Guest Commentary: Encapsulation as Year 2000 Alternative

The 20th century is drawing to a close, faster as each week passes, and with time the choices available in handling the century change are ever more limited. Among the available options there is one that may prove to be the difference between business failure and survival during the century transition.

The basic choice facing companies today is whether to upgrade their applications or cope with the difficulties posed by systems that are unable to handle the century change. In some quarters this will mean re-implementation of paper systems long since put aside which are unable to deal with the volume of transactions required by a modern competitive business.

The alternative is a technology which amends the business systems to handle the year 2000 without the huge financial and human resource investments normally associated with large IT projects. This technology, commonly known as encapsulation, defers the century change by downdating all dates in a database by a fixed number of years, and updating these dates wherever they are displayed or printed.

The key problem with the traditional approaches to the year 2000 issue is the effect the change has on the core logic of the application. In implementing an expansion or windowed approach, there is signification effort required to review, change and correct program logic. This is logic that has been written by a number of programmers often without documentation or adequate quality control. Given the normal performance of IT projects this makes it more unlikely that the project can be completed on schedule let alone on time.

Encapsulation on the other handle does not impact logic and as such requires far less manual review and correction. It is also less likely that errors in the conversion will have a significant effect on the business as the changes are largely cosmetic and may be quickly corrected or removed.

There are two major advantages in considering encapsulation as an approach to the year 2000. Firstly it is technically very simple and can be implemented in a very short period of time, and secondly it defers the upgrade or replacement of systems to a time when it is more appropriate to the business in general.

The process of applying encapsulation as a Year 2000 remedy is also better defined and requires less manual effort than the alternative approaches. Encapsulation is also totally reversible and has no impact on existing code or database structures.

In creating testing scenarios, encapsulation does not require the system dates to be changed as the system date is automatically downdated at each insertion point. This results in a shorted test cycle as well as reduced costs in hardware, whether leased or purchased.

Considering all the alternatives, encapsulation my prove to be the only solution left for those companies that are about to hit or have already passed their Year 2000 critical horizon. Even those who do not plan to be year 2000 ready until the first or second quarter of ’99 may not be able to complete their conversion project and should consider encapsulation as a viable contingency plan.

Encapsulation is a simple but ingenious solution to the year 2000 without incurring the costs and risks of a major application development project. It is already in use by a large number of businesses throughout the world, was implemented at a fraction of the cost of any of the alternatives and it works. With a little under 18 months to the year 2000 and an ever dwindling pool of skilled IT resources available, time is running out. You need act now to identify a solution that can be implemented and tested on time and on budget and the options available to you are becoming more and more limited as time passes.

Brendan O'Connell has been working with the IBM mid-range systems for 16 years. He has been involved in the development and implementation of a number of major systems throughout the world, and worked with most of the leading AS/400 package solutions available. For the last two years he has specialised on year 2000 projects with particular emphasis in the implementation of Y2K compliant solutions.