https://ir.unisa.ac.za/handle/10500/24002 Wed, 06 May 2026 12:46:23 GMT 2026-05-06T12:46:23Z https://ir.unisa.ac.za/handle/10500/24049 Restructuring of the Conceptual Schema to produce DBMS Schemata Wulf, S. An overall methodology for database design is presented. This includes creation of a data dictionary and relational analysis. An algorithm is presented to create a conceptual schema or logical database design from the set of third normal form relations. The primary emphasis of this paper is to present algorithms for restructuring the conceptual schema to create first pass designs (FPO) for particular DBMS. These FPDs can then be used as the basis for iterative physical database design and performance predictions. Examples are given for creating FPDs for TOTAL and IMS. These can be extended for any other DBMS. Indications are given for future research in extending the methodology. Sat, 01 Jan 1983 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/24049 1983-01-01T00:00:00Z https://ir.unisa.ac.za/handle/10500/24048 Managing and Documenting 10-20 Man Year Projects Visser, P. This paper presents a summary of tutorial material on accepted management techniques and philosophies as applied to the development of large software systems, based upon the documentation for such systems. Sat, 01 Jan 1983 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/24048 1983-01-01T00:00:00Z https://ir.unisa.ac.za/handle/10500/24047 Die Operasionele Enkelbedienermodel Van Niekerk, J.C. The operational approach to the analysis of the single server model is discusses. Firstly Buzen's operational equivalent to the stochastic birth/death Markov model is presented, whereafter generalisations of the model are examined. Lastly, the performance indicators are derived for these models. Sat, 01 Jan 1983 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/24047 1983-01-01T00:00:00Z https://ir.unisa.ac.za/handle/10500/24046 Data Structure Traces Schach, Stephen R. Three levels of traces for data structures (as opposed to simple variables) are defined. A machine-code core dump is essentially a low level trace. A high level trace reflects the high level language in which the data structure being traced has been implemented. A very high level trace displays the data structure in the format in which the programmer conceptualizes it. Three traces written by the author (a graphical FORTRAN array trace, a portable trace for the Pascal heap, and a graphical Pascal data structure trace) are described, and the level of each trace is then analyzed. Sat, 01 Jan 1983 00:00:00 GMT https://ir.unisa.ac.za/handle/10500/24046 1983-01-01T00:00:00Z