PROLOG-REQUEST@SUSHI.STANFORD.EDU.UUCP (05/28/87)
PROLOG Digest Thursday, 28 May 1987 Volume 5 : Issue 39 Today's Topics: Query - Test Programs & GNU & Quintus, Announcement - New System Available, LP Library - Tech Report ---------------------------------------------------------------------- Date: 19 May 87 17:34:55 GMT From: Goran Bage <mcvax!enea!ttds!draken!sics!goran@seismo.css.gov> Subject: GC test programs wanted I need a set of programs that I can use to test and evaluate a garbage collector for SICStus Prolog. Please let me know if you have or know of any suitable programs. -- Goran Bage ------------------------------ Date: 20 May 87 03:07:07 GMT From: Michael Gleicher <duke!gleicher@mcnc.org> Subject: GNU and Quintus Prolog Does anyone have a GNU emacs Prolog mode that will run Quintus Prolog in a manner similar to the way that Quintus runs under Unipress ? Thank you. -- Mike Gleicher ------------------------------ Date: Fri, 08 May 87 12:19:01 +1000 From: munnari!mulga.oz!jas@seismo.CSS.GOV Subject: Academic Release of NU-Prolog System Version 1.1 of the NU-Prolog system is now available for release to academic institutions (schools, colleges, universities). NU-Prolog is a second generation Prolog system which incorporates a number of important advances in Logic Programming implementation. NU-Prolog was implemented as part of the Machine Intelligence Project in the Department of Computer Science at the University of Melbourne. It is the successor to Lee Naish's successful MU-Prolog system and attempts to move Prolog closer to the ideals of Logic Programming by allowing the user to program in a style closer to first order logic. In addition, it provides substantial performance gains over interpreted systems such as MU-Prolog. NU-Prolog has the following features: * compiles Prolog programs into machine code for an enhanced version of the Warren abstract machine (implementing the delay/coroutine style of programming of MU-Prolog) * incorporates a database system based on superimposed codeword indexing which can store general Prolog terms in external databases for fast retrieval by NU-Prolog programs; the database system makes use of the superjoin algorithm to perform efficient join operations * uses "when" declarations (the successor to MU-Prolog's "wait") to control the execution of NU-Prolog programs according to the availability of data * implements a large set of built-in predicates, including many Quintus Prolog predicates; most DEC-10/Edinburgh/MU-Prolog library predicates are available through compatibility libraries The NU-Prolog system contains the following major components: * "nc", the NU-Prolog compiler * "np", a simple interpreter-style interface which implements the standard Edinburgh Prolog style debugging facilities and has a sophisticated query language for accessing external database predicates * "nac", a program for adding control information to NU-Prolog programs written in a purely logical style * "nit", a program for reporting common errors in NU-Prolog programs (cf. Unix/C's "lint") NU-Prolog runs under Unix System V and Berkeley BSD Unix 4.?. It has been implemented on the following machines: Elxsi 6400, Vax 11/780, Perkin Elmer 3240, Sun workstations, Pyramid 98x, Integrated Solutions Workstations. The system comes complete with a manual and all source code. The preferred distribution medium is 1/2" tape, Unix tar-format at 1600bpi. There is a A$400.00 fee to cover distribution costs. In order to obtain a copy of the system, you must first complete a licence agreement with the University of Melbourne. Licences can be obtained by contacting: NU-Prolog Distribution Department of Computer Science University of Melbourne Parkville, Victoria, 3052 AUSTRALIA or CSNET: mip%munnari.oz@australia ARPA: mip%munnari.oz@seismo.css.gov UUCP: ...!munnari!mip (maybe, mip@munnari.uucp) ACSnet: mip@munnari.oz The system will be demonstrated at the Fourth International Conference on Logic Progrmaming in Melbourne later in May. The Machine Intelligence Project has been assisted in the development of NU-Prolog by: the Commonwealth Department of Science, the Australian Research Grants Scheme, the University of Melbourne and Pyramid Technology, Aust. ------------------------------ Date: Sun, 10 May 87 17:08:23 PDT From: Mike Newton <newton@vlsi.caltech.edu> Subject: Tech Report The following paper has been submitted to a journal. However, copies of the Caltech TR version are available now. For ordering information, please contact Nancy Zachariasen (nancy@csvax.caltech.edu or 818 356 6704). A High Performance Implementation of Prolog Caltech Computer Science Department, Technical Report TR:5234:86 % abstract We discuss an efficient implementation of the Warren Abstract Machine (WAM) in detail. Special attention is given to data formats, memory layout, WAM optimizations and code generation techniques. A final section describes some hardware considerations for even higher performance execution. Currently the compiler produces code that runs at approximately 900,000 logical inferences per second (LIPS) on a single processor of an IBM 3090 using the naive reverse benchmark. Using several of the yet unimplemented optimizations, we expect this figure to top one million LIPS. -- mike ------------------------------ End of PROLOG Digest ********************