Skip to main content
Article
Resource Bundling for Distributed Computing
Journal of Object Technology (2005)
  • Douglas A. Lyon
Abstract

This paper describes techniques for integrating programs and their resources. The goal is to distribute the programs, to a variety of platforms, without loosing the resources that they need in order to run. Programs so integrated are less fragile than their nonintegrated counterparts. The techniques described include the use of a semi-automatic source code synthesizer, XML-based serialization and a base-64 GZIP encoded string format. The approach is suitable for small data objects (i.e., icons, short audio signals, etc.). It has been used, with good success, on a variety of projects. One drawback of the technique is that an added step is required during program development in order to integrate resources into the code. Another drawback is that integrating resources into the source code can dramatically increase the size of the class files. On the other hand, once the class files are loaded, the resources are available in memory (and hence, quickly accessible). The techniques described are a part of the Kahindu project, a joint project between the skunk works of DocJava, Inc. and Fairfield University. Kahindu is the name of a village in Kenya known for its fine coffee.

Publication Date
Winter January, 2005
Citation Information
Douglas A. Lyon. "Resource Bundling for Distributed Computing" Journal of Object Technology Vol. 4 Iss. 1 (2005)
Available at: http://works.bepress.com/douglas_lyon/61/