Section 2.4. Business Computing


2.4. Business Computing

Although the evolution of programming languages and distributed computing eventually provided the technical concepts that today are the first cornerstone of Service-Oriented Architectures, it is equally important to look at the developments in business computing that provided the content that represents the second cornerstone of Service-Oriented Architectures: business data and business logic.

The history of computing was always closely related to solving business problems, starting as early as 1940 with computers being used as large-scale calculators and as a replacement for large filing cabinets. Functions that nowadays are considered "technical" provided immediate business value in the advent of computing.

The following decades created further levels of abstraction, making it easier to think of a computer in terms of a provider for business services. However, business computing maintained its focus on mainframe computers. Most software was custom-built, originally written in machine language and later in functional languages such as COBOL or FORTRAN. Yet business computing proved a crucial success factor for enterprises. For example, logistics companies used computers to compute routes for shipments through their vast international transport networks. Retail giant Wal-Mart was among the first to create custom-made supply-chain management systems to optimize the purchase and distribution of goods.

In the 1970s, the original homegrown filing cabinet applications were gradually replaced using fully fledged database systems, including relational databases, which encapsulate the storage of complex interrelated data. Although today we regard the storage mechanism itself as a technical concept, it seemed fairly business-oriented when it was created. In fact, SQL was developed as a language that was intended to be used mainly by business analysts, not database programmers.

In 1972, four former IBM employees founded SAP in Germany, an event that marked a milestone for business computing. SAP introduced R/2 in 1981, the first business-computing platform that enabled enterprise-wide real time processing of financial data and resource planning information.

However, by the mid 1980s, corporate software development seemingly reached saturation level. Most companies were convinced that they had achieved most of what was possible with computing in their environment. College students were discouraged from majoring in software development because people assumed that only maintenance would be needed in the future, which probably would be performed in some remote offshore location.

Then computers began to reach employee desktops. Systems such as the Commodore PET and others introduced a new concept to computing. Data was obtained from remote storage, while computation and visualization were performed locally. This paradigm was boosted by the huge success of the IBM PC, which was launched in 1984. Through several stages of hardware and software development at both the client and server sides, the client/server paradigm held steady. The focus for advancement shifted back and forth between the two. On one hand, client computers became more powerful, sporting graphical user interfaces and raw computational power, and networks became faster. On the other hand, database vendors worked hard to add value to their systems by providing fault tolerance, scalability, and load balancing using cluster techniques and procedures that were executed within the database.

Driven by an increasing economical globalization and new manufacturing models such as Just-in-Time production, the supply and distribution chains of companies became increasingly sophisticated, relying more on complex IT systems to manage these processes. The software market reacted to this newly awakened demand in enterprise computing by developing complex enterprise applications, such as Enterprise Resources Planning (ERP) and Supply Chain Management (SCM). Over two decades, the market for enterprise solutions has become increasingly complex, offering applications ranging from Customer Relationship Management and Product Lifecycle Management to highly specialized applications, such as applications that manage complex transportation networks or billing systems for telecommunications companies. As a result, a plethora of new enterprise software companies emerged, and old ones grew even bigger, including SAP, Siebel, Oracle, PeopleSoft, J.D. Edwards, Baan, Manugistics, and others.

With the emergence of the Internet, many of these companies transformed their enterprise applications to make use of the new technology (e.g., mySAP) and to cater for new business models in the area of eCommerce and B2B. New companies, such as salesforce.com, emerged even after the heydays of the Internet were over, taking on established companies such as Siebel by delivering its applications entirely over the Internet.

However, the availability of new enterprise software often caused as many problems as it aimed to solve. Often, ERP, CRM, and Enterprise Portal solutions competed within a single organization to become the central repositories of business data and processes. Keeping in mind that as late as 1990, most enterprise IT systems served no more than a single department, it comes as no surprise that many challenges emerged on the organizational level along with a huge demand for Enterprise Application Integration (EAI). However, suffering from problems similar to the Enterprise Software Bus model, the EAI-typical hub-and-spoke model was often limited to solve the problems of individual application integration projects, failing to deliver a more holistic view to the problems of having to integrate across organizational boundaries.

An enterprise SOA aims to leverage the business logic and data that resides in the many applications, databases, and legacy systems of today's enterprises, offering them a sustainable strategy for flexibly adapting their IT systems to changes in functional requirements and business strategy without imposing onto them the use of a single middleware or EAI platform. The services exposed in an SOA are intended to be mapping directly to business entities, thus enabling enterprise integration on the business level, not the technical level.



    Enterprise SOA. Service-Oriented Architecture Best Practices
    Enterprise SOA: Service-Oriented Architecture Best Practices
    ISBN: 0131465759
    EAN: 2147483647
    Year: 2003
    Pages: 142

    flylib.com © 2008-2017.
    If you may any questions please contact us: flylib@qtcs.net