I am studying patterns and anti-patterns. I have a clear idea about patterns, but I don't get anti-patterns. Definitions from the web and Wikipedia confuse me a lot.
Can anybody explain to me in simple words what an anti-pattern is? What is the purpose? What do they do? Is it a bad thing or good thing?
Anti-patterns are certain patterns in software development that are considered bad programming practices.
As opposed to design patterns which are common approaches to common problems which have been formalized and are generally considered a good development practice, anti-patterns are the opposite and are undesirable.
For example, in object-oriented programming, the idea is to separate the software into small pieces called objects. An anti-pattern in object-oriented programming is a God object which performs a lot of functions which would be better separated into different objects.
For example:
class GodObject {
function PerformInitialization() {}
function ReadFromFile() {}
function WriteToFile() {}
function DisplayToScreen() {}
function PerformCalculation() {}
function ValidateInput() {}
// and so on... //
}
The example above has an object that does everything. In object-oriented programming, it would be preferable to have well-defined responsibilities for different objects to keep the code less coupled and ultimately more maintainable:
class FileInputOutput {
function ReadFromFile() {}
function WriteToFile() {}
}
class UserInputOutput {
function DisplayToScreen() {}
function ValidateInput() {}
}
class Logic {
function PerformInitialization() {}
function PerformCalculation() {}
}
The bottom line is there are good ways to develop software with commonly used patterns (design patterns), but there are also ways software is developed and implemented which can lead to problems. Patterns that are considered bad software development practices are anti-patterns.
Whenever I hear about Anti-patterns, I recollect another term viz. Design smell.
"Design smells are certain structures in the design that indicate violation of fundamental design principles and negatively impact design quality”. (From "Refactoring for Software Design Smells: Managing technical debt")
There are many design smells classified based on violating design principles:
Abstraction smells
Missing Abstraction: This smell arises when clumps of data or encoded strings are used instead of creating a class or an interface.
Imperative Abstraction: This smell arises when an operation is turned into a class.
Incomplete Abstraction: This smell arises when an abstraction does not support complementary or interrelated methods completely.
Multifaceted Abstraction: This smell arises when an abstraction has more than one responsibility assigned to it.
Unnecessary Abstraction: This smell occurs when an abstraction that is actually not needed (and thus could have been avoided) gets introduced in a software design.
Unutilized Abstraction: This smell arises when an abstraction is left unused (either not directly used or not reachable).
Duplicate Abstraction: This smell arises when two or more abstractions have identical names or identical implementation or both.
Encapsulation smells
Deficient Encapsulation: This smell occurs when the declared accessibility of one or more members of an abstraction is more permissive than actually required.
Leaky Encapsulation: This smell arises when an abstraction “exposes” or “leaks” implementation details through its public interface.
Missing Encapsulation: This smell occurs when implementation variations are not encapsulated within an abstraction or hierarchy.
Unexploited Encapsulation: This smell arises when client code uses explicit type checks (using chained if-else or switch statements that check for the type of the object) instead of exploiting the variation in types already encapsulated within a hierarchy.
Modularization smells
Broken Modularization: This smell arises when data and/or methods that ideally should have been localized into a single abstraction are separated and spread across multiple abstractions.
Insufficient Modularization: This smell arises when an abstraction exists that has not been completely decomposed, and a further decomposition could reduce its size, implementation complexity, or both.
Cyclically-Dependent Modularization: This smell arises when two or more abstractions depend on each other directly or indirectly (creating a tight coupling between the abstractions).
Hub-Like Modularization: This smell arises when an abstraction has dependencies (both incoming and outgoing) with a large number of other abstractions.
Hierarchy smells
Missing Hierarchy: This smell arises when a code segment uses conditional logic (typically in conjunction with “tagged types”) to explicitly manage variation in behavior where a hierarchy could have been created and used to encapsulate those variations.
Unnecessary Hierarchy: This smell arises when the whole inheritance hierarchy is unnecessary, indicating that inheritance has been applied needlessly for the particular design context.
Unfactored Hierarchy: This smell arises when there is unnecessary duplication among types in a hierarchy.
Wide Hierarchy: This smell arises when an inheritance hierarchy is “too” wide indicating that intermediate types may be missing.
Speculative Hierarchy: This smell arises when one or more types in a hierarchy are provided speculatively (i.e., based on imagined needs rather than real requirements).
Deep Hierarchy: This smell arises when an inheritance hierarchy is “excessively” deep.
Rebellious Hierarchy: This smell arises when a subtype rejects the methods provided by its supertype(s).
Broken Hierarchy: This smell arises when a supertype and its subtype conceptually do not share an “IS- A” relationship resulting in broken substitutability.
Multipath Hierarchy: This smell arises when a subtype inherits both directly as well as indirectly from a supertype leading to unnecessary inheritance paths in the hierarchy.
Cyclic Hierarchy: This smell arises when a supertype in a hierarchy depends on any of its subtypes.
The above definition and classification is described in "Refactoring for software design smells: Managing technical debt". Some more relevant resources could be found here.
A pattern is an idea of how to solve a problem of some class. An anti-pattern is an idea of how not to solve it because implementing that idea would result in bad design.
An example: a "pattern" would be to use a function for code reuse, an "anti-pattern" would be to use copy-paste for the same. Both solve the same problem, but using a function usually leads to more readable and maintainable code than copy-paste.
If you really wish to study AntiPatterns, get the book AntiPatterns (ISBN-13: 978-0471197133).
In it, they define "An AntiPattern is a literary form that describes a commonly occurring solution to a problem that generates decidedly negative consequences."
So, if it's a bad programming practice but not a common one— very limited in frequency of occurrence, it does not meet the "Pattern" part of the AntiPattern definition.
Today, software engineering researchers and practitioners often use the terms “anti-pattern” and “smell” interchangeably. However, they are conceptually not the same. The Wikipedia entry of anti-pattern states that an anti-pattern is different from a bad practice or a bad idea by at least two factors. An anti-pattern is
"A commonly used process, structure or pattern of action that despite initially appearing to be an appropriate and effective response to a problem, typically has more bad consequences than beneficial results.”
It clearly indicates that an anti-pattern is chosen in the belief that it is a good solution (as a pattern) to the presented problem; however, it brings more liabilities than benefits. On the other hand, a smell is simply a bad practice that negatively affects the quality of a software system. For example, Singleton is an anti-pattern and God class (or Insufficient Modularization) is a design smell.
It is sometimes used when you misuse design patterns in an illegal way, or you don't know the actual usage of it. For example, having builder pattern for simple classes, or obsessively defining one singleton instance for each Active class you use in your code.
Also it might be beyond design patterns. For example, defining local variables in Java as final, or using try / catch for NullPointerException when you could simply check input against being null, or nulling objects after they are used (like what you do in some other languages) that you don't notice about garbage collection mechanism, or calling system.gc() to make memory empty, and many other misunderstandings, which are quite likely to be considered as Cargo-Cult phenomena.
In Microservices based area:
Everything is micro except for data is an anti-pattern. It means that if every thing is decomposed reasonable and fully based on DevOps and CI/CD. Maybe some distributed design patterns are in place and even fully replicated, but there is one giant data store behind all of the services, so it is still a monolithic data structure.
Any design pattern that is doing more harm than good to the given software development environment would be considered as anti-pattern.
Some anti-pattern are obvious but some are not. For example Singleton, even though many consider it good old design pattern but there are others who don't.
You can check question What is so bad about singletons? to better understand the different opinions on it.
