Database Reference
In-Depth Information
and, more recently, web pages themselves as the JavaScript engine built
into every web browser has become sufficiently powerful to implement rich
applications.
Although it lost on the client side, Java did find a niche on the server side of
web applications because it's easier to deploy than its primary competitor,
C++. It was particularly easy to deploy when database connections were
involved, having introduced the Java Database Connectivity (JDBC)
standard fairly early in its life. Most web applications are, essentially,
attractive connections to a database, and Java became a common choice of
server implementation language as a result.
The development of a staggering number of software libraries and the slow
but steady improvements in performance made Java a natural choice when
many of the first Big Data applications, especially Hadoop, were being
developed. (Google's internal Map-Reduce application, however, was
written in C++.) It made sense to develop the rest of the stack in Java
because it was on the front end and on the back end in the form of Hadoop.
Because most of the packages in this topic were written in Java, or at least
written to run on the Java Virtual Machine (JVM), Java is used as the
example language for most of this topic. The reasons for this decision are
largely the same as those used to implement the packages in the first place.
The Java language has a variety of software libraries that are fairly easy to
obtain thanks to its similarly wide array of build management systems.
In particular, this topic uses a build system called Maven to handle the
management of packages and their dependencies. Maven is nice because
it provides the ability for distributors of software packages to make them
available either through Maven's own central repository or through
repositories that library authors maintain on their own. These repositories
maintain the software itself and also describe the dependencies of that
software package. Maven can then obtain all the appropriate software
required to build the package automatically.
Scala and Clojure
In the last section it was mentioned that most of the software was built in
Java or to “run on the Java Virtual Machine.” This is because some of the
software was written in languages that compile to run on the JVM and can
interact with Java software packages but are not actually Java.
