Developing Security Systems in the Real World

Abstract

The debut of Java technology in 1995 was a significant event for the computer security field. First, the claim that ”Java is secure” immediately attracted the intense scrutiny by the computer security research community. Numerous security bugs were later found (and fixed). Second, Java and the Internet once again highlighted the issue of mobile code security and the need for a comprehensive solution - the original ”sandbox” security model for Java is not sophiscated enough for many applications that Java programmers would like to develop. All these put tremendous pressure on the subsequent commercial releases of the Java Development Kit, where any new security solution must be technically sound and also meet commercial needs (time to market, backward compatibility, solving a real problem while maintaining simplicity and extensibility, etc.) This talk covers my experience and lessons learned during the development of JDK 1.1 and JDK 1.2 that might be useful to those who are bridging the gap between academic research and commercial product development.