Luke Cassady-Dorion

A common occurrence on comp.lang. java is a post questioning the ability to create growable data structures in Java. The common belief tends to be that pointers are necessary to implement a growable data structure. This obviously stems from experience with languages like C or Pascal, where one is forced to manipulate pointers to track memory in growable data structures. If you think about the data structures, however, all that is really necessary to implement these systems is the ability to dynamically allocate memory. Dynamic memory allocation, a fundamental part of any modern programming language, is actually much easier to handle in Java than in C or Pascal. This is obviously due, in part, to Java's automatic garbage collection. Not only does Java provide for easier memory management, but it also provides several useful classes and interfaces which aid in the cre... (more)

Java and The Mac

When Java Development tools initially appeared on the market, Mac programmers were some of the last people to have tools available to them. While Sun chose to release the JDK for most popular platforms, it waited quite a while before releasing the MacJDK. A variety of third party developers began to release Java development tools, but they too have decided to wait before releasing tools for the Mac programmer. Fortunately, there are options open to the Mac programmer now, but these options are nowhere near as vast as those options open for users of Windows95 or UNIX platforms. Ge... (more)

Pointer Free Data Structures

Last month, we began looking at building data structures in Java. The idea for the article was inspired by the constant posts to comp.lang.java from people who were lost without pointers. The data structures which we introduced were useful, but were really more of a starting point for some more advanced data structures. The third data structure covered last month was a binary search tree. As you will remember, binary search trees are great for storing large amounts of information, as search time is much faster than for a linked-list. The problem with binary trees, however, is the... (more)