Logic Oriented Program Synthesis - goals and realization.

Christoph Kreitz, Gerd Neugebauer, Bertram Fronhöfer.

Forschungsberichte Künstliche Intelligenz FKI-90-88, Technische Universität München, 1988.


Automated Program Synthesis from logical specifications nowadays has to face attacks coming from two directions. "Real" Programmers often argue that it is a nice academic toy, good to generate a handful of small examples, but of no use at all in the hard real world of software technology. On the other hand, due to the advent of Logic Programming the distinction between specification language and program language got blurred and some people believe that Program Synthesis has become an obsolete field of research. In view of the existing papers and systems both opinions appear to be quite natural. Therefore it has to be clarified where the real great challenges to Program Synthesis are today.
Our paper intends to open the discussion on the topic again by expounding a view of the field which arose from experiences with implementing a Logic Oriented Program Synthesizer (LOPS). After stating what the tasks of a Program Synthesis System should be we will give a methodological guideline for the practical realization of such a system. By following our suggestions, we believe, both attacks against the field can be countered successfully. Firstly, Program Synthesis will indeed have useful applications in industry and secondly, as we will show, Logic Programming languages are not at all a solution to the problems of the field.

Back to overview of papers

Bibtex Entry

@Techreport{tr:KreitzNeugebauerFronhofer88a, author = "Christoph Kreitz and Gerd Neugebauer and Bertram Fronh{\"o}fer", title = "{Logic Oriented Program Synthesis} - goals and realization", institution = "Technische Universit{\"a}t M{\"u}nchen, Institut f{\"u}r Mathematik und Informatik" type = "Forschungsberichte K{\"u}nstliche Intelligenz", year = 1988, number = "FKI-90-88" }