Information Centre as Strategy

Transcribed from the printed issue; not yet reviewed. Read it in the PDF of the issue, page 90

The challenge facing the DP Department today is that of meeting demands for flexible and efficient access to, and control of, the information resources at the heart of any successful company’s operation. The Information Centre potentially does meet the challenge. There are however many pitfalls, and as the issues are very much at the centre of an organisation’s total strategy for computing, communications, and office automation, it is important to understand the history of the development of Information Centre concepts. There are perhaps more accounts of failure than of success, but this should not deter an organisation from seeking to follow what is fundamentally a very sound strategy.

The Information Centre concept came about in the early 1970’s. Traditionally, DP departments had been predominantly batch or transaction processing in their orientation. Unhappily, interactive users were generally unwelcome on such systems, as their demands were spasmodic, unpredictable, and all too often liable to cause system crashes. Controllable, regular jobs run as card decks were much more acceptable to the DP manager with his scant resources.

The gulf that arose between the DP department and the end users is one familiar to most large organisations. It benefited only the external service companies, who were able to charge premium prices for access to systems which end users felt were much more responsive to their real needs.

However, as computer hardware costs came down, and minicomputers (and later, microcomputers) came almost within the budget capabilities of user depatments, serious problems of anarchy arose, as users set about solving their problems entirely outside the influence of the DP department. At the same time, the growth of such systems, isolated by their incompatibilities, began to cause real difficulties in the efficient control and integration of the information resources of the organisation.

The concept of the Information Centre¶

The Information Centre was a concept, originally developed by IBM, which attempted to bring order to the impending chaos. Its fundamental aim was to shift the DP focus back on to the end user, but under the control and direction of the Data Processing department. This aim would be achieved by providing tools and basic support from the DP group, together with sufficient education to allow the user to solve his own problems, largely in a “Do It Yourself” mode.

The Information Centre was a management philosophy, implemented in software. It offered a small number of “strategic” software packages and building blocks aimed at end users, typically in the M.I.S. department. Basic database mechanisms were provided to allow the user access to corporate data.

Hence, the increasing levels of end user expertise in the use of computer systems was channelled into a relatively small number of basic application systems, and the DP department role moved from one of total responsibility for the development of systems to one of encouraging user self sufficiency within a controllable environment. Backlogs in system implementation timescales would largely disappear, and in any event they were now on the shoulders of the user himself.

The concept was excellent. It was welcomed from all sides of the debate, and quite a few organisations established Information Centres which even today still serve their companies well. A truce was established between the users and the DP department.

Defects in the original concept¶

Computing however is probably one of the most rapidly changing environments of the last decade, and the flaws in the Information Centres soon became apparent.

It has been said of the Information Centre that “never in the field of human endeavour have so many waited so expensively, for so little.” This may indeed be true. It would be naive to believe that a major new concept propounded by a hardware manufacturer was done out of purely philanthropic motives. The underlying motive was undoubtedly to promote the sale of large computer systems, and associated software packages that would perhaps encourage rapid consumption of hardware resources.

Perhaps the single most obvious flaw in the original offerings was the lack of integration between the elements that made up the Information Centre. Indeed, the most cohesive product on offer was probably the marketing brochure itself!

Users were quick to discover the “missing links” of I.C. implementation. These fell into 3 main areas:

• End user terminal hardware
• Communications
• Implementation software

Hardware:¶

Professional DP users were quite happy to utilise bulky complex terminals, generally from a single supplier, and heavily oriented towards use by computer programmers. End users however, were demanding flexibility and portability of user terminals, and wanted the ability to take advantage of all manner of user-friendly and application-specific devices, like flatbed plotters, viewdata terminals, and communicating word processors and micros. The Information Centre did little to promote the attachment of anything other than a small range of “standard” devices.

Communications:¶

The early Information Centre tended to assume that all users would be in relatively close proximity to the central mainframe. But organisations were becoming more international in their outlook. Information pertinent to the running of today’s companies is typically highly geographically distributed. Communications facilities available to the Information Centre implementors were woefully inadequate. The network was more often a barrier to communication than an interface between remote sites.

Implementation software:¶

Although “integration” was a watchword, the basic offerings of the Information Centre gave little hope of achieving this. The applications packages, being a collection from different historic origins, were implemented in the whole range of computer languages, and to entirely different standards of design and user interface. So, while the user might be content if his requirements were met entirely by one package, attempts to interface between them, or extend and customise them, were liable to meet with frustration. The implementation backlogs, which had begun to decrease with the arrival of the Information Centre, increased again along with the growth of demands for the ability to properly exploit, or extend the range of, the basic facilities.

The “missing links” manifested themselves in several problem areas:

User Isolation: I.C. users were victims of the ultimate in “Personal Computing”, but unable to communicate with each other, or with each other’s data.

Lack of Portability: Software applications, often with significant development effort, were rarely made available to other parts of the organisation, and “reinventing the wheel” became the rule rather than the exception.

Growth and Migration Problems: Ideas and solutions were typically incapable of growing beyond quite tight constraints, dictated by the implementation mechanisms.

Integration Difficulties: With few standards or guidelines, the “Do It Yourself” approach made any attempts at integration extremely difficult.

Insufficient Tools/Data: The Personal Computing philosophy, by encouraging users to think small, denied them access to the main strategic systems which often contained the most valuable data.

Add to these problems the arrival of the cheap, but very powerful, micro, with highly “user friendly” software packages, and it can be seen that the DP department is once again under attack. These micros can be bought at prices often well within an individual’s unqualified signing approval, and yet they often provide power and sophistication above that achievable on the company’s largest computers.

Towards a better strategy¶

We are very nearly at a stage where it will be possible, if we want, to place on every executive’s desk a user workstation which can provide virtually infinite computing power, together with essentially unlimited memory capacity. But what do we want to do with it? All too often, this fundamental question is left unanswered.

For the end user, the cost of the hardware technology will no longer be any obstacle to the accessibility of computer systems. As computerisation spreads within organisations, there will be increasing demands for provision of new data and information to help the user in making decisions. End user systems will spread and overlap with each other.

The micro itself is NOT the answer to these new problems and opportunities. In a short time, the anarchy which has typified the growth of standalone microcomputer systems will be replaced by a need to re-integrate scattered, and often incompatible, systems. Equally, users will come to realise that their Lotus 1-2-3 or Multiplan systems are not an end in themselves, but should be only front ends to larger corporate databases and systems, since their inputs and outputs belong in, and relate to, other parts of the organisation. The end user does not function in isolation.

The Global Information Centre¶

The Global Information Centre (GIC) takes the basic framework of the Information Centre, addresses its problems, and adds the missing links in the earlier concept. There are several fundamental components which are critical in achieving true acceptability by both end users and DP managers — though each may regard the relative priorities as different.

The GIC assumes that in the typical situation, users may well be scattered around the world; that they will want to use a variety of terminal technology to access corporate systems, including intelligent terminals and micros; that there will be a requirement at a later stage to integrate applications systems, software, and data, even though the eventual needs may not be apparent at the outset; and that the DP department will want to offer a growth and development strategy with new products attractive to end users, while at the same time maintaining central coordinated control over corporate DP development.

The GIC strategy therefore does not prohibit end user acquisition of micros, though it insists that they must be capable of interfacing, at hardware and software levels, into the other elements of the GIC. Far from being seen as a restriction, many users will gain a feeling of comfort from the knowledge that their chosen micro system forms part of the long term strategy in their organisation.

Account should also be taken of Word Processing, automated PABX’s, Local Area Networks, and the full range of office automation technology. They all can, and should, be integrated into the GIC strategy. This does not imply a commitment to a single-vendor policy, but simply a requirement that all selected vendors comply with INTERFACING standards. Careful attention should be paid to a vendor’s commitment to Open System Interconnection (OSI).

The principal missing links which then, should be added to the original definition of the Information Centre are:

• Worldwide Communications
• Corporate AND external public data
• Technology Independence
• INTEGRATION
• CO-ORDINATION

The best examples of Global Information Centre implementation are today probably in the service companies, whose historical development has been along Information Centre lines even before the phrase was coined.

Case Study: I P Sharp Associates¶

I P Sharp has in effect been running a Global Information Centre for some 15 years. It has developed almost entirely driven by client demands, serving users in a variety of organisations and industries who looked outside their own organisations for solutions to their problems.

The key elements of the Sharp GIC are as follows:

WORLDWIDE COMMUNICATIONS NETWORK: A private packet-switched network, developed internally in the absence of suitable available technology. A principal design goal was that it should be invisible to the end user, yet flexible in terms of the technology supported. Also interfaced to public networks, it provides local call access from some 600 cities worldwide.

SINGLE IMPLEMENTATION LANGUAGE: Utilising only SHARP APL, a proprietary enhancement to APL, all applications are implemented in the same language. Hence maintenance, integration, and interfacing of packages becomes relatively straightforward. The APL language (James Martin has said it is the only serious contender for the title of Fifth Generation Language) provides an extremely powerful user-friendly environment, with great attractions for systems programmers and end users alike. Applications, simple or complex, can be written quickly and elegantly — and can be easily modified when required.

INTEGRATED APPLICATIONS PACKAGES: Although a vast suite of packages and building blocks is available, a kernel of applications are seen as essential for the GIC:

• Financial Consolidation
• Financial Modelling
• Generalised Record Management
• Time Series Database Maintenance and Analysis
• Business Graphics
• Electronic Mail

— and these applications all readily interface to each other.

DATABASES: While the basis of the corporate Information Centre is very much an organisation’s own private data, an increasingly common demand is arising for external data. The I P Sharp GIC offers some 50 million time series in areas like Finance, Economics, Energy, Aviation, Demography, and others. These databases can all be accessed using the same tools as are used for private data, and data from several sources, private and public, can be combined.

IBM PC INTEGRATION: An identical implementation of the Sharp APL environment on the IBM PC, together with software to interconnect mainframes to PC’s allows applications and data to migrate from one to the other, and also permits interfacing to other PC applications.

SUPPORT and EDUCATION: Undoubtedly, the key to any successful DP strategy lies in the level of support and education available to the end user. The GIC user is encouraged to utilise the services of some 60 local branch offices around the world, where he can be assured, by virtue of the GIC structure, that the staff are acquainted with the user’s applications (and can help him in his native language!).

The DP department too, requires support, but at a system level. This is provided 24 hours a day by staff specialising in, and using, precisely the same software environment as on the client machine.

A Strategy for the future¶

Whether following the I P Sharp philosophy, or that of another vendor, the fundamental issues remain the same. End users are increasingly becoming unconcerned with the technical implementation issues, whilst their demands continue to increase. On the one hand, they will say they want more “simplicity” in systems they use, but at the same time they expect much more sophistication.

The need is therefore very clearly for an implementation infrastructure that meets the needs of both the DP department and of the end user, even though their goals often conflict.

The Information Centre concept recognised that the end user was becoming more important in the computing environment. There can be no doubt that this trend is continuing, and so it is evident that the major issue of the next decade will be that of software development, with the end user as the main driving force.

The Global Information Centre, and the principles which underly it, do, it is suggested, offer a viable strategy for the future.