Claude Henri Pesquet,
Engineering Group Manager,
International Systems Engineering

In the late 1970s, Digital began to ship its first office products outside the U.S. We realized then that it was no longer an option to provide users with the ability to input, view, edit, and print foreign language text; it was instead a necessity, as well as a passport for Digital into world markets.

The foreign-language requirement came as a shock to the application developers who had been trained in the late 1960s, at a time when the U.S. English-speaking market represented more than 70 percent of the total worldwide information technology market. Today's reality is quite different. The English-speaking IT market is below 40 percent, and trends indicate that it will continue to decline. This is not surprising, because only 8.41 percent of the world's population is native English speaking. Moreover, each year the commoditization of computers lowers the entry point for the acquisition of computer products; consequently the market is expanding to encompass a much broader socioeconomic community. Further, starting in the 1980s, the creation of global markets -- for labor, materials, intellectual talent, financing, and distribution channels -- has forced businesses to continually reach outside their domestic markets. Government mandates also have an impact, requiring that products sold within country boundaries have local-language capability. Together these factors will increase the demand for and requirements of international products -- products that will provide users with linguistic choices.

In recent years, Digital has broadened its market focus to include not only the scientific/technical, mainly English-speaking markets, but also commercial markets -- a large market comprising many languages. To serve these markets well, we are compelled to adopt a strategy for the internationalization* of our products.

The strategy, i.e., to develop products that "speak" the local language, has evolved from a fastidious reengineering of a product after the fact to an architectural definition that ensures products are designed originally to meet local-language requirements. Digital had three goals:

• Reduce development costs.
• Shorten the time to market.
• Increase product quality.

The cost of reengineering products that were designed based upon a North American paradigm is similar to the cost of maintaining an application that was designed without regard to future maintenance. Such costs could meet, if not exceed, the original product development cost. This was discouraging, because the markets outside the U.S. were smaller and emerging; producing the local product compared in cost to producing the original U.S. product. It became obvious that it was too expensive to continually rebuild products that sold only to a small market.

Local-language products were late to market when compared with availability of the same products in the U.S. This presented a twofold problem. It denied our multinational customers the capability of installing products and applications simultaneously in their worldwide operations. Further, product launches, training, selling, support, and retirement had to be addressed one country at a time because specific local-language components were not available simultaneously.

In addressing short-term "surface" issues, we had utilized the brute force of reengineering to produce one language version at a time. As a result, we delayed addressing the "deep" quality issue of originally designing and building into our products the internationalization features that would allow for easy adaptation to any language without modification of a product's core.

A vision on how to address the internationalization of products was developed by a worldwide team of architects led by Gayn Winters. The majority of this team was located outside the U.S. and had been closely involved in Digital's reengineering efforts for many years. The team's prime motivation was to eliminate the need for reengineering. The vision they developed is one in which all Digital integrated systems can process electronic information containing multiple languages and character sets, and can satisfy end-user linguistic preferences. An inherent part of this vision is to make all systems available simultaneously worldwide.

One of the major difficulties in implementing the vision was that internationalization was not aimed at specific products, rather it was a pervasive attribute required across systems. For product development groups trained to develop components, this represented a difficult change in mind-set. The implementation also required a huge paradigm shift --

From one character and... To one character and... One input method Many input methods One cell Multiple cells One collation point Several collation points One geometry Many geometries Alphabet Ideograms "Frozen" alphabet User-defined characters

The paradigm shift led to a redefinition of the elements to be incorporated in the basic design of new products. The strategy from a product perspective was to start with the base system (CPU, peripherals, network, and operating system), and then move to the application side. From an engineering-resource perspective, we would start with parallel internationalization development, and then inject internationalization expertise into the original product development group. The strategy from a process perspective was to customize code for specific countries, and then roll back the country-specific code into the original product code base and continue future development from this unique code base. The implementation has resulted in major achievements, for example, the simultaneous shipment of products to which this approach was applied.

To illustrate our progress, the latest version of Rdb (relational database application) was developed with the injection of internationalization expertise. The approach resulted in one common code base and achieved worldwide simultaneous shipment.

Many challenges remain. Standards have to be defined and implemented in areas such as naming conventions, user profiles, and character attributes. Emerging technologies such as object-oriented software and multimedia need to be addressed. And real-time multilinguality (the simultaneous translation from one language to another) must be tackled.

This issue of the Journal provides a broad sampling of our product internationalization efforts -- from the concept of cultural differences to the specific internationalization of our Rdb product. The papers herein represent only a few of the hundreds of projects dedicated to the internationalization of Digital's products.

* The term internationalization as it is used within the context of this Journal includes both the technologies and the processes applied to enable a product to meet the need of any local-language market without requiring modification of the base functionality of the product.