Dean Wampler - Programming Scala, 3rd Edition

Programming Scala, Second Edition was published six years ago, in the fall of 2014. At that time, interest in Scala was surging, driven by two factors.

First, alternative languages for the JVM instead of Java were very appealing. Javas evolution had slowed, in part because its steward, Sun Microsystems was sold to Oracle Corporation a few years previously. Developers wanted improvements like more concise syntax for some constructs and features they saw in other languages, like support for functional programming.

Second, big data was a hot sector of the software industry and some of the most popular tools in that sector, especially Apache Spark and Apache Kafka, where written in Scala and offered elegant Scala APIs, along with APIs in other languages.

A lot has changed in six years. Oracle deserves a lot of credit for reinvigorating Java after the Sun Microsystems acquisition. The pace of innovation has improved considerably. Java 8 was a ground-breaking release, as it introduced two of the most important improvements needed to address limitations compared to Scala. One was support for anonymous functions, called lambdas, which addressed the biggest missing feature needed for functional programming. The second feature was support for default implementations of the methods declared in interfaces, which made Java interfaces more useful as composable mixins.

Also, the Kotlin language was created by the tool vendor Jet Brains, as a better Java that isnt as sophisticated as Scala. Kotlin received a big boost when Google endorsed it as the preferred language for Android apps. Around the same time, Apple introduced a language called Swift primarily for iOS development that has a very Scala-like syntax, although it does not target the JVM.

Big data drove the emergence of data science as a profession. Actually, this was just a rebranding and refinement of what data analysts and statisticians had been doing for years. The specialties of data science called deep learning (DL - i.e., using neural networks), reinforcement learning (RL), and artificial intelligence (AI) are currently the hottest topics in the data world. All fit under the umbrella of machine learning (ML). A large percentage of the popular tools for data science, especially ML, are written in Python or expose Python APIs on top of C++ kernels. As a result, interest in Python is growing strongly again, while Scalas growth in the data world has slowed.