Terminal Services has been a part of the Windows server OS since NT 4.0, but its history as a technology goes back further to the late 80s when Ed Iacobucci founded Citrix. Ed had been working with IBM trying to develop a multi-user version of OS/2, but IBM wasn't much interested so Ed left to start his own company (good move). In the early 90s Microsoft came out with Windows NT 3.51, their first really solid 32-bit Windows server platform, but Microsoft too wasn't much interested in developing a multi-user version of their OS at the time, so in 1994 Microsoft granted Citrix access to their NT source code so Citrix could develop this technology as a third-party vendor. Soon afterward Citrix released their revolutionary Win Frame product, and its immediate success soon led Microsoft to shift gears and in 1997 Microsoft licensed Citrix technology so they could incorporate it into NT 4.0. The result was the release in 1998 of Hydra, Microsoft's code name for Windows NT 4.0 Terminal Server Edition. Hydra of course has since been superseded by the built-in Terminal Services feature of Windows 2000 Server and more recently in Windows Server 2003.

And there we are today. But why all the fuss? What good is Terminal Services technology anyway?

The Terminal Services Advantage

From the operations side of running a business, there are really only two things that are important: time and money. The great thing about Terminal Services is that it can save you both.

From a time-saving perspective, Terminal Services lets administrators install, configure, manage and maintain applications centrally on a few servers. This is usually much faster and easier to do than deploying applications on hundreds or thousands of desktop machines at different sites across an enterprise. And by making the job of the administrator easier, companies also save money on IT support costs for their information systems infrastructure. Furthermore, centrally-deployed applications are usually easier to maintain (for example, patching and upgrading) and simpler to troubleshoot when things go wrong. As a result, downtime is reduced, users are more productive, and business booms along.

Another cost-saving perspective is that since in a terminal server environment all application logic runs on the server, the processing and storage requirements for client machines are minimal. This means you can save money by keeping in service older desktop computers running legacy versions of Windows, and focus your limited IT budget on a few high-powered systems to run as your terminal servers. Or you can toss your old desktop PCs and buy thin clients like Windows-Based Terminals (WBTs) instead, such as those from Neoware, Wyse, and other vendors. The options are almost limitless since terminal servers let you run almost any DOS, Win32, or Web-based application from almost any client platform as we'll see in a moment.

So what are the options for running terminal servers in today's enterprise environment? There are two major players in the landscape, Microsoft and Citrix, and we'll start by looking at Microsoft's latest offering, the built-in Terminal Services component of Windows Server 2003.