced or no efforts are made to eliminate.

Page 3: Why users must co-operate internationally on standardization

R. Walker / Why l~ers Mt~s't Co-operate lnternatinallv 59

5. Integrated Services Digital Networks

ISDN must be a critical area for users. If a standard basic interface is not agreed the supplier of an integrated services private branch exchange (ISPBX) may hope to control the sale of a high proportion of the devices for use on that ISPBX. Even at the simple level of the physical connector there have been problems and there could still be many difficulties ahead.

The digital networks are being based on exist- ing technology and because this varies from coun- try to country there is a tendency to think in terms of national variations, for example, in the char- acteristics of the power supplies for line powered devices. In some cases it may be that governments see an advantage for their national suppliers in promoting a national difference.

The investment in digital networks is at the moment negligible compared with the investment say 10 years from now. It is therefore important that users should make is quite clear to their PTT's, governments, and in Europe to the European Commission, that they are not prepared to tolerate national variations. Where the PTT's operate under a national monopoly or with some protective legislation, as they do in nearly all countries in Europe, the national controlling agency could, if necessary by legislation, enforce the provision of an international standard inter- face at no greater cost than a national variant. Where there are claims that the national variant has marked advantages for the users it would be interesting to see which the users select given the freedom of choice.

There is a problem with this proposal. There is still no international standard to be encouraged, let alone enforced.

Already digital networks are being established and users wish to create networks on them linking their ISPBX's to give the same facilities over the network as can be obtained on a modern ISPBX. Unfortunately there is no recognised international standard for digital private network signalling sys- tems (DPNSS). Suppliers and PTT's are therefore setting their own specifications. ISPBX's from dif- ferent suppliers and different countries will not interwork. International organisations wishing to create international networks and buy ISPBX's from local suppliers to obtain the benefit of local support are frustrated.

It is suggested that the CCITT will not, left to their own devices, get around to producing a standard before the middle '90's at the earliest. There are users who would have wished to have invoked the standard in their procurement policy a year ago had it been available because the PBX's being acquired now are likely to be in use well into the 90's.

The problem may be that the CCITT have little interest in creating the standards required. They are interested in interconnecting networks and within their own networks enforcing their own standards. DPNSS may have little attraction for them.

6. Need for Standards to Aid Software Portability

To provide portability of software it is essential that there are standard interfaces to hardware, operating systems, and other software packages but while setting the standards let us not forget the operator. How many variations are there in the commands just to copy a data set or a disc?

There is clearly a requirement for a standard for general purpose operating systems. At first sight UNIX comes close to satisfying that require- ment but on closer examination this is unfor- tunately not the case. It is a proprietary product whereas users would wish to be able to acquire their standard operating systems from competing sources. Unfortunately also, UNIX as currently marketed has a number of deficiencies. Suppliers are making these good in their particular versions of UN IX but as one supplier said with remarkable frankness " i f users take up the valuable extras we provide they will be hooked".

7. The Role of Users in Standards Making

The foregoing are only examples and there are a number of other critical areas where users re- quire standards to enable them to operate their businesses more efficiently or procure IT equip- ment more objectively.

If users recognise the importance of standards why do they not play a larger part in the stan- dards making process? The short answer is that many of them have tried but applying the normal criteria which they apply to their other activities,

Page 4: Why users must co-operate internationally on standardization

60 R. Walker / Why Users Must Co-operate Internatinally

cost effectiveness, they have concluded that they cannot justify continued involvement. If they calculate the cost of their efforts in highly skilled manpower, travel, hotel and other expenses and compare it to the value of the standard to them discounted by the probability of achieving a useful standard within their timescale the answer be- comes clear.

The answer is even clearer if they also take into account the fact that if to their surprise a useful standard is developed their commercial competi- tors who have not incurred the expenses of par- ticipating in the creation of the standard will equally benefit.

8. What is Wrong With the Standards Making Process?

wards the creation of the standards but by much more immediate business objectives. Wherever conflicting pressures arise it is inevitably the work on the standard which suffers.

This analysis ignores the less benign factors which may be involved. Some suggest that repre- sentatives of suppliers who may have more to lose by the creation of a standard than to gain from it undertake to carry out work on the standard be- tween meetings. Such offers are invariably accepted by others on the committee but somehow the work never actually materialises or is hope- lessly inadequate. Meanwhile technology is mov- ing on faster than the standard is created.

When examined in detail the surprise is not that standards making is so inefficient but that standards are ever created at all.

The question of what is wrong with the stan- dards making process can best be illustrated by comparing it with the way normal business pro- jects are conducted.

First an organisation identifies a profit making opportunity by taking into account the capital investment required, time to develop the product, size of the potential market and probable competi- tion.

If the omens are good it will then establish a business plan with a manager to implement it and a mechanism to review progress against the busi- ness plan. Remedial action will be taken if signifi- cant variances from the business plan are de- tected. The whole project may be axed if at any stage further expenditure seems unlikely to pro- vide an acceptable return on investment.

All those involved expect that success will be rewarded and failure dealt with appropriately.

The decision to proceed with a standard is the consequence of a democratic appreciation of a need. The people that make the decision seldom have money or careers at stake. There is no busi- ness plan, no undertaking to provide risk capital, no appointment of a manager, no timescale, no machinery for reviewing progress against the plan - if there was one, no jobs are at stake and the remuneration of those involved is seldom related to the outcome.

The work is even undertaken on a part-time basis by individuals whose career prospects are not usually determined by their contribution to-

9. Suggestions for Change

Having identified the problem what is the answer?

In order to create a standard the need must first be identified by one or more parties who should then specify the scope of the standard in reasonable detail together with an outline of the work required to create it. If not self apparent the work should include research into the require- ments of users and suppliers.

It is then important to ensure there is a com- mercial demand for the standard. The commercial demand may come from suppliers, from users or from governments but if that demand as indicated by a willingness to put up sufficient money to generate the standard does not exist there is no cost justification for the work.

To determine the finance required it is neces- sary to prepare an estimate of the resources needed to produce the standard and if an IT standard cannot be created within two years it may not be worthwhile starting. Information technology is moving too quickly. The idea that if a standard is started before the technology is mature it will inhibit progress is fallacious. A standard is after all no more than the choice of one of a number of technical possibilities at a particular moment of time. Every standard becomes obsolete and when it has served its purpose it should be discarded. The time to produce a standard is when there are sufficient parties interested to pay for its creation.

Page 5: Why users must co-operate internationally on standardization

R. Walker / Why Users Must Co-operate lnternatinallv 61

Sponsors of the standard should then raise the necessary finance and a manager and full time staff should be appointed to create the standard. The fact that the standard is in course of produc- tion should be well publicised in those places where interested parties are likely to be made aware of it. Regretfully this is not the journals of the National Standards bodies.

Those interested in participating should, on payment of a modest fee designed to cover the cost of distribution and processing receive drafts by electronic mail, not international post which may take at least five days.

Those receiving drafts should also be entitled to comment on them and the professional team creating the standard should attempt to reconcile these comments as far as possible, explaining the reasons for their decisions.

It is important that standards do not incorpo- rate avoidable options. The justification for an option should be the need to provide variety of function.

When it comes to voting on the completed document acceptance should not depend on na- tional voting. At the risk of offending almost everybody, who really cares what Switzerland, Sweden and Spain think about an IT standard, or for that matter the UK or France or Germany? As individual countries their markets are just not large enough to exert a major influence on the production policy of an international IT supplier. Even more important, the individuals who par- ticipate in the national committees that decide how to vote are seldom representatives of major purchasers. The success of a standard depends upon a significant percentage of the suppliers of that category of hardware or software being pre- pared to take it up and a significant percentage of users being prepared to buy to it. Indeed once the draft is stable it can be used for procurement whether it has international approval or not.

If anyone is wondering where the existing na- tional standards bodies stand in relation to this proposal the short answer is that they should be concerned only as providers of services. They may wish to provide interpretation of documents so that they can be readily understood by the inter- ested parties in their countries. They may also wish to provide national facilities for international teleconferencing.

There is no evidence that development of a

national view does anything but hamper the crea- tion of IT standards.

Most users do not want national standards. They seek widely respected international stan- dards and with the possible exception of the PTT's views do not coincide with national boundaries. It is much more likely that the large international aircraft manufacturing companies, petroleum companies or even car manufacturers have more in common than their respective countries. They are, after all, international organisations currently stuck with the problems of either buying interna- tional products which lack national support or buying local products which lack international compatibility.

10. Conclusion

The establishment of the Corporation for Open Systems, (for whatever reason), the creation of MAP by General Motors as a matter of necessity and TOP by Boeing indicates that something has to be done quickly. These initiatives are by sup- pliers, or users, or a mixture of both who are saying in effect that the existing standards making machinery is not satisfying their requirements.

Unfortunately in satisfying their business needs it is not unlikely that they will create a standards anarchy.

The time has come to try to restore, even extend, the international authority and respect of ISO. ISO has a place to play for perhaps it is ISO that should be appointing the managers and those technical teams of professionals. It is ISO that should be at the centre of the communications hub organising the teleconferencing, publishing the standards and drafts or rather making them available on a centrally held data bank so that anybody can access them. Incidentally how many users really do want a paper copy of an IT stan- dard? They may rarely want to refer to it but their real interest is in knowing how to procure to it, and what they should expect if they do.

There are those who say that the MAP and TOP initiatives could not have taken place had it not been for the fundamental work carried out over the last 10 years on the OSI framework. This is undoubtedly true but how many people realised even three years ago that having completed the OSI model and a large number of technical stan-

Page 6: Why users must co-operate internationally on standardization

62 R. Walker / Why Users Must Co-operate lnternatinally

dards it would then be necessary to agree func- tional standards? Creation of those functional standards or profiles is now revealing a number of gaps in the framework of technical standards which must be filled one way or another before the function can be performed.

Surely if users had really been in control or even understood the situation they would have started with a functional profile or at least a statement of functional requirements because it is that which is fundamental to their use of IT. The technical standards could then have been written within that profile not the other way about.

ITUSA is trying in a small way to pilot some aspects of the changes proposed above in connec- tion with the creation of standards for a General Purpose Operating System. This operating system will undoubtedly be based on the characteristics of the A T & T product UNIX and will draw heavily on the work of A T & T the IEEE and the X / O P E N Group.

ITUSA is seeking as allies suppliers who should benefit from the development but much more needs to be done by users in all countries if we are to achieve the standards we require.

Users should be able to find common ground more easily than suppliers because the actual con- tent of a standard usually matters little to them. They just require standards that are adequate, work, and are available in time.

Some of those standards may be commercially

unattractive to some large suppliers although there are undoubtedly other suppliers who would regard them as presenting an excellent marketing oppor- tunity. Users should therefore not only combine with other users. They should where appropriate~ seek alliances with selected suppliers in order to create an open and competitive IT market.

11. Status of This Paper

It is the policy of ITUSA to support ap- propriate international standards where they exist in preference to national or even European stan- dards. National standards are seen as having little relevance for IT.

It is the view of the ITUSA Board that the existing standards making machinery is seriously defective in several respects and is incapable of providing the IT standards users require in an acceptable timescale.

The suggestions for change are not ITUSA policy. They are presented in this paper to attract comment, whether critical or supportive. Reaction of international users, suppliers, standards making bodies and governments will be taken into account in developing ITUSA policy and the consequential action plan to achieve change. Change there must certainly be if users are to get the standards they want in a timescale which makes some sense in relation to the speed of IT product development.