Modeling Paradigms

Modeling Paradigms

MODEL-DRIVEN DESIGN calls for an implementation technology in tune with the particular modeling paradigm being applied. Many such paradigms have been experimented with, but only a few have been widely used in practice. At present, the dominant paradigm is object-oriented design, and most complex projects these days set out to use objects. This predominance has come about for a variety of reasons: some factors are intrinsic to objects, some are circumstantial, and others derive from the advantages that come from wide usage itself.

Why the Object Paradigm Predominates

Many of the reasons teams choose the object paradigm are not technical, or even intrinsic to objects. But right out of the gate, object modeling does strike a nice balance of simplicity and sophistication.

If a modeling paradigm is too esoteric, not enough developers will master it, and they will use it badly. If the nontechnical members of the team can't grasp at least the rudiments of the paradigm, they will not understand the model, and the UBIQUITOUS LANGUAGE will be lost. The fundamentals of object-oriented design seem to come naturally to most people. Although some developers miss the subtleties of modeling, even nontechnologists can follow a diagram of an object model.

Yet, simple as the concept of object modeling is, it has proven rich enough to capture important domain knowledge. And it has been supported from the outset by development tools that allowed a model to be expressed in software.

Today, the object paradigm also has some significant circumstantial advantages deriving from maturity and widespread adoption. Without mature infrastructure and tool support, a project can get sidetracked into technological R&D, delaying and diverting resources away from application development and introducing technical risks. Some technologies don't play well with others, and it may not be possible to integrate them with industry-standard solutions, forcing the team to reinvent common utilities. But over the years, many of these problems have been solved for objects, or made irrelevant by widespread adoption. (Now it falls on other approaches to integrate with mainstream object technology.) Most new technologies provide the means to integrate with the popular object-oriented platforms. This makes integration easier and even leaves open the option of mixing in subsystems based on other modeling paradigms (which we will discuss later in this chapter).

Equally important is the maturity of the developer community and the design culture itself. A project that adopts a novel paradigm may be unable to find developers with expertise in the technology, or with the experience to create effective models in the chosen paradigm. It may not be feasible to educate developers in a reasonable amount of time because the patterns for making the most of the paradigm and technology haven't gelled yet. Perhaps the pioneers of the field are effective but haven't yet published their insights in an accessible form.

Objects are already understood by a community of thousands of developers, project managers, and all the other specialists involved in project work.

A story from an object-oriented project of only a decade ago illustrates the risks of working in an immature paradigm. In the early 1990s, this project committed itself to several cutting-edge technologies, including use of an object-oriented database on a large scale. It was exciting. People on the team would proudly tell visitors that we were deploying the biggest database this technology had ever supported. When I joined the project, different teams were spinning out object-oriented designs and storing their objects in the database effortlessly. But gradually the realization crept upon us that we were beginning to absorb a significant fraction of the database's capacity—with test data! The actual database would be dozens of times larger. The actual transaction volume would be dozens of times higher. Was it impossible to use this technology for this application? Had we used it improperly? We were out of our depth.

Fortunately, we were able to bring onto the team one of a handful of people in the world with the skills to extricate us from the problem. He named his price and we paid it. There were three sources of the problem. First, the off-the-shelf infrastructure provided with the database simply didn't scale up to our needs. Second, storage of fine-grained objects turned out to be much more costly than we had realized. Third, parts of the object model had such a tangle of interdependencies that contention became a problem with a relatively small number of concurrent transactions.

With the help of this hired expert, we enhanced the infrastructure. The team, now aware of the impact of fine-grained objects, began to find models that worked better with this technology. All of us deepened our thinking about the importance of limiting the web of relationships in a model, and we began applying this new understanding to making better models with more decoupling between closely interrelated aggregates.

Several months were lost in this recovery, in addition to the earlier months spent going down a failed path. And this had not been the team's first setback resulting from the immaturity of the chosen technologies and our own lack of experience with the associated learning curve. Sadly, this project eventually retrenched and became quite conservative. To this day they use the exotic technologies, but for cautiously scoped applications that probably don't really benefit from them.

A decade later, object-oriented technology is relatively mature. Most common infrastructure needs can be met with off-the-shelf solutions that have been used in the field. Mission-critical tools come from major vendors, often multiple vendors, or from stable open-source projects. Many of these infrastructure pieces themselves are used widely enough that there is a base of people who already understand them, as well as books explaining them, and so forth. The limitations of these established technologies are fairly well understood, so that knowledgeable teams are less likely to overreach.

Other interesting modeling paradigms just don't have this maturity. Some are too hard to master and will never be used outside small specialties. Others have potential, but the technical infrastructure is still patchy or shaky, and few people understand the subtleties of creating good models for them. These may come of age, but they are not ready for most projects.

This is why, for the present, most projects attempting MODEL-DRIVEN DESIGN are wise to use object-oriented technology as the core of their system. They will not be locked into an object-only system—because objects have become the mainstream of the industry, integration tools are available to connect with almost any other technology in current use.

Yet this doesn't mean that people should restrict themselves to objects forever. Traveling with the crowd provides some safety, but it isn't always the way to go. Object models address a large number of practical software problems, but there are domains that are not natural to model as discrete packets of encapsulated behavior. For example, domains that are intensely mathematical or that are dominated by global logical reasoning do not fit well into the object-oriented paradigm.

Nonobjects in an Object World

A domain model does not have to be an object model. There are MODEL-DRIVEN DESIGNS implemented in Prolog, for example, with a model made up of logical rules and facts. Model paradigms have been conceived to address certain ways people like to think about domains. Then the models of those domains are shaped by the paradigm. The result is a model that conforms to the paradigm so that it can be effectively implemented in the tools that support that modeling style.

Whatever the dominant model paradigm may be on a project, there are bound to be parts of the domain that would be much easier to express in some other paradigm. When there are just a few anomalous elements of a domain that otherwise works well in a paradigm, developers can live with a few awkward objects in an otherwise consistent model. (Or, on the other extreme, if the greater part of the problem domain is more naturally expressed in a particular other paradigm, it may make sense to switch paradigms altogether and choose a different implementation platform.) But when major parts of the domain seem to belong to different paradigms, it is intellectually appealing to model each part in a paradigm that fits, using a mixture of tool sets to support implementation. When the interdependence is small, a subsystem in the other paradigm can be encapsulated, such as a complex math calculation that simply needs to be called by an object. Other times the different aspects are more intertwined, such as when the interaction of the objects depends on some mathematical relationships.

This is what motivates the integration into object systems of such nonobject components as business rules engines and workflow engines. Mixing paradigms allows developers to model particular concepts in the style that fits best. Furthermore, most systems must use some nonobject technical infrastructure, most commonly relational databases. But making a coherent model that spans paradigms is hard, and making the supporting tools coexist is complicated. When developers can't clearly see a coherent model embodied in the software, MODEL-DRIVEN DESIGN can go out the window, even as this mixture increases the need for it.

Sticking with MODEL-DRIVEN DESIGN When Mixing Paradigms

Rules engines will serve as an example of a technology sometimes mixed into an object-oriented application development project. A knowledge-rich domain model probably contains explicit rules, yet the object paradigm lacks specific semantics for stating rules and their interactions. Although rules can be modeled as objects, and often are successfully, object encapsulation makes it awkward to apply global rules that cross the whole system. Rules engine technology is appealing because it promises to provide a more natural and declarative way to define rules, effectively allowing the rules paradigm to be mixed into the object paradigm. The logic paradigm is well developed and powerful, and it seems like a good complement to the strengths and weaknesses of objects.

But people don't always get what they hope for out of rules engines. Some products just don't work very well. Some lack a seamless view that can show the relatedness of model concepts that run between the two implementation environments. One common outcome is an application fractured in two: a static data storage system using objects, and an ad hoc rules processing application that has lost almost all connection with the object model.

It is important to continue to think in terms of models while working with rules. The team has to find a single model that can work with both implementation paradigms. This is not easy, but it should be possible if the rules engine allows expressive implementation. Otherwise, the data and the rules become unconnected. The rules in the engine end up more like little programs than conceptual rules in the domain model. With tight, clear relationships between the rules and the objects, the meaning of both pieces is retained.

Without a seamless environment, it falls on the developers to distill a model made up of clear, fundamental concepts to hold the whole design together.

The most effective tool for holding the parts together is a robust UBIQUITOUS LANGUAGE that underlies the whole heterogeneous model. Consistently applying names in the two environments and exercising those names in the UBIQUITOUS LANGUAGE can help bridge the gap.

This is a topic that deserves a book of its own. The goal of this section is merely to show that it isn't necessary to give up MODEL-DRIVEN DESIGN, and that it is worth the effort to keep it.

Although a MODEL-DRIVEN DESIGN does not have to be object oriented, it does depend on having an expressive implementation of the model constructs, be they objects, rules, or workflows. If the available tool does not facilitate that expressiveness, reconsider the choice of tools. An unexpressive implementation negates the advantage of the extra paradigm.

Here are four rules of thumb for mixing nonobject elements into a predominantly object-oriented system:

• Don't fight the implementation paradigm. There's always another way to think about a domain. Find model concepts that fit the paradigm.

• Lean on the ubiquitous language. Even when there is no rigorous connection between tools, very consistent use of language can keep parts of the design from diverging.

• Don't get hung up on UML. Sometimes the fixation on a tool, such as UML diagramming, leads people to distort the model to make it fit what can easily be drawn. For example, UML does have some features for representing constraints, but they are not always sufficient. Some other style of drawing (perhaps conventional for the other paradigm), or simple English descriptions, are better than tortuous adaptation of a drawing style intended for a certain view of objects.

• Be skeptical. Is the tool really pulling its weight? Just because you have some rules, that doesn't necessarily mean you need the overhead of a rules engine. Rules can be expressed as objects, perhaps a little less neatly; multiple paradigms complicate matters enormously.

Before taking on the burden of mixed paradigms, the options within the dominant paradigm should be exhausted. Even though some domain concepts don't present themselves as obvious objects, they often can be modeled within the paradigm. Chapter 9 will discuss the modeling of unconventional types of concepts using object technology

The relational paradigm is a special case of paradigm mixing. The most common nonobject technology, the relational database is also more intimately related to the object model than other components, because it acts as the persistent store of the data that makes up the objects themselves. Storing object data in relational databases will be discussed in Chapter 6, along with the many other challenges of the object life cycle.