Two important challenges for educators of software engineering are to ensure the relevance of what they teach for the job the students will eventually get, and to ensure that they teach fundamental skills that go beyond, for instance, the syntax of a particular programming language. These twin challenges often work at cross purposes. For instance, should extra time be devoted to ensuring that a student has a solid grounding in algorithms, or should the extra time be devoted to a large programming project? The former would provide the student with a basis of knowledge that he may never again have the opportunity to learn; the latter will enable him to hit the ground running on his next job. Should the student be given a solid grounding in the principles of programming languages, or should he be taught the syntax of three different programming languages? The former may keep him versatile for life, as languages come and go, but the latter will ensure that he gets his next job! These trade-offs are difficult, and judging by the results, there is a tendency to classify this as a choice between the theoretical and the practical, which to our mind, makes matters worse. In this talk we will examine these trade-offs and offer our perspective.

The Kerberos authentication system is based on the trusted third-party Needham-Schroeder (1978) authentication protocol. The system is one of the few industry standards for authentication systems and its use is becoming fairly widespread. The system has some limitations, many of which are traceable to the decision of the Kerberos designers to solely use symmetric key cryptosystems. Using asymmetric (public-key) cryptosystems in an authentication protocol would prevent some of the shortcomings. Several such protocols have been proposed and some have been implemented. However, all these designs are either completely different from the Kerberos system, or require major changes to the basic system. Any attempts to improve Kerberos would do so with only minimal impact to the protocol and the source tree. In this work, we describe Yaksha, a new approach to achieving these goals. Yaksha uses as its building block an RSA (Rivest, Shamir & Adelman, 1978) algorithm variant independently invented by Boyd (1989) and by Ganesan and Yacobi (1994), in which the RSA private key is split into two portions. One portion becomes a user's Yaksha password, and the other the Yaksha server's password for that user. Using this simple but useful primitive, we show how we can blend the Kerberos system with a public-key infrastructure to create Yaksha, a more secure version of Kerberos, with minimal changes to the protocol.>

Interleaved memories are essential in pipelined computers to attain high memory bandwidth. As a memory bank is accessed, a reservation is placed on the bank for the duration of the memory cycle, which is often considerably longer than the processor cycle time. This additional parameter, namely, the bank reservation time or the bank busy time, adds to the complexity of the memory model. For Markov models, exact solutions are not feasible even without this additional parameter due to the very large state space of the Markov chain. The authors develop a Markov model which explicitly tracks the bank reservation time. Because only one processor and the requested bank are modeled, the transition probabilities are not known and have to be approximated. The performance predicted by the model is in close agreement with simulation results.>