Many problems faced in implementing computer-based system applications are related to the deployment of commercially available general-purpose computer system platforms. In the provisioning application software products and services, the platforms have often proven to be the most unreliable, costly, and risk-laden system element. There are historical reasons for the current problems that must be understood in order to map out a path for a future where stable and secure platforms can be provided as reliable commercially available system elements. This paper extends a model of function distribution between various hardware and software levels from a previously published result (Lawson, "Function distribution in computer system architectures," presented at the 3rd Annu. Symp. Computer Architecture, 1976). After a review of the revised model, related principles are "reapplied" to the current situation. Alternative paths that can lead to stable and secure computer platforms via the rational distribution of functions are presented. Finally, the need for as well as the potential impact of autonomic computing is described

Summary form only given, as follows. Today's computer systems contain significant unnecessary complexity. The unnecessary complexity is due to two factors. Firstly, 'Yeature-itis' (too much functionality), and secondly, the level-wise mapping of application functions - via programming languages, operating systems, protocols and middleware onto poor or inappropriate platforms. The unnecessary complexity leads to significant risks and costs (particularly in maintenance). In this presentation, we focus upon the risks associated with poor and often inappropriate platforms upon which application software (often critical) is dependent. The following aspects will be addressed: The Problem of Platform Stability Historical Perspective (Or - It Did Not Have To Happen This Way) What To Do About The Situation?

BASEMENT/sup TM/ is a distributed real-time architecture developed for vehicle internal use in the automotive industry. BASEMENT covers application development, as well as the hardware and software that provide execution and communication support. This paper gives an overview of the BASEMENT concept, as well as presenting two system realizations. The first realization is based on the commercial real-time kernel Rubus, while the second is an ultra-dependable architecture (DACAPO) with provisions for fault tolerance at various system levels. BASEMENT is designed for the automotive systems of the future. These systems will be required to simultaneously handle multiple safety critical functions and a large number of less critical functions. All of these features are to be provided at a production cost substantially lower than that of current systems, and, at the same time, with a reliability allowing vehicles to be built without mechanical backup systems, even for safety critical subsystems such as braking and steering.

The major concepts of an engineering oriented software methodology to be employed in the development of application functions for multiplexed automotive control systems are presented. The methodology provides for an orderly engineering approach to application function development.>

Abstract The specification of human/machine dialogues for general purpose terminals and special purpose consoles in a natural language is difficult and frequently leads to ambiguities and misinterpretations between terminal and dialogue constructors and users. In this paper, the problem of operator console definition is related to the problem of language and language translator definition. An efficient metalanguage tool (TBNF) for the handling of the latter problem is applied to the solution of the first problem. An example from the Telecontrol of a Power Network, being developed at ENHER (Barcelona, Spain), is described as an illustration.

The levelwise structuring and complexity of a computer system is presented informally and as a general model based upon the notion of abstract machines (processors), processes and interpreters. The important domains of the computer architect are considered along with historical perspectives of certain stimulae and decisions that have affected the distribution of functions amongst the various levels of computer system implementations.

This paper describes the functional characteristics of the Standard Computer MLP-900 processor. This microprogrammable processor is designed for general-purpose emulation of a wide variety of known processors as well as being adaptable to new general-purpose and special-purpose applications. Two approaches to the design of microprogrammed processors, namely minimally encoded and highly encoded, are considered. The significant architectural properties that provide the multilingual capability are discussed. A target language-execution strategy for the MLP-900 is presented followed by a problem-oriented target instruction set which provides for packed field operations. Finally, some projected uses of the processor are considered.

Abstract—The possible introduction of changeable control memories in fourth-generation computers may color the manner in which programming systems are designed and implemented. With this type of machine, we will be able to implement varieties of instruction sets by writing microprograms to interpret them. The trade-offs involved in deciding what part of a system is to be handled by the hardware and what part is to be handled by software will provide a challenge for future systems designers and implementers.