Getting the most out of your software is something all developers strive for, and concurrency, and the art of concurrent programming, happens to be one of the best ways in order for you to improve the performance of your applications. Through the careful application of concurrent concepts into our previously single-threaded applications, we can start to realize the full power of our underlying hardware, and strive to solve problems that were unsolvable in days gone past.
With concurrency, we are able to improve the perceived performance of our applications by concurrently dealing with requests, and updating the frontend instead of just hanging until the backend task is complete. Gone are the days of unresponsive programs that give you no indication as to whether they've crashed or are still silently working.
This improvement in the performance of our applications comes at a heavy price though. By choosing to implement systems in a concurrent fashion, we typically see an increase in the overall complexity of our code, and a heightened risk for bugs to appear within this new code. In order to successfully implement concurrent systems, we must first understand some of the key concurrency primitives and concepts at a deeper level in order to ensure that our applications are safe from these new inherent threats.
In this chapter, I'll be covering some of the fundamental topics that every programmer needs to know before going on to develop concurrent software systems. This includes the following:
- A brief history of concurrency
- Threads and how multithreading works
- Processes and multiprocessing
- The basics of event-driven, reactive, and GPU-based programming
- A few examples to demonstrate the power of concurrency in simple programs
- The limitations of Python when it comes to programming concurrent systems
