When to use OOP over procedural coding

Object-Oriented Programming

Commonly used because it allows developers to work on the same project without colliding with one another as easily. The video covered many of the advantages of OOP already, but let me outline

Advantages include:

• Modular: Provides a clear modular structure for programs which makes it good for defining abstract datatypes where implementation details are hidden and the unit has a clearly defined interface.
• Scaleable: Adding developers to a project often easier because the don't have to understand the entire code base, just the section they're working on. Adding Hardware resources can be more cost efective because you can have different resources for each modular piece.
• Maintainable: Makes it easy to maintain and modify existing code as new objects can be created with small differences to existing ones.
• Extensible: Provides a good framework for extending a project through libraries where these components can be easily adapted and modified by the programmer. This is particularly useful for developing graphical user interfaces, or GUIs.
• Reusable: Each module works independently of the surrounding module. This allows you take take one section, such as the user login, and use it for other projects.

Disadvantages include:

• The real world refuses to divide up into neat classes and subclasses.
• Sometimes several objects will interact in complex ways - maybe even ways we didn't necessarily anticipate when writing the program.
• Can require more code and be more complicated for smaller projects or project in which there are very few repeated tasks.
• Decreased performance. This is one of the most heated debates. Although a well designed procedural site CAN slightly out perform a well designed object-oriented site, there are many factors to consider and this should not be your main concern.

OOP is often the best use when:

• You have multiple programmers who don’t need to understand each component.
• There is a lot of code that could be shared and reused.
• The project is anticipated to change often and be added to over time.
• Different sections can benefit from different resources like datasource or hardware.

You probably don't what to use it when you have several developers who need to share implementation specifics, for example if you were writing a keyboard driver, you wouldn't want to break it into fragments which would hide implementation specifics from a developer working on the driver.

Procedural Programming

Revolves around keeping code as concise as possible, and focuses on a very specific end result.

Advantages include:

• It is written in a step-by-step function, smaller programs written this way are very easy to follow.
• Easy to maintain, as each procedure or function can be debugged in isolation
• Since it is written for a very specific purpose the code often gets you extremely efficient and high-performance applications.
• Code can be more flexible as you can change a specific procedure that gets implemented across the program

Disadvantages include:

• Procedural coding tends to get very difficult to maintain the larger the code gets.
• As programs grow in size, the temptation arises to start using more and more global variable. This can introduce sometimes-devastating results, particularly in a large program worked on by multiple programmers.
• Portions of the code are so interdependent that the code in one application will not be useable in another, meaning despite being somewhat similar the code for one program will not able to be carried to a new one, which OOP can do.
• Procedural code is difficult to relate with real world objects.
• Modifying one part of the code requires modification of the entire code
• As code grows, it gets harder to understand and modify
• Hard to apply code from one program to another

Procedural Programming is often the best choice when:

• There is a complex operation which includes dependencies between operations and a need for clear visibility of different application states ('SQL loading', 'SQL loaded', 'Network online', 'No audio hardware', etc). This is usually appropriate for application startup and shutdown.
• The program is very unique and few elements were shared.
• The program is static and not expected to change much over time.
• None or only a few features are expected to be added to the project over time.

You probably don't what to use it when there are many simple independent tasks to perform or to manage user interface.

Summary

In short, when it comes to short programs, Procedural Programming can be faster and easier both to maintain and to read.

When it comes to large and complex modular projects, being worked on by multiple programmers or even teams, using OOP lets you compartmentalize and organize, again making it faster and easier both to maintain and to read.

PS: There is also the debate about whether a program is object-oriented just because it uses objects... @ircmaxell has a good post