Will this be the year of the fat network?

Regular IS Survivalists will recall this term. Introduced last year as a more accurate alternative to “thin client” (which originally referred to the presentation module of a well-designed n-tier application, but which has, through misuse, lost all of its meaning) fat network architectures come in three flavors:

• Windows terminals, in which client modules that range from slender to obese execute on servers, relegating desktop systems to manage keystrokes, mouse clicks, and the display.
• Network computers (NCs), which download applications stored on servers to the desktop for execution.
• Applications that use a browser as their user interface platform, either in vanilla form or enhanced through the use of some kind of plug-in.

I call them fat network architectures because the only thing they have in common is the need for bigger, more reliable servers and (usually) more network bandwidth than applications designed to be stored on and executed from local hard drives.Fat network architectures (do the translation in your head every time you read “thin client”) won’t hit the jackpot this year … or if they do, something is seriously wrong with how IS sets its priorities. Here’s why:

First of all, fat network architectures place a premium on sturdy, high-performance distributed systems. From my conversations with several specialists in the enterprise systems management discipline, the state of distributed systems management in most data centers is deficient. Even if moving to a fat network architecture is your top priority, you’ll do it next year — this year, you’ll focus on stabilizing your servers and managing them better.

Secondly, most CEOs won’t give you the time, budget and resources for a major architectural change this year. They just spent a king’s ransom on infrastructure in the form of Y2K remediation, often delaying new business initiatives as a result. Now, they need you to focus on these business initiatives, and the new applications and application integration needed to support them. With Y2K, the CIO dictated the company’s IS priorities. Now it’s time for a different balance.

Yes, you can build the new stuff using one of the three fat network architectures. This strategy is less risky than wholesale conversion anyway, so more power to you. If you choose a fat network technique that replaces PCs with something else, though, make sure you first inventory the applications used by affected employees to do their jobs today. Otherwise, you could accidentally cripple them when you roll out the new stuff.

Oh, and don’t forget to fortify your distributed systems.

The third reason this won’t be the year of the fat network is perhaps the most intriguing. The personal computer is more than a “client” these days (remembering that in the device sense, “client” means “device hosting client processes” just as “server” means “device hosting server processes”). The personal computer is now a server as well — it has become a personal information hub, storing and synchronizing with an expanding array of information appliances. Examples?

• The personal digital assistance (PDA) is the most visible information appliance these days. Millions of employees rely on them where they used to rely on a paper day planner for their calendar, task management, and address book.
• MP3 players will find uses beyond downloading music. Whether for “talking books” or executive briefings, information managed by the PC and downloaded to an MP3 player can be a productive alternative to the dangerous practice of conducting business via cell phone while commuting.
• Speaking of books, how about electronic ones? The PC becomes the device that manages the user’s library, downloading books of current interest into the electronic book reader as needed. Start thinking about how your business could use electronic books instead of printed paper — the possibilities are stupendous.

Years of bushwah about “total cost of ownership” have conditioned a lot of IS managers to think of the PC as a liability to be minimized. In reality, it’s exactly the opposite.It’s an asset, ready and waiting to be leveraged.