Background

Global Perspective

The availability of appropriate software has been a major factor in the diffusion of information technology into almost all aspects of the current environment. Cane (1992) classified software and its utilization as follows:

• Embedded software for the electronic products and telecommunications industry

• Customized software for commercial and financial sectors and for government services

• Packaged software, the driving force behind many personal computers employed in all areas of human endeavour in today's information age

The acquisition of computers for use in business was concentrated primarily in large corporations where centralized computing was successfully implemented. As technological improvements came in the form of mini computers and later PCs, small and medium sized businesses and eventually homes were able to afford to acquire computers. Three market segments for software development could be identified:

• Large corporations that favoured the building of proprietary systems from scratch using basic languages such as COBOL to provide systems for the management of their operations. Maintenance and upgrading of the legacy systems constituted an important activity.

• Small and medium sized businesses for which packaged software was ideally suited. Application software has included spreadsheets, accounting and other management packages, which were bought and then adapted to the special needs of the business.

• The personal computer market for which off-the-shelf products and user-friendly programs mainly on CD-ROMs were mandatory.

This assessment of the market for software globally reflects the overall situation that prevailed up until the start of the Internet era. Web-based business opportunities have for the past five years been sought by corporations and small businesses alike. In this radically new environment, software applications and services utilize easy to use graphical web browsers with new companies constantly seeking to network their businesses.

For those which seek a mere presence on the web all that is required after the page has been designed, possibly by the client, is the integration of the graphics and sound. To achieve greater benefit there will have to be the integration of the web page with the databases and the legacy systems. Successful entry into the electronic commerce environment necessitates a complex information technology environment. There is, according to Hoch et al., a coexistence challenge as one seeks greater advantages from the web technologies. Several platforms, programming languages and strong de facto standards coexist and often must be managed in parallel. The high level of complexity in the IT and communications architectures along with variable commitment to standards in software production have led to difficulties in the management of today's IT infrastructure (Hoch, 2000).

The industry has become quite fluid and unstable with the frequent entry and demise of companies and constant search for innovation among the leading players. The ever-present threat of viruses and worms and the persistence of hackers have not helped the management of current systems.

Another challenge relates to the business skills competence of current software developers in this networked environment. A decade ago software engineers would most likely have had experience in the development of information systems for businesses prior to their involvement in software for export. They would have been quite knowledgeable about business applications. Today's software developers brought up in this networked and object-oriented environment and fast-tracked into dealing with web-based applications are less likely to be familiar with basic business practices.

This trend poses difficulties when, in electronic commerce applications, web-based interfaces to legacy systems have to be developed and maintained. Such challenges must be addressed in the education and training of current software developers. In particular greater emphasis on business skills and project management should be reflected in the respective training programmes.

Domestic Market

In both Jamaica and Barbados, businesses entered the computer era using in-house software and with expertise and experiences gained moved on to the acquisition of packaged software that had to be customized. U.S. software corporations mainly supplied these packages, although companies throughout the Caribbean have acquired a few local packages. In a few cases major businesses, such as a utility company in Barbados, which have successfully implemented computer applications now feel ready to acquire best practice software globally and modify their procedures to fit the software.

The banking sector has been a major user of information technology with the internationally affiliated banks using the same systems as found in their respective head offices. The indigenous banks have attempted to innovate with their systems which are developed by local or international software producers.

In the public sector, which is a major user of information technology the specialized nature of their information systems has meant that, until recently, most systems had to be developed in-house. Local software developers and others from North America have also been contracted to develop application packages, mainly payroll and national insurance systems, for government departments. With growing demand for better information provision expected in an electronic government environment much interest is being shown in Intranets and Internets.

One particular area where a local demand for software is expected in the near future in both countries is in the education sector as both have committed themselves to greater implementation of educational technology. Over a decade ago, a few companies were developing courseware for the North American education system but they have all ceased operations, mainly due to the high cost of production. It is uncertain how much of the available software can be described as being culturally relevant or desirable or whether domestic production of courseware will be required. Any production of courseware for the local environment would have to be preceded by substantial training in an area where there is already much concern about current quality. Unfortunately, the focus has been on the purchase of hardware with considerably more complex questions of software evaluation and acquisition being given minimum attention.

Software Engineering Process

The software engineering process consists of various phases and has, over the past decades, been modified in a continuing effort to arrive at a more efficient model. However there is commonality among the various approaches as described in the following five phases by Reichgelt (2000):

• Problem definition where the problem to be solved is formulated as accurately as possible and in terms relevant to organizational needs.

• Requirements analysis where the requirements of the software are described.

• Specifications phase where the detailed specifications of what the software will look like and how it will meet the requirements are outlined.

• Programming phase where those specifications are converted into program code.

• Maintenance phase where bugs detected during operational use of software are fixed and modifications are introduced to improve and optimise the software's functionality.

The first three phases, representing analysis and design, are deemed to require higher skills and often reflect subjectivity including cultural values of the organization and its environment. The other two phases, representing programming, are seen as requiring relatively lower skills in that a prescribed algorithmic outline has to be coded in the language of choice.

The maintenance phase, in spite of the continuing enhancement efforts including automation of the software process, remains a very labor intensive one. As applications have moved from the earlier accounting ones with specific and static processes to modern online ones incorporating complex decision making aspects, it has proven very difficult to cater for these multiple and dynamic requirements. As a result software maintenance accounts for over 50 percent of the programming effort for most software projects and developing countries have benefited from the outsourcing of this maintenance work.

The pattern has therefore been that the analysis and design for software projects have tended to be confined to the developed countries. Reasons for this pattern include:

• A desire to incorporate cultural affinity and values that can best be ensured by utilizing IT professionals from the same developed countries.

• A lack of trust in unknown and inexperienced IT professionals from developing countries whose work ethic has unfortunately been seen as not being up to the same standard of professionals in developed countries. There is, however, ample evidence from companies in the Information Services Sector to negate this suggestion.

In the competitive global software environment, critical project management skills are needed to complete projects on time, within budget and to the required quality assurance level. Programming, on the other hand, is a relatively more prescribed and objective activity where competence in the particular hardware and software environments are mandatory. This competence can more readily be acquired from education and training in academic and professional institutions. Universities in developing countries after years of insistence on traditional Computer Science programs have modified their offerings and are now able to provide basic and even advanced programming instruction suited for the professional software development. In addition initiatives in the form of short, high-impact and focussed training have been undertaken recently by the governments of Jamaica and Barbados to enhance the professional programming competence in these countries (Bishop, 2001).

Off-Shore Development

Yourdon (1994) was one of the earliest to recognise the potential of developing countries for producing world-class software and possibly becoming a threat to the American programmer. He identified the following growth stages for foreign software industries:

• Body shopping with offshore programmers working on-site in U.S. corporations;

• On-site systems analysis and design done in the U.S. with programmers working offshore to develop that software;

• Offshore development of generic software products to be marketed in the U.S. and in other developed countries; and

• Global marketing of indigenous applications developed offshore. Egypt with its multimedia cataloguing of its historical and cultural artefacts and Singapore with its port management systems have been successful in this area.

India, with its large numbers of trained scientists and programmers, has been the only developing country capable of promoting body shopping as a development strategy. Most developing countries have had to be content with utilizing their relatively cheap labor resources to provide programming support in collaboration with partners in the developed countries.

Typically the applications which are outsourced have tended to be of a financial nature and are based on on-line, web-based and multimedia technologies. In many developing countries, only a few applications, for example in the banking sector, depend heavily on these technologies. Programming standards and practices vary greatly in the two markets with world-class expectations for the export market. Another key requirement would be the existence of an appropriate legal and regulatory environment to safeguard against piracy and copyright infringements.

Global Software Outsourcing

Global software outsourcing is the outsourcing of software development to subcontractors or developers who operate away from the client's home country. Clients tend to be mainly U.S. corporations but may originate from Europe or even Asia. India can claim to be the original developer and, with over 800 IT service firms, has dominated the market, its share being worth almost $4 billion (U.S.) in 1999/2000 (Heeks, 2001). However almost 50 countries now participate as developers with Ireland, Israel and Singapore being among the more successful ones. In the Caribbean Jamaica and Barbados have been the leaders in this pursuit.

Global software outsourcing has been undertaken by the clients for two main strategic reasons, namely cost considerations and access to a larger pool of competent professionals. This has meant that business relationships have had to be established and maintained between these clients in the developed countries and the developers who might be, physically and otherwise, poles apart. Various studies have been undertaken to evaluate the strategies for achieving the most satisfactory relationships.

Global software outsourcing has been fraught with difficulties, chief of which are communication between partners and coordination of the various tasks in the software development process. Distance between the partners can play a major role in exacerbating the problem. Modern communications technologies, it has been claimed, can remove physical or geographical distance as a constraining factor. However experience has shown that, with respect to communication between business partners, video-conferencing and other technologies cannot substitute for actual face-to-face contact. Carmel et al. have identified, in addition to physical and temporal distance, national, cultural and organizational differences as factors that reduce the effectiveness in coordinating the several tasks (Carmel, 2001).

Developers in the Caribbean clearly have great advantages with North American clients in having English as the spoken language and through close cultural and business ties. To capitalize on this Caribbean countries would be advised to market themselves as "near-shore" rather than offshore developers.

Strategies for Involvement

Schware's Strategy

Schware's "walking on two legs" proposition has been advocated as a development strategy for developing countries wishing to become involved. The production of software for domestic markets, he argues, provides opportunities for domestic companies to develop their skills, capabilities and possibly market niches. This could result in the establishment of successful consultancies. The transition to software for export can then be approached with greater confidence and expertise gained from the domestic production effort (Schware, 1992).

There is great variability between developing countries according to their size, stage of economic and technological development and such factors. Thus, the application of Schware's proposition to any particular developing country would depend on the similarity between its domestic applications and those being demanded by corporations in developed countries.

Indigenisation

Heeks (1992) posits that developing countries need to indigenise their software production efforts for the following reasons:

• To avoid reliance on the whims and fancies of foreign suppliers and operators through the creation of technological awareness.

• To allow adaptation of the technology for local conditions. Certain software could be adapted to local languages and scripts.

• To ensure the development of a local capability at both the low- and high-skilled aspects of software production.

Here Heeks is emphasizing hands-on experience for workers rather mere involvement in training sessions. A representative measure of a country's indigenous IT industry development is its technological capability, which he defines as the general ability to undertake a range of tasks that involve the consumption and production of any particular technology. These tasks are:

• Choice and use of a technology

• Adaptation without production, that is, product modification and minor product innovation for local needs

• Basic production, that is, copying and assembly using existing products and processes

• Inward-looking skilled production, that is, modification, redesign or innovation to meet local needs

• Outward-looking skilled production, that is, modification, redesign and innovation to meet global needs

• Global production to compete through innovation

Technological development in a country then is the growth in its technological capability as evidenced by movement up the various levels listed earlier. When applied to IT production or more specifically, to software production, this quantitative measurement of indigenous development is similar to the Software Engineering Index developed at Carnegie-Mellon University which is currently regarded as the industry's standard.

Small developing countries cannot attain the technological development of larger ones such as India and thus have to set for themselves more modest and realistic goals in their involvement in software for export. There has therefore been a reliance on multinationals or collaboration with foreign institutions and organizations to develop their local capability. Governments have offered incentives such as tax concessions and subsidized office space to foreign investors, sometimes in excess to those available to local IT companies. This reliance on multinationals and other foreign investors has proven to provide mixed results. Appropriate advice in this regard should be:

• Approach foreign investors with caution since they bring hidden costs as well as benefits.

• Monitor and evaluate the role of foreign investors.

• Assist local technological development through negotiation with these investors to provide transfer of technology or training as condition of their access to government incentives.