TCP/IP First-Step [Electronic resources] نسخه متنی

How Do You Not Keep a Secret?

Having explored the ins and outs of open versus proprietary architectures, it should be fairly obvious why the benefits of openness have proven so compelling.

The next logical question is this: How do you develop a technology without keeping it secret? That sounds funnyusually keeping a secret is difficult. As you see in this section, keeping a secret can be child's play in comparison to the complex process by which new open technologies are hammered out in political committees.

Open Architectures Equals Open Standards

Every open technology is based on an open standard. An open standard is simply one not kept secret. That open standard might have been developed privately and then introduced publicly, but the result is the same: It is a public technology.reference model.

Reference Models: Keeping Things Organized

One of the simplest, yet most confusing, concepts to understand is a reference model. A reference model is a simple tool. In essence, this framework enables you to dissect a complicated function into a series of individual steps and keep those tasks organized relative to each other. A common framework enables different people or companies to collaborate even if they aren't working directly with each other. That's a powerful and valuable capability!

When you start dissecting a function into its individual components, one thing becomes remarkably clear: Those tasks usually have to occur in the same sequence for things to work right. Using a rather mundane and not necessarily technical example to illustrate this point, consider driving a car. The first step is gaining entry to the inside of the vehicle. Usually that's done by unlocking the door, but not always. Some people leave their doors unlocked, or it might be a top-down convertible whose driver can hop inside. (Alternatively, an entire profession is dedicated to gaining entrance to cars without having to rely on keys, but that's another issue altogether!) The point is that entering a vehicleregardless of how you achieve that taskis necessarily the first step in the function of driving.

The next steps are sitting in the driver's seat, buckling up, and making sure your seat and mirrors are properly adjusted. The sequence of these events doesn't leave a lot of room for debate. You could, for example, adjust your mirrors before adjusting your seat, but then you'd probably have to readjust the mirrors after getting the seat in the right position.

Thus, these tasks have a natural or logical sequence. A more subtle point is that they can be lumped together and generally regarded as a set of related functions. Their relationship is that people do these things to prepare to drive. For the sake of example, call this set of related functions Personal Preparations. The next set of functions comes immediately after completing Personal Preparations.

Start visualizing your reference model for driving a car. Figure 3-1 shows the Personal Preparations in their logical order.

Despite completing Personal Preparations, you're still not quite ready to drive. Take your preparations a bit further and actually engage the mechanical systems that let you drive the car. Again, letting logic guide you, the next thing to do is start the car. Thus, inserting the key in the ignition and turning it should result in the engine starting. Release the parking brake and put the car into gear!

These steps should be separate and distinct from Physical Preparations because they represent a higher level of readiness: You are preparing the car's mechanical systems for operation. Call this group of functions Mechanical Preparations.

Figure 3-2 shows the reference model with Personal Preparations and Mechanical Preparations. At a glance, you should see how they relate to each other and why you want to complete all Personal Preparations before starting Mechanical Preparations.

After making yourself comfortable behind the steering wheel and engaging (or disengaging) all the necessary mechanical systems, you are finally ready to drive! Rather than try to deconstruct the act of driving into discrete tasks, take one last look at your fictitious reference model. It shows how driving is yet another higher level of functionality and how that functionality relates to other functions.

Driving can be thought of as the application and the reason you're bothering with both Personal and Mechanical Preparations. The relationship between these tiers of functions is depicted in Figure 3-3.

Note that the logic works in two directions. This model is based on the logical steps needed to start and drive a car. It works in reverse, too. The model shows the logical sequence of events in which you would stop driving a car. You might not need all the same steps, but the sequencing of steps you do need remains valid, as does the grouping around layers of functions.

For example, you stop driving (which means stop the car), disengage the mechanical systems (first the transmission, then the engine), and deal with Personal Preparations (unbuckle your seatbelt and get out of the car). This important concept takes on greater meaning when you look at reference models for data communications later in this chapter.

Although this rather simple example demonstrates the concept of a dividing a complex function into a logically arranged sequence of individual tasks, it's important to recognize that no standards body governs this function! The layered sequence of tasks might be a de facto standard, but that's logic and common sense at work, not a political committee.

The next section looks at some real reference models that are indispensable when it comes to understanding open standard networking technologies and the standards bodies that developed them.