Quarks; Binary Encoding for Dummies

Contrary to widespread popular belief, only a fraction of the data on your disk and network is encoded in angle brackets. In fact, the more interesting things are encoded in a compressed binary way. Now, with XML, there is a wealth of libraries out there to bind documents to your application's object model. With binary encoded data however, you are pretty much left on your own, which is odd, since parsing binary compressed data is in fact considerably harder than parsing a markup language.

Quarks intends to change that. It allows you declaratively bind a Java based object model to its compressed binary encoded representation. It aims to be to binary encoding formats what JAXB is to XML, and Hibernate to relational databases.

Quarks buys you a lot. Simply declaring the mapping between the Java based representation and the encoded representation will give you the decoder, the encoder and even hyperlinked documentation on the encoding format. And it is readable documentation. Not JavaDoc of your classes, but a document that will carefully document every piece of your binary encoding format. , including its dependencies on other pieces. Needless to say that takes away a huge burden from the developer.

Wilfred Springer  Xebia Wilfred Springer works as a consultant and chief scientist (a.k.a. general loud mouth) at Xebia in the Netherlands. Before joining Xebia, he led several teams of developers at TomTom, and worked as a software architect at Sun Microsystems. Wilfred is an active contributor to the open source domain and a Codehaus Despot. Most of his open source involvement is currently contributed to the development of Preon, an open source project aiming to provide a data binding framework for bitstream compressed data. Wilfred is active in a team of people leading the DuSE (Dutch Scala Enthusiasts) and heads the Dutch NoSQL meetup group in Amsterdam.