Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site gitpyr.UUCP
Path: utzoo!linus!gatech!gitpyr!allen
From: allen@gitpyr.UUCP (P. Allen Jensen)
Newsgroups: net.ai,net.lang.prolog
Subject: Specification and the Synthesis of Logic Programs
Message-ID: <663@gitpyr.UUCP>
Date: Sun, 18-Aug-85 14:38:21 EDT
Article-I.D.: gitpyr.663
Posted: Sun Aug 18 14:38:21 1985
Date-Received: Tue, 20-Aug-85 01:06:00 EDT
Organization: Georgia Institute of Technology
Lines: 18
Keywords: specification, synthesis, logic
Xref: linus net.ai:2724 net.lang.prolog:514

It seems to me that using Logic as both a specification language and a
programming language would make the problem of Verification and
program Synthesis eaiser.  Deduction may be used to both derive
programs from their specification and to run those programs.  On the
other hand, wouldn't using Logic for both tend to make the difference
between the specification and the program fuzzy ?

I understand how a specification can be used to verify a program.  This
assumes, however, that the specification is correct and self-consistent.
It would seem that this is just as difficult as writing a correct
program !  How does one verify a specification ?

P. Allen Jensen
-- 
P. Allen Jensen
Department of Civil Engineering
Georgia Insitute of Technology, Atlanta Georgia, 30332
...!{akgua,allegra,amd,hplabs,ihnp4,masscomp,ut-ngp}!gatech!gitpyr!allen