Client/Server computing has brought about a whole new standard of corporate computing productivity, but at the same time it has introduced many new problems for corporate IT organizations.

The advent of low-cost desktop computers makes Client/Server possible. No longer is it necessary for users to wait even seconds to interact with their corporate data. With their own computer on their desktop they do not need to wait for the mainframe to respond.

All of this cheap computer power has also made it possible to support computation- intensive graphical interfaces, which are much easier for users to understand.

In order for Client/Server to work, the application program which used to reside entirely on the mainframe has been split into two pieces, the client piece and the server piece.

The portion of the application that resides in the Client personal computer includes logic for the presentation of information to the user and mechanisms for accepting user input. It also includes logic for data interaction at the Client, such as changing the appearance of the graphical interface as the user make various choices.

The Client software often includes corporate business rules that allow data to be validated before it is sent to the server.

The portion of the application that remains on the Server is usually just the portion that stores data at a central location, accessible to other users. The once mighty mainframes have been reduced to database servers, while the rest of the application has migrated out to the clients.

This migration has caused many problems. The Client machines have become "fat". As the complexity of applications has risen it has become necessary to supply users with more and more powerful computers with faster processors, more disk storage, and more RAM. It is true that personal computer manufacturers have been able to deliver better and better computers for the same amount of money, but the constant upgrades required are costly for corporations. It takes more people and time to upgrade hundreds or thousands of personal computers than it used to take to just upgrade the mainframes.

Setting aside the cost of Client equipment, the proliferation of corporate applications, including business logic, to hundreds or thousands of machines has become a major cost factor for corporations. Where IT staffs used to be able to maintain all the corporate software assets in one central location, they now need to maintain corporate software spread all across the corporation, often housed in computers that are out of IT control.

There are many estimates available that the cost of Client/Server computing is eight or more times the cost of the equipment alone. Several companies offer large, expensive systems that have the sole purpose of maintaining corporate software spread out all over the corporation.

The issue here is not personal-productivity software such as word processors and spreadsheets. Personal software products have enabled knowledge workers to attain new heights of productivity. The issue is the maintenance of corporate software, such as order-entry and accounting. Imagine the difficulty of making a schema change in a corporate database, and synchronizing that change with the updating of thousands of copies of the corresponding Client software.

It is true that the use of department-level LANs can bring down the number of copies that need to be updated to hundreds, instead of thousands. The size of the LANs must be limited, because the large size of the fat client software requires a responsive LAN or else the users will just copy the enterprise software to their own machines, bringing us back to potentialy thousands of copies to be maintained.

The task of making simultaneous changes to even hundreds of copies of client software is still a daunting one.