From comsat@vpics1 Mon Jun 24 21:59:25 1985 Date: Mon, 24 Jun 85 21:59:18 edt From: comsat@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by csnet-relay.arpa id a001158; 24 Jun 85 18:04 EDT Date: Mon 24 Jun 1985 09:14-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #82 To: AIList@SRI-AI Received: from rand-relay by vpi; Mon, 24 Jun 85 22:44 EST AIList Digest Monday, 24 Jun 1985 Volume 3 : Issue 82 Today's Topics: Queries - VAX Lisp & PC Lisps & McDonnell Douglas NL Breakthrough, Games - Optimal Scrabble, Automata - Predation/Cooperation, Psychology - Common Sense, Analogy - Bibliography, Seminar - Evaluating Expert Forecasts (NASA) ---------------------------------------------------------------------- Date: Mon, 24 Jun 85 07:38:35 EDT From: cugini@NBS-VMS Subject: VAX Lisp Just looking for a little consumer information here - does anyone have any experience with Digital's VAX LISP ? DEC advertises it as a full-fledged implementation of CommonLisp. Any remarks on price, performance, quality, etc are appreciated. John Cugini Institute for Computer Sciences and Technology National Bureau of Standards Bldg 225 Room A-265 Gaithersburg, MD 20899 phone: (301) 921-2431 ------------------------------ Date: Sun 23 Jun 85 15:09:12-EDT From: Jonathan Delatizky Subject: PC Lisps [Forwarded from the MIT bboard by SASW@MIT-MC.] Can some of you out there who have used Lisp implementations on IBM PC type machines give me some recommendations as to the best PC Lisp? I plan to run it on a PC/XT and a PC/AT if possible. Also, any expert systems shells that run on the same machines, real or toy-like. ...jon ------------------------------ Date: 22 Jun 1985 13:20-EST From: George Cross Subject: McDonnell Douglas NL Breakthrough The following is the text of a full page color ad on page 49 in the June 24, 1985 New Yorker. It has also been run in the Wall Street Journal. Does anyone know what the breakthrough is? This was mentioned on the ailist some time ago but I didn't notice a response. There is a photo of a hand holding the chin of smiling boy. BREAKTHROUGH: A COMPUTER THAT UNDERSTANDS YOU LIKE YOUR MOTHER Having to learn letter-perfect software languages can be frustrating to the average person trying to tap the power of a computer. But practical thinkers at our McDonnell Douglas Computer Systems Company have created the first computer that accepts you as you are - human. They emulated the two halves of the brain with two-level software: One level with a dictionary of facts and a second level to interpret them. The resulting Natural Language processor understands everyday conversational English. So it knows what you mean, no matter how you express yourself. It also learns your idiosyncrasies, forgives your errors, and tells you how to find out what you're looking for. Now, virtually anyone who can read and write can use a computer. We're creating breakthroughs not only in Artificial Intelligence but also in health care, space manufacturing and aircraft. We're McDonnell Douglas. How can I learn more? Write P.O. Box 19501 Irvine, CA 92713 ------------------------------ Date: 22 Jun 1985 13:07-EDT From: Jon.Webb@CMU-CS-IUS2.ARPA Subject: Optimal Scrabble Anyone interested in computer Scrabble should be aware that Guy Jacobson and Andrew Appel (some of the people that did Rog-o-matic) have written a program which in some sense solves the problem. Using a clever data structure, their program makes plays in a few seconds and always makes the best possible play. Their dictionary is the official Scrabble dictionary. The program is not completely optimal because it doesn't take into account how the placement of its words near things like triple word scores may help the other player, but in all other senses it always makes the best play. I suppose some simple strategic techniques could be added using a penalty function, but as the program almost always wins anyway, this hasn't been done. It regularly gets bingos (all seven letters used), makes clever plays that create three or more words, and so on. The version they have now runs on Vax/Unix. There was some work to port it to the (Fat) Macintosh but that is not finished, mainly for lack of interest. Jon ------------------------------ Date: Fri, 21 Jun 85 17:17:58 EDT From: David_West%UMich-MTS.Mailnet@MIT-MULTICS.ARPA Subject: Predation/Cooperation (AIL v3 #78) Re: enquiry of sdmartin@bbng about learning cooperation in predation: For an extensive investigation of a minimal-domain model (prisoner's dilemma),see _The Evolution of Co-operation_ (NY: Basic Books, 1984; LC 83-45255, ISBN 0-465-02122-0) by Robert Axelrod (of the U of Mich). He is in the Institute of Public Policy Studies, but one of his more interesting methods was the use of the genetic algorithms of John Holland (also of the U of Mich) to breed automata to have improved strategies for playing Prisoner's dilemma. A one-sentence summary of his results is that cooperation can displace non-cooperation if individuals remember each other's behavior and have a high enough probability of meeting again. An intermediate-length summary can be found in Science _211_ (27 Mar 81) 1390-1396. ------------------------------ Date: Fri 21 Jun 85 19:23:03-PDT From: Calton Pu Subject: definition of common sense I had a discussion with a friend on this exact topic just a few weeks ago. My conclusions can be phrased as an elaboration of V. Pratt's two criteria. 1. common knowledge basis (all facts depended on must be common knowledge) I think the (abstract) common knowledge basis can be more concretely described as "cultural background". Your Formico's Pizza example shows clearly that anybody not familiar with San Francisco will not have the "common sense" to go there. The term "cultural background" admits many levels of interpretation (national, provincial, etc.) so most of REALLY COMMON knowledge basis will be encompassed. 2. low computational complexity (easy to check the conclusion). I think the key here is not the checking (NP), but the finding (P) of the solution. So here I differ from Vaughan, in that I believe common sense is something "obvious" to a lot of people, by their own reasoning power. There are two factors involved: the first is the amount of reasoning power; the second is the amount of deductive processing involved. On the first factor, unfortunately usual words to describe people with the adequate reasoning power such as "sensible", "reasonable", and "objective" have also the connotation of being "emotionless". Let's leave out the emotional aspects and use the term "reasonable" to include everybody who is able to apply elementary logic to normal situations. On the second factor, typical words to picture easy deductive efforts are "obvious", "clear", and "evident". So my definition of common sense is: that which is obvious to a reasonable person with an adequate cultural background. I should point out that the three parameters of common sense, cultural background, reasoning power, and deductive effort, vary from place to place and from person to person. If we agreed more on each other's common sense, it might be easier to negotiate peace. ------------------------------ Date: Monday, 24 Jun 85 01:38:08 EDT From: shrager (jeff shrager) @ cmu-psy-a Subject: Analogy Bibliography [Someone asked for an analogy bibliography a while back. This was compiled about two years (maybe more) ago so it's partial and somewhat out of date, but might serve as a starter for people interested in the topic. I've added a couple of thing just now in looking it over. The focus is primarily psychological, but readers will recognize some of the principle AI work as well. I've got annotations for quite a few of these, but the remarks are quite long and detailed so I won't burden AIList with them. -- Jeff] ANALOGY (A partial bibliography) Compiled by Jeff Shrager CMU Psychology 24 June 1985 (Send recommendations to Shrager@CMU-PSY-A.) Bobrow, D. G. & Winograd, T. (1977). An Overview of KRL: A Knowledge Representation Language. Cognitive Science, 1, 3-46. Bott, R.A. A study of complex learning: Theories and Methodologies. Univ. of Calif. at San Diego, Center for Human Information Processing report No. 7901. Brown, D. (1977). Use of Analogy to Acheive New Experience. Technical Report 403, MIT AI Laboratory. Burstein, M. H. (June, 1983). Concept Formation by Incremental Analogical Reasoning and Debugging. Proceedings of the International Machine Learning Workshop. pp. 19-25. Carbonell, J. G. (August, 1981). A computational model of analogical problem solving. Proceedings of the Seventh International Joint Conference on Artificial Intelligence, Vancouver. pp. 147-152. Carbonell, J.G. (1983). Learning by Analogy: Formulating and Generalizing Plans from Past Experience. In Michalski, R.S., Carbonell, J.G., & Mitchell, T.M. (Ed.), Machine Learning, an Aritificial Intelligence Approach Palo Alto: Tioga Press. Carnap, R. (1963). Variety, analogy and periodicity in inductive logic. Philosophy of Science, 30, 222-227. Darden, L. (June, 1983). Reasoning by Analogy in Scientific Theory Construction. Proceedings of the International Machine Learning Workshop. pp. 32-40. de Kleer, J. & Brown, J.S. Foundations of Envisioning. Xerox PARC report. Douglas, S. A., & Moran, T. P. (August, 1983). Learning operator semantics by analogy. Proceedings of the National Conference on Artificial Intelligence. Douglas, S. A., & Moran, T. P. (December, 1983b). Learning text editor semantics by analogy. Proceedings of the Second Annual Conference on Computer Human Interaction. pp. 207-211. Dunker, K. (1945). On Problem Solving. Psychological Monographs, 58, . Evans, T. G. (1968). A program for the solution of a class of geometric analogy intelligence test questions. In Minsky, M. (Ed.), Semantic Information Processing Cambridge, Mass.: MIT Press. pp. 271-253. Gentner, D. (July, 1980). The Structure of Analogical Models in Science. Report 4451, Bolt Beraneck and Newman. Gentner, D. (1981). Generative Analogies as Mental Models. Proceedings of the 3rd National Cognitive Science Conference. pp. 97-100. Proceedings of the 3rd annual conference. Gentner, D. (1982). Are Scientific Analogies Metaphors? In D. S. Miall (Ed.), Metaphor: Problems and Perspectives New York: Harvester Press Ltd. pp. 106-132. Gentner, D., & Gentner, D. R. (1983). Flowing Waters or Teeming Crowds: Mental Models of Electricity. In Gentner, D. & Stevens, A. L. (Ed.), Mental Models Hillsdale, NJ: Lawrence Earlbaum Associates. pp. 99-129. Gick, M. L. & Holyoak, K. J. (1980). Analogic Problem Solving. Cognitive Psychology, 12, 306-355. Gick, M. L. & Holyoak, K. J. (1983). Schema Induction and Analogic Transfer. Cognitive Psychology, 15, 1-38. Halasz, F. & Moran, T. P. (1982). Analogy Considered Harmful. Proceedings of the Conference on Human Factors in Computer Systems, New York. Hesse, Mary. (1955). Science and the Human Imagination. New York: Philisophical Library. Hesse, Mary. (1974). The Structure of Scientific Inference. Berkeley: Univ. of Calif. Press. Kling, R. E. (1971). A Paradigm for Reasoning by Analogy. Artificial Intelligence, 2, 147-178. Lenat, D.B. & Greiner, R.D. (1980). RLL: A representation language language. Proc. of the first annual meeting. Stanford. McDermott, J. (December, 1978). ANA: An assimilating and accomodatiing production system. Technical Report CMU-CS-78-156, Carnegie-Mellon University. McDermott, J. (1979). Learning to use analogies. Sixth Internation Joint Conference on Artificial Intelligence. Medin, D. L. and Schaffer, M. M. (1978). Context Theory of Classification Learning. Psychological Review, 85(3), 207-238. Minsky, M. (1975). A Framework for Representing Knowledge. In Winston, P.H. (Ed.), The Psychology of Computer Vision New York: McGraw Hill. Minsky, M. (July, 1982). Learning Meaning. Technical Report, . Unpublished MIT AI Lab techinical report. Moore, J. & Newell, A. (1974). How can MERLIN Understand? In L.W.Gregg (Ed.), Knowledge and Cognition Potomic, Md.: Erlbaum Associates. Ortony, A. (1979). Beyond Literal Similarity. Psych Review, 86(3), 161-179. Pirolli, P. & Anderson, J.R. (1985) The role of Learning from Examples in the Acquisition of Recursive Programming Skills. Canadian Journal of Psychology. Vol. 39, no. 4; pgs. 240-272. Polya, G. (1945). How to solve it. Princton, N.J.: Princeton U. Press. Quine, W. V. O. (1960). Word and Object. Cambridge: MIT Press. Reed, S. K., Ernst, G. W., & Banerji, R. (1974). The Role of Analogy in Transfer Between Similar Problem States. Cognitive Psychology, 6, 436-450. Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M., & Boynes-Braem, P. (1976). Basic Objects on Natural Kinds. Cog Psych, 8, 382-439. Ross, B. (1982). Remindings and Their Effects in Learning a Cognitive Skill. PhD thesis, Stanford. Rumelhart, D.E., & Norman, D.A. (?DATE?). Accretion, tuning, and restructuring: Three modes of learning. In R.Klatsky and J.W.Cotton (Eds.), Semantic Factors in Cognition Hillsdale, N.J.: Erlbaum Associates. Rumerlhart, D.E. & Norman, D.A. (1981). Analogical Processes in Learning. In J.R. Anderson (Ed.), Cognitive Skills and Their Acquisition Hillsdale, N.J.: Lawrence Earlbaum Associates. pp. 335-360. Schustack, M., & Anderson, J. R. (1979). Effects of analogy to prior knowledge on memory for new information. Journal of Verbal Learning and Verbal Behavior, 18, 565-583. Sembugamoorthy, V. (August, 1981). Analogy-based acquisition of utterances relating to temporal aspects. Proceedings of the 7th International Joint Conference on Artificial Intelligence. pp. 106-108. Shrager, J. & Klahr, D. (December, 1983). A Model of Learning in the Instructionless Environment. Proceedings of the Conference on Human Factors in Computing Systems. pp. 226-229. Shrager, J. & Klahr, D. Instructionless Learning: Hypothesis Generation and Experimental Performance. In preparation. Sternberg, R. (1977). Intelligence, information processing, and analogical reasoning: The componential analysis of human abilities. Hillsdale, N.J.: Lawrence Erlbaum Associates. VanLehn, K., & Brown, J. S. (1978). Planning nets: A representation for formalizing analogies and semantic models of procedural skills. In Snow, R. E., Frederico, P. A. and Montague, W. E. (Ed.), Aptitude Learning and Instruction: Cognitive Process Analyses Hillsdale, NJ: Lawrence Erlbaum Associates. Weiner, E. J. A Computational Approach to Metaphore Comprehension. In the Penn Review of Linguistics. Winston, P. H. (December, 1980). Learning and Reasoning by Analogy. Communications of the ACM, 23(12), 689-703. Winston, P. H. Learning and Reasoning by Analogy: The details. MIT AI Memo number 520. ------------------------------ Date: Fri, 21 Jun 85 11:42:26 pdt From: gabor!amyjo@RIACS.ARPA (Amy Jo Bilson) Subject: Seminar - Evaluating Expert Forecasts (NASA) NASA PERCEPTION AND COGNITION SEMINARS Who: Keith Levi From: University of Michigan When: 10 am, Tuesday, June 25, 1985 Where: Room 177, Building 239, NASA Ames Research Center What: Evaluating Expert Forecasts Abstract: Probabilistic forecasts, often generated by an expert, are critical to many decision aids and expert systems. The quality of such inputs has usually been evaluated in terms of logical consistency. However, in terms of real-world implications, the external correspondence of probabilistic forecasts is usually much more important than internal consistency. I will discuss recently developed procedures for evaluating external correspondence and present research on the topic. Non-citizens (except permanent residents) must have prior approval from the Directors Office one week in advance. Permanent residents must show Alien Registration Card at the time of registration. To request approval or obtain further information, call 415-694-6584. ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Thu Jun 27 03:22:01 1985 Date: Thu, 27 Jun 85 03:21:57 edt From: comsat@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by csnet-relay.arpa id a003734; 26 Jun 85 2:06 EDT Date: Tue 25 Jun 1985 22:22-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #83 To: AIList@SRI-AI Received: from rand-relay by vpi; Thu, 27 Jun 85 04:43 EST AIList Digest Wednesday, 26 Jun 1985 Volume 3 : Issue 83 Today's Topics: Queries - Lisps for VAX, Book - Logic Programming Text, Seminars - A Situational Theory of Analogy (CSLI) & Implementing Dempster's Rule (SU), Conference - 2nd ACM N.E. Regional Conference ---------------------------------------------------------------------- Date: Tue, 25 Jun 85 06:51:42 EDT From: cugini@NBS-VMS Subject: Lisps for VAX Does anyone have recommendations for incarnations of Lisp to run under VAX/VMS, especially ones with features for object-oriented programming? Is there something called XLISP which fits this description, and if so, where does it live? Thanks for any help. John Cugini Institute for Computer Sciences and Technology National Bureau of Standards Bldg 225 Room A-265 Gaithersburg, MD 20899 phone: (301) 921-2431 ------------------------------ Date: Tue, 25 Jun 85 09:02:59 mdt From: cib%f@LANL.ARPA (C.I. Browne) Subject: Common Lisp on VAX/UNIX (Query) We would be most grateful for pointers to a source of Common Lisp for a VAX 11/780 running under UNIX 4.2bsd. Thank you. cib cib@lanl cib@lanl.arpa ------------------------------ Date: 22 Jun 85 1842 PDT From: Yoni Malachi Subject: Logic Programming Text [Excerpted from the Prolog Digest by Laws@SRI-AI. The original contained a lengthy abstract for each section of the book; to get a copy, FTP file logicprog.txt on SRI-AI, or write to me at AIList-Request@SRI-AI.ARPA. -- KIL] LOGIC PROGRAMMING: RELATIONS, FUNCTIONS, AND EQUATIONS Doug DeGroot Gary Lindstrom Editors Prentice-Hall, Inc. Publication date: Summer 1985 June 14, 1985 1. Concept This book addresses the topical and rapidly developing areas of logic, functional, and equational programming, with special emphasis on their relationships and prospects for fruitful amalgamation. A distinguished set of researchers have contributed fourteen articles addressing this field from a wide variety of perspectives. The book will be approximately 500 pages, published in hard cover form, with foreword by the editors and combined index. 2. Table of Contents 2.1. Setting the Stage - Uday Reddy: On the Relationship between Logic and Functional Languages (34 pp.). - J. Darlington, A.J. Field, and H. Pull: The Unification of Functional and Logic Languages (34 pp.). 2.2. Unification and Functional Programming - Harvey Abramson: A Prological Definition of HASL, a Purely Functional Language with Unification Based Conditional Binding Expressions (57 pp.). - M. Sato and T. Sakurai: QUTE: a Functional Language Based on Unification (24 pp.). - P.A. Subrahmanyam and J.-H. You: FUNLOG: a Computational Model Integrating Logic Programming and Functional Programming (42 pp.). 2.3. Symmetric Combinations - R. Barbuti, M. Bellia, G. Levi, and M. Martelli: LEAF: a Language which Integrates Logic, Equations and Functions (33 pp.). - Shimon Cohen: The APPLOG Language (38 pp.). 2.4. Programming with Equality - Wm. Kornfeld: Equality for Prolog (15 pp.). - Joseph Goguen and Jose Meseguer: EQLOG: Equality, Types, and Generic Modules for Logic Programming (69 pp). - Y. Malachi, Z. Manna and R. Waldinger: TABLOG: a New Approach to Logic Programming (30 pp.). 2.5. Augmented Unification - Robert G. Bandes (deceased): Constraining-Unification and the Programming Language Unicorn (14 pp.). - Ken Kahn: Uniform -- A Language Based upon Unification which Unfies (much of) Lisp, Prolog, and Act 1 (28 pp.). 2.6. Semantic Foundations - Joxan Jaffar, Jean-Louis Lassez and Michael J. Maher: A Logic Programming Language Scheme (27 pp.). - Gert Smolka: Fresh: A Higher-Order Language with Unification and Multiple Results (56 pp.). ------------------------------ Date: Mon 24 Jun 85 16:03:14-PDT From: Emma Pease Reply-to: davies@csli Subject: Seminar - A Situational Theory of Analogy (CSLI) "A Situational Theory of Analogy" Todd Davies Conference Room, Ventura Hall CSLI, Stanford University Monday, July 1, 1985 1:15 p.m. Analogy in logic is generally given the form: P(A)&Q(A) and P(B) are premises --------- therefore Q(B) can be concluded, where P is a property or set of properties held by the analogous situation A in common with the present situation B, and where Q is a property which is initially held to be true of A. The question is: What justifies the conclusion? Sometimes the conclusion is clearly bogus, but for other pairs of situations and properties it seems quite plausible. I will give examples of both intuitively good and intuitively bad analogies as a way to argue that theories of analogy hitherto proposed are inadequate, and that the rationale for analogy which has been assumed for most early work on analogy in AI -- namely, that the inference is good if and only if the situations being compared are similar enough -- is based on a mistake. I will also point to traditional logic's inadequacies as a formal language for analogy and develop a theory which incorporates ideas from (and finds its easiest expression in) the theory of situations of Barwise and Perry. The theory suggests a general means by which computers can infer conclusions about problems which have analogues for which the solution is known, when failing to inspect the analogue would make such an inference impossible. ------------------------------ Date: Mon 24 Jun 85 15:16:42-PDT From: Alison Grant Subject: Seminar - Implementing Dempster's Rule (SU) SPECIAL MEDICAL INFORMATION SCIENCES COLLOQUIUM Tuesday, June 25, 1985 3:00 - 4:00 P.M. Room M-112, Stanford University Medical Center Speaker: Professor Glenn Shafer University of Kansas Title: IMPLEMENTING DEMPSTER'S RULE FOR HIERARCHICAL EVIDENCE Abstract: Gordon and Shortliffe have asked whether the computational complexity of Dempster's rule makes it impossible to combine belief functions based on evidence for and against hypotheses that can be arranged in a hierarchical or tree-like structure. In this talk I show that the special features of hierarchical evidence make it possible to compute Dempster's rule in linear time. The actual computations are quite straightforward, but they depend on a delicate understanding of the interactions of evidence. ------------------------------ Date: Mon, 24 Jun 85 10:29:34 edt From: Alan Gunderson Subject: Call For Papers-2nd ACM N.E. Reg. Conf. -- AI Track CALL FOR PAPERS SECOND ANNUAL ACM NORTHEAST REGIONAL CONFERENCE Integrating the Information Workplace: the Key to Productivity 28-30 October 1985 Sheraton-Tara Hotel Framingham, Mass. and The Computer Museum Boston, Mass. The conference sessions are grouped into tracks corresponding to major areas of interest in the computer field. Papers are solicited for the Conference's Artificial Intelligence Track. The Track's program will emphasize "real world" approaches and applications of A. I. Topics of interest include: - Expert Systems - Natural Language - Man-Machine Interface - Tools/Environments - A. I. Hardware - Robotics and Vision Submit papers by: July 22, 1985 Please send three copies of your paper to: Dr. David S. Prerau Track Chairman Artificial Intelligence Track ACM Northeast Regional Conference GTE Laboratories Inc. 40 Sylvan Road Waltham MA 02254 For additional information on the Conference, write: ACM Northeast Regional Conference P.O. Box 499 Sharon MA 02067 ------------------------------ End of AIList Digest ******************** From csvpi@vpics1 Mon Jul 1 04:42:45 1985 Date: Mon, 1 Jul 85 04:42:39 edt From: csvpi@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by csnet-relay.arpa id a024546; 1 Jul 85 1:04 EDT Date: Sun 30 Jun 1985 21:14-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #84 To: AIList@SRI-AI Received: from rand-relay by vpi; Mon, 1 Jul 85 04:31 EST AIList Digest Monday, 1 Jul 1985 Volume 3 : Issue 84 Today's Topics: Queries - Expert System Validation & LISP Productivity, Psychology - Predation/Cooperation & Common Sense, Business - TI and Sperry Join Forces, Games - Chess Programs and Cheating, Seminars - Learning in Expert Systems (Rutgers) & How to Clear a Block (SRI) ---------------------------------------------------------------------- Date: Sat, 29 Jun 85 01:58:21 edt From: Walter Maner Subject: Expert System Validation I would appreciate pointers to research dealing with answers to questions about expert system bugs, e.g., How can expert-system advice be validated? Are there failure modes specific to expert systems? What classes of error can be prevented by consistency enforcers? I am primarily interested in how these answers would apply to very large rule-based systems which have evolved under multiple authorship. Walter Maner CSNet maner@bgsuvax UseNet ...cbosgd!osu-eddie!bgsuvax!maner SnailMail Department of Computer Scinece Bowling Green State University Bowling Green, OH 43403 ------------------------------ Date: Wed, 26 Jun 85 13:02 CDT From: Patrick_Duff Subject: requested: papers concerning LISP programmer man-hours I am trying to locate articles which discuss the differences between LISP and non-AI languages in terms of the time and effort required to create prototype systems, to make additions or revisions to a design after much of the programming is completed, total programming time from start to finish, etc.. My opinion is that in general, it takes fewer man-hours to create a LISP program than to create a program to do the same task using languages such as Ada, Pascal, FORTRAN, or an assembly language. Note that I am *not* claiming that the program will also be "better", more efficient, or faster--just that most relatively large programs will take less time to write in LISP. I have been asked to come up with justification for using LISP based upon the total man-hours required. Does anyone know of a paper which would support or undercut my opinion? Has there been a convincing demonstration or test of the power of LISP (and its powerful programming environment) versus more traditional languages? regards, Patrick Patrick S. Duff, ***CR 5621*** pduff.ti-eg@csnet-relay 5049 Walker Dr. #91103 214/480-1659 (work) The Colony, TX 75056-1120 214/370-5363 (home) (a suburb of Dallas, TX) ------------------------------ Date: Saturday, 29 Jun 1985 22:22-EST From: munnari!psych.uq.oz!ross@seismo Subject: Predation/Cooperation (AIL v3 #78) David West (AIL v3 #82) mentioned the work of Robert Axelrod on the evolution of cooperation. Another good summary of Axelrod's work can be found in Douglas Hofstadter's Metamagical Themas column in Scientific American, May 1983, v248 #5, pp 14-20. UUCP: {decvax,vax135,eagle,pesnta}!mulga!psych.uq.oz!ross ARPA: ross%psych.uq.oz@seismo.arpa CSNET: ross@psych.uq.oz ACSnet: ross@psych.uq.oz Mail: Ross Gayler Phone: +61 7 224 7060 Division of Research & Planning Queensland Department of Health GPO Box 48 Brisbane 4001 AUSTRALIA ------------------------------ Date: Thu, 27 Jun 85 14:08:05 pdt From: Evan C. Evans Subject: Common Sense Common sense = conclusions reached thru the processes of natural reasoning (or behaviors resulting from such). I borrow heavily from Julian Jaynes, The Origin of Consciousness in the Breakdown of the Bicameral Mind. Natural reasoning is neither conscious nor rigorous in the sense of formal logic. For in- stance, upon observing a piece of wood floating on a given pond one will conclude directly that ANOTHER piece of wood will float on ANOTHER pond. This is sometimes called reasoning from partic- ulars. More simply, it's expectation based on subliminal gen- eralization. A baby quickly concludes that objects will fall without being AWARE of that conclusion. We're constantly ex- ercising natural reasoning to reach conclusions about others' feelings or motives based on their expressions or actions. Such reasoning was early recognized as unconscious & called automatic inference or COMMON SENSE, see John Steward Mill or James Sully. Pu's elaboration on Pratt stands, but it is well to remember that natural reasoning is usually unconscious & does not necessarily proceed by logical means. In fact, automatic infer- ence sometimes achieves correct conclusions by demonstrably il- logical means. evans@nosc-cc ------------------------------ Date: Thu 27 Jun 85 15:54:15-CDT From: Werner Uhrig Subject: news: TI and SPERRY join forces to sell AI [ from Austin American Statesman - June 26, 1985 ] TI captures computer deal with Sperry ===================================== (Kirk Ladendorf - Statesman staff) - TI has landed what it calls its biggest ever sales contract in the still-infant artificial intelligence industry - a three-year, $42 million deal to supply computers and related equipment to Sperry Corp. Sperry, one of the largest computer makers with $5.7 billion in sales last year, plans a large-scale campaign to develop specialized, salable uses for the TI machine, which Sperry will call the Knowledge Workstation. For TI the contract gives credibility that its well-regarded artificial intelligence system, called Explorer, is more than just an esoteric product with limited sales potential. ..... Sperry will combine the TI machine with a software system called the Knowledge Engineering Environment software developed by Intellicorp of Menlo Park, Calif. Intellicorp software is regarded as a very sophisticated tool for building specialized AI-programs. The new system can be used to create so-called "expert systems" which ... Such programs have been used on a demonstration basis to perform such tasks as the running of an electrical power-plant and experimental weather forecasting. Sperry's 26 specialized applications programs will be aimed at areas that have been difficult to serve with traditional computers. Those areas include software development; testing and debugging; navigation; communications sognal processing; CAD/CAM; and scheduling and resource allocation. Sperry chose TI over 2 principal competitors in the field, Symbolics and Lisp Machine Inc, because TI "has the best AI hardware available," a Sperry spokesman said. ..... Sales of AI Lisp-machines totaled only about $85 million last year, but Sperry projects the AI market will mushroom to more than $4 billion by 1990. ..... TI has announced no major additions to its 3,000 person Austin staff because of the new contract, but ... it has already begun to build the staff it needs to support the Explorer and the Business Pro. TI is already at work developing new features for the Explorer. They include developing computer communications links so that the AI machine can interact with Sperry and other IBM-compatible mainframes. .... ------------------------------ Date: Fri 21 Jun 85 21:44:18-EDT From: Andrew M. Liao Reply-to: LIAO%Weslyn.Bitnet@WISCVM.ARPA Subject: Chess, Programs And Cheating A Consideration Of "Do Computers Cheat At Chess?" I've been giving some thought to the question, "Do computers actually cheat at chess?". To start, I'm going to assume that what is at issue in the first objection is a chess program's use of a game tree whose nodes are representations of potential board/piece/move configurations. I think the objection that computers cheat because they use "external boards" (albeit represented internally) can be answered by saying, "No - there is no cheating involved because humans 'look ahead' in some way and since no physical external boards are allowed, the only way to 'look ahead' is to represent an 'image' of potential board positions in one's mind [though in a real limited way]. But isn't this just what a program does - only better?" I think that, in some sense, the argument that programs cheat at chess by virtue of having "internally represented 'external boards'" is just wrong. What a program tries to do on one aspect is to simulate what is going on inside a person's mind and, in a limited sense, this is actually achieved (albeit by brute force game trees). The second objection concerns the problems of "moves-made- by-reference". The objection, if I understand it correctly, is that (1) one cannot refer to moves that have been pre- recorded for the player's use during a match and that (2) such moves are encoded into a program (we disallow an external database file of moves since it is, in some way, a set of moves that have been pre-recorded for future use), and without these encoded moves, a program does not know what opening move(s)/strategy(ies) is optimal. Presumably, the reason for this rule is to force a player to rely on his experience and no one else's (i.e. no outside help) and at the same time, prevent any player being put at an unfair disadvantage. But I think it cannot be denied that encoding any move into a chess program is tantamount to making the program dependent upon the author's experience and not its own - a clear violation of the spirit of the rule. The question remains - Is it cheating? I am of the opinion that such a program is cheating on the basis that the program cannot decide during the opening of the game what strategy is optimal for it and hence must rely on outside help, in the form of stored data, given to it by its author. Although I feel the first objection is easily answered, I am still not happy with my reply to the second, although my intuition tells me that my reply to the second objection is, at least in spirit, on the right track. The motivation for my second reply is due (in great part) to J.R. Searle's conception of the Background which directly relates to the problem of "experience" and the like. ------------------------------ Date: Wed 26 Jun 85 09:58:03-PDT From: Ken Laws Subject: Re: Chess, Programs And Cheating Reply to Andrew Liao: When I open with a king pawn, I am relying on the experience and knowledge of others -- that doesn't seem to be cheating. I prefer an interpretation of the rules as "you run what you brung" -- namely that you cannot access external help >>during the match<<. I do admit that stored book openings seem questionable (although chess masters certainly memorize such material), but to say that a computer's superior memory gives it an advantage is no more damning than to say that its superior speed gives it an advantage. In just a few years it will be obvious that computers are inherently better "chess machines" than people are, and people will stop quibbling about handicaping the computer in one way or another to make the contest "fair". -- Ken Laws ------------------------------ Date: 28 Jun 85 11:07:37 EDT From: PRASAD@RUTGERS.ARPA Subject: Seminar - Learning in Expert Systems (Rutgers) LEARNING IN SECOND GENERATION EXPERT SYSTEMS Walter Van De Velde AI Laboratory, Vrije Universiteit Brussel This talk discusses a learning mechanism for second generation expert systems: rule-learning by progressive refinement. Second generation expert systems not only use heuristic rules, but also have a model of the domain of expertise so that deeper reasoning is possible whenever the rules are deficient. A learning component is described that abstracts new rules out of the results of deep reasoning. Gradually, the rule set is refined and restructured so that the expert system can solve more problems in a more effecient way. The approach is illustrated with concrete implemented examples. Date: Friday, June 28, 1985 Time: 11 AM Place: Hill Center, Room 423 ------------------------------ Date: Thu 27 Jun 85 12:21:03-PDT From: LANSKY@SRI-AI.ARPA Subject: Seminar - How to Clear a Block (SRI) "HOW TO CLEAR A BLOCK" or Unsolved Problems in the Blocks World #17 Richard Waldinger -- SRI AI Center 11:00 am, WEDNESDAY, July 3 Room EJ232, SRI International ABSTRACT: Apparently simple problems in the blocks world get more complicated when we look at them closely. Take the problem of clearing a block. In general, it requires forming conditionals and loops and even strengthening the specifications; no planner has solved it. We consider how such problems might be approached by bending a theorem prover a little bit. ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Wed Jul 3 10:02:37 1985 Date: Wed, 3 Jul 85 10:02:30 edt From: comsat@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by csnet-relay.arpa id aa00221; 2 Jul 85 13:41 EDT Date: Tue 2 Jul 1985 09:33-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #85 To: AIList@SRI-AI Received: from rand-relay by vpi; Wed, 3 Jul 85 09:52 EST AIList Digest Tuesday, 2 Jul 1985 Volume 3 : Issue 85 Today's Topics: Query - Othello, Games - Hitech Chess Performance & Computer Cheating, Psychology & AI Techniques - Contextual Reasoning ---------------------------------------------------------------------- Date: 1 Jul 85 17:36:29 EDT From: Kai-Fu.Lee@CMU-CS-SPEECH2 Subject: Othello (the game) [Forwarded from the CMU bboard by Laws@SRI-AI.] I am leading a PGSS project (for high school students) that will implement an Othello program in Common Lisp on the IBM PC. Any program source, clever tricks, and good evaluation functions that you're willing to share will be appreciated. /Kai-Fu ------------------------------ Date: 30 June 1985 2144-EDT From: Hans Berliner@CMU-CS-A Subject: Computer Chess (Hitech) [Forwarded from the CMU bboard by Laws@SRI-AI.] CLOSE BUT NO BIG CIGAR is an appropriate summary of the performance of Hitech in this week-ends Pittsburgh Summer Classic. In a field of 24, including 3 master and several experts, Hitech won its first three games against an 1833 (Class A), 1802 (Class A), and 2256 (Master) before losing in the final round to another Master (2263) who won the tournament. This was Hitech's first win against a Master. Its overall performance in two tournaments is 6 1/2 - 2 1/2; better than 70 percent. As it was, it finished 2nd in the tournament. Its provisional rating is now around 2100; middle expert. We will hold a show and tell on Friday at a time and place to be announced. ------------------------------ Date: 1 Jul 85 10:29:54 EDT From: Murray.Campbell@CMU-CS-K Subject: More on Hitech Result [Forwarded from the CMU bboard by Laws@SRI-AI.] The 3-1 result Hitech achieved in the chess tournament this past weekend was a performance rating of about 2300, well above the master standard of 2200. And it should be noted that the last round loss was to Kimball Nedved. After 2 tournaments, Hitech's performance rating is approximately 2210. ------------------------------ Date: Mon, 1 Jul 85 22:37:20 PDT From: Richard K. Jennings Subject: Computer Cheating. Computer programs are the consequences of people, like novels, as ripples are to rocks thrown in a smooth pond. Hence, it is the authors of these programs who are the cheaters! In fact, computer chess is simply another iteration, as is speed chess, double bug-house, and probably several other versions I have not heard of. Consenting adults can do what every they want as long as they don't bill the government. So, questions as to whether computers are cheating are really questions about whether a programmer who writes a program and watches it play is on the same footing as the live opponent. Practically I think he is, philosophically I think not. If the chess program is going to take more responsibility for its actions, I think it should have to 'learn' as you or I (are all of you, if expert systems, *learning* ones?). Of course, the author of the program still is partly responsible for how efficiently the program learns, and in finite time how good his creation will become. So, to apply the principles of recursion, let the program have to learn how to learn. At this level it is easy to see that skilled opponenets will cause a 2nd-loop learner to learn faster; hence the product is less a function of its original architecture and more a function of its environment -- which I guess we can (by default) attribute to the program (not its creator). At his point, if we put it in a closet and let it vegetate, it probably will not be very good at chess. This is certainly true in the limit as n (as in nth-loop learner) approaches infinity. It is easy to see that man is effectively an n-loop learner which cannot comprehend an o-loop learner for o>n. To be precise I should have said a *single* man. Groups of people, similarly and perhaps even including women, can function at some level p>n. Hence it should be possible for teams of people to beat individuals (and generally is). I see no reason for p or q to be bounded (where q is the class of learning evidenced by machine), and the problem has been reduced to a point: that is that man is just a transient form of intelligence which cannot be quantified (by himself anyway), only *measured*. Chess in its various forms does so (well you think when you win, poorly when you lose): and in its various forms is *fun*. Just remember, computer chess wouldn't be around if several smart people were not whipped at the game through careless errors by people so dumb that 'even a computer could beat them' or 'except for one careless mistake...' RKJ. ------------------------------ Date: Saturday, 29 Jun 1985 22:16-EST From: munnari!psych.uq.oz!ross@seismo Subject: Use of context to allow reasoning about time David Sherman (AIList V3 #71: Suggestions needed for tax rules) writes: > I am trying to design a system which will apply the rules of the Income > Tax Act (Canada) to a set of facts and transactions in the area of > corporate reorganizations. > ... > The passage of time is very important: steps happen in a particular > sequence, and the "state of the world" at the moment a step is taken is > crucial to determining the tax effects. > ... > The problem I have ... is how to deal with _time_. The following is just a suggestion. I have not actually tried it and I am not familiar enough with the literature to even say whether it is an old idea. However, it seems plausible to me and might be a useful approach. Time is not directly perceptible. It is perceived indirectly by noting that the environment (physical and cognitive) changes. There is a lot of survival advantage to believing in causality so the brain likes to attribute a cause to every change and when there is nothing obvious around to attribute causality to, we invoke the concept of time. As Dave Sherman pointed out, time is bound up with changes in the "state of the world", what I just called the environment. Lets shift into psychology and call it the context. Context plays a very important role in psychology. All the human and animal decision processes that I know of are context dependent. Consider a classic and simple memory experiment. The subject is given a list of nonsense words to memorise and is then given a new list of words some of which are from the memorised list to judge as old or new. This process may be repeated a dozen or more times in a session. How does the subject restrict his definition of old/new to the list he has just seen? It seems that the words are not remembered as individual and isolated objects but are remembered along with associative links to the context, where the context contains everything else that happened simultaneously. So, when memorising words in a list the subject links the words to each other, any extraneous noises or thoughts, even small changes in posture and discomfort. It has been shown that recognition and recall are greatly enhanced by reconstruction of the context in which memorisation occurred. Context is also evolutionarily important. It obviously enhances survival to be able to form associative links between the centre of attention and possibly anything else. The nasty thing about many environments is that you can't tell before hand what the important associations are going to be. Lets look at how context might be applicable to AI. In MYCIN, data are stored as triples. This is also a reasonable way to do things in PROLOG because it allows the data to be treated in a more uniform fashion than having each attribute (for instance) as a separate predicate. The objects in MYCIN are related by a context tree, but this has nothing to do with the sense in which I am using "context" so I will continue to call them objects. An object is a more or less permanent association of a bundle of attributes. That is, there is some constancy about it, which is why we can recognize it as an object (although not necessarily a physical one). By contrast the context is an amorphous mass of other things which happen to be going on at the same moment. There is little constancy to the structure of the context. The MYCIN triple cannot be related to it's context other than through the values of its object, attribute or value fields. There is no explicit way of showing that a fact was true only for a certain interval of time or only when a particular goal was active. I propose that the triple be extended to explicitly represent the context so it becomes . The values of the context variable would normally be unique identifiers to allow a particular context to be referred to. The context does not actually store any information, but many facts may be tied to that context. A context is a snapshot of the facts at some stage in the computation. Obviously there needs to be a lot of thought put into when to take the snapshots and the appropriate strategy will vary from application to application. The context will contain the facts being reasoned about at the time of the snapshot (probably when they had been whipped into consistency) but would also contain other relevant information such as goal states and clock times. For Dave Sherman's application there would probably be a new context snapshot when each transaction occurred (e.g. transfer of property in exchange for shares). Two additional facts within the context would be the earliest and latest clock times for which the context is valid. This would allow reasoning about changes of state and elapsing of time because the before and after states are simultaneously present in the fact base along with the clock times for which they were true. A couple of other uses of contexts suggest themselves. One is the possibility of implementing "possible worlds" and answering "what if" questions. If the system is capable of manipulating contexts it could duplicate an existing context (but with a new context ID of course), modify a few of the facts in the new context and then start reasoning in that context to see what might have happened if things had been a little different. Another possibility is that it might be useful in "truth maintenance systems". I have heard of them but not had a chance to study them. However their references to assumption sets and dependency directed backtracking sound to me like the idea of tracking the context, attributing changes in the context to various facts within the context and then using that information to intelligently manipulate the context to implement backtracking in a computation. UUCP: {decvax,vax135,eagle,pesnta}!mulga!psych.uq.oz!ross ARPA: ross%psych.uq.oz@seismo.arpa CSNET: ross@psych.uq.oz ACSnet: ross@psych.uq.oz Mail: Ross Gayler Phone: +61 7 224 7060 Division of Research & Planning Queensland Department of Health GPO Box 48 Brisbane 4001 AUSTRALIA ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Wed Jul 3 10:15:46 1985 Date: Wed, 3 Jul 85 10:15:35 edt From: comsat@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by csnet-relay.arpa id a000336; 3 Jul 85 2:19 EDT Date: Tue 2 Jul 1985 22:10-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #86 To: AIList@SRI-AI Received: from rand-relay by vpi; Wed, 3 Jul 85 09:59 EST AIList Digest Wednesday, 3 Jul 1985 Volume 3 : Issue 86 Today's Topics: Administrivia - Addresses of Seminar Presenters, Queries - Statistics of Syntactic Structures & Spatial Reasoning & UNIVAC 1100 LISP & Symbolics's User Interface, Expert Systems - Validation, AI Tools - Interlisp Comments & C vs. LISP ---------------------------------------------------------------------- Date: Monday, 1 Jul 1985 20:49-EST From: munnari!psych.uq.oz!ross@seismo.ARPA Subject: addresses of seminar presenters Being from out of town I find it a little difficult to get to most of the seminars advertised in the AIList. However, there are some I would like a little more information on by contacting the presenter to get a copy of the talk or, more likely, a related paper or report. Unfortunately, most of the seminar announcements give no network address for the presenter and an inadequately specified postal address. Would it be possible to exhort seminar hosts to put complete addresses in the announcements or at least make sure that they ask the presenter for an address so that others may find out from the host? -- Ross ------------------------------ Date: Tue, 2 Jul 85 10:26:31 EDT From: "Ben P. Yuhas" Subject: Read My Lips Here at the Sensory Aids Lab, we are beginning to explore some of the strategies used by lip readers to decode the visual speech signal. One of the questions we want to answer is to what degree does syntactic structure influence a sentence's lip read- ability. In developing a data base of test sentences on laserdisk, we began to wonder whether anyone had ever attempted to find the statistical distribution of syntactic structures in spoken English. I realize this distribution might vary greatly from group to group. If there are any computational linguists with references or thoughts on this matter I would appreciate hearing from you. yuhas@hopkins-eecs-bravo Snailmail: Ben Yuhas Dept EECS Johns Hopkins University Baltimore, MD 21218 ------------------------------ Date: 2 Jul 85 08:27:00 EDT From: " CUGINI, JOHN" Subject: spatial reasoning Can anyone suggest a good survey article or textbook that covers AI for spatial reasoning, especially for 3-D? I have in mind things like, "will this refrigerator fit thru that door", etc. Thanks for any help. John Cugini Institute for Computer Sciences and Technology National Bureau of Standards Bldg 225 Room A-265 Gaithersburg, MD 20899 phone: (301) 921-2431 ------------------------------ Date: Mon, 1 Jul 85 11:51:36 EDT From: Marty Hall Subject: UNIVAC 1100 Series LISP I am trying to find where I can find documentation or info on the Univac 1100 LISP. We are referring to a system that was written in that dialect, and converting it to Common LISP. However, there are two functions that appear that are found in neither MACLISP or LISP 1.5 (which this LSP was supposed to be), that we can't find out what they do. The functions are "AMB" and "STACK". They are used: (setq (amb (stack ))) We have called lots of people, including the local Sperry Corp folks, and no one seems to know. Any suggestions on where to look or who to call/send to ? -Marty Hall hall@hopkins aplvax!m2lab!hall@maryland ------------------------------ Date: Tue, 2 Jul 85 11:26 EDT From: susan watkins Subject: Symbolics's application user interface I'm working as a developer at Symbolics, Inc. in Cambridge, Ma. I would like to get opinions, reactions, problems encountered, constraints the system s/w imposes, etc. from programmers who have use the Symbolics s/w to develop a reasonably sized product. e.g. what problems they have run into trying to use the window system. I'll be at IJCAI, so I'll more than happy to meet and talk with anyone who is interested. My mail-stop is chaowatkins@SCRC-STONY-BROOK.ARPA. ------------------------------ Date: Mon 1 Jul 85 09:54:22-PDT From: Bruce Buchanan Subject: validation of expert systems Ted Shortliffe & I tried to address the issues surrounding evaluation of expert systems in chapter 30 of RULE-BASED EXPERT SYSTEMS. We did not specifically talk about what to do in the case of very large knowledge bases built by multiple experts, but chapter 8 does discuss some knowledge-base editing facilities that should help. I would like to know of work on these problems. B.Buchanan ------------------------------ Date: 26 Jun 1985 1215-PDT (Wednesday) From: Steven Tepper Subject: Putting comments in Interlisp programs (flame) [Forwarded from the Stanford bboard by Laws@SRI-AI. This is part of an exchange on hacking and software engineering.] I didn't say that comments are impossible in Interlisp -- merely that it's painful to put them in. For the edification of those who have not had the privilege of being subjected to Interlisp's slavish adherence to the principle that it should constitute an entire programming environment (as opposed to being just another programming language living on a general purpose computer system), one of the concomitant requirements of this philosophy is that all operations, including editing, be done on Lisp objects. This means that comments (which are handled by a function called * that does not evaluate its arguments) are a part of the running program. Thus, extreme care is required in the placement of comments. For example, the following function fails: (DEFINEQ (FOO (LAMBDA (X Y) (COND ((GREATERP X Y) X) (* Return the maximum value) (T Y] because the comment is treated as a clause to the COND. Similarly, a comment placed as the last form in a function (Interlisp provides an implicit PROGN in function definitions) will return the first word of the comment as the value of the function. In fact, because Lisp is largely a functional language, there are relatively few safe places to put comments. A further indication of the low repute with which comments are held in Interlisp is the fact that the common way of displaying a function at the top level, PP (pretty-print), replaces all comments with the symbol **COMMENT**. To me, this is backwards -- if anything, the comments should be given prominence over the Lisp code. Similarly, in the display editor on Interlisp-D, comments are kept as far away from the executable code as possible (on the same line) and displayed in a font which is considerably less readable than that used for non-comments. This is the basis on which I justify my earlier claim that Interlisp "discourages" comments, which I consider an undesirable goal. ------------------------------ Date: Thu 27 Jun 85 11:47:41-PDT From: Liam Peyton Subject: Interlisp [Forwarded from the Stanford bboard by Laws@SRI-AI.] If Interlisp's approach to comments prevents one from inserting mindless comments like the following: (DEFINEQ (FOO (LAMBDA (X Y) (COND ((GREATERP X Y) X) (* Return the maximum value) (T Y))))) then it should be praised not criticized. A step by step translation is not helpful commenting and certainly does not give the comments more prominence than the actual code. (if anything it reduces the relevance of comments). A short summary before the cond explaining what the code is doing is far neater and more useful. Why would one ever want to have a comment in the last line of a PROG? (* comment: this is the end of the prog) This is not to say by any means that Interlisp has the ideal means of handling comments or that Interlisp doesn't have its problems. It does, but they are certainly not a basis for rejecting it as a programming environment. Some of the things that result from "Interlisp's slavish adherence to the principle that it should constitute an entire programming environment" are incremental execution for debugging purposes, sophisticated mouse and window system with interactions between windows, online text processing, and online graphics. A general purpose computer is a computer that can do anything painfully. ------------------------------ Date: Mon, 1 Jul 85 21:46:21 PDT From: Richard K. Jennings Subject: C vs LISP We have continuing debates about that subject all the time, and I think for us we have come to the conclusion (for now) that C is better than LISP. Currently we have MS-DOS 2.0, XLISP 2.0, and Lattice C compiler version 2.0. A new man (fish@aerospace) was given the simple task to write a plotter driver. He did this by first writing a plot spooler of sorts in assembly language (after failing miserably in basic). At this point, he was a virgin programmer, save some fortran programming on large machines. He then started on another program to help people interactively develop a 'plotter language' file which his plot spooler could plot. At this point he was manually generating the plotter language, to produce vu-graphs (as I said he was *the* new man). Using a copy of Winston's Lisp text, he set out with XLISP to produce this translator. After a month or so, his incentive to write an interactive translator (to get him out of the vu-graph making loop) dissipated. About a month ago, he went to NARDAC and attended a 1-day C tutorial. Then he brought up the C compiler, and now is just about done. There is no doubt in my mind that he prefers Lattice C to XLISP. Before all the LISP people flame let me make a few comments. I am familliar with both C and XLISP, and have programmed in each. Both are pretty basic, but in my opinion I would chose XLISP to write the basic program, and then recode it to C if it was going to be maintained. During this effort I acted as the trouble shooter, and let me say that if I was going to supervise a programming team, THEY WOULD USE C. In fact, from the management level, I think Lisp is only marginally better than assembly language: perhaps. By September we should be using PC-AT's with GC-LISP, and the new Microsoft C compiler. By December we will probably have our Symbolics, without a C compiler (although the salesman evidently has one to sell which ought to say something). So shoot me a note about Jan and I may have changed my tune. To conclude, if you have a one man project, and the lisp environment you plan to use has a lot of functionality you need in you application built-in, lisp can probably be justified. The source libraries now available in C (or Pascal, real soon now for Ada) will be increasing difficult to beat, especially in the context to C interpreters and incremental compilers, and the fact C runs on *everything*. If you are starting from scratch -- common Lisp -- (even XLISP is extended to support object oriented programming) ***good luck***! Richard Jennings AFSCF/XRP Sunnyvale ARPA: jennings@aerospace ->standard disclaimer applies<- ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Mon Jul 8 18:26:40 1985 Date: Mon, 8 Jul 85 18:26:35 edt From: comsat@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by csnet-relay.arpa id a000702; 6 Jul 85 18:24 EDT Date: Sat 6 Jul 1985 14:47-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #87 To: AIList@SRI-AI Received: from rand-relay by vpi; Sat, 6 Jul 85 21:01 EST AIList Digest Saturday, 6 Jul 1985 Volume 3 : Issue 87 Today's Topics: AI Scholarship - NMSU, Seminars - Logic Programming with Functions (BBN) & Shape from Function (GE), Conferences - Expert Systems Application in Business & Intelligent Simulation Environments & North American Fuzzy Information Processing Society ---------------------------------------------------------------------- Date: Thu, 4 Jul 85 11:25:17 mdt From: yorick%nmsu.csnet@csnet-relay.arpa Subject: AI Scholarship - NMSU GRADUATE AND UNDERGRADUATE SCHOLARSHIPS: New Mexico State University, Computing Research Laboratory, invites applications from excellent graduate and undergraduate students interested in Artificial Intelligence, including Expert Systems, Natural Language, Cognitive Modelling, Intelligent User-Interfaces, Vision and Robotics, and interdisciplinary projects that integrate these fundamental aspects of computing science. The CRL offers scholarships of up to $12,000 for graduates and $3,000 for undergraduates per year including tuition and cash. Successful applicants will additionally be employed for up to 20 hours per week during the academic year, and 40 hours per week during the summertime on CRL sponsored research programs. Applications should include a letter indicating your intent to be considered for one of these scholarships, a statement of your experiences, a statement of your interests and future goals, transcripts of all undergraduate work, and names and addresses of 3 references who know your abilities in computing science. Please send applications, by 20 July 1985 to: Dr. Yorick Wilks, Director, Computing Research Laboratory, Box 3CRL, New Mexico State University, Las Cruces, NM 88003. ------------------------------ Date: 25 Jun 1985 16:11-EDT From: Brad Goodman Subject: Seminar - Logic Programming with Functions (BBN) [Forwarded from the MIT bboard by SASW@MIT-MC.] BBN Labs SDP AI Seminar Speaker: Uday S. Reddy University of Utah Title: Logic Programming with Functions Time: Friday, July 19th, 10:30 a.m. Place: 3rd Floor Large Conference Room 10 Moulton Street, Cambridge While functional programming has been with us for more than two decades, logic programming is a relatively new programming language concept. A comparison of the two styles shows that functional programming is done by rewriting expressions to semantically equivalent ones, while, on the other hand, logic programming is done by solving formulas for values of their free variables. Thus, logic programming provides significantly more expressive power than functional programming. However, it is possible to perform logic programming in functional languages. Whereas Horn-clause logic languages use resolution as the operational mechanism, functional logic languages use a mechanism called "narrowing". Given an expression with free variables, the narrowing mechanism answers the question "for what values of the variables does the expression reduce to a value?". Narrowing is a generalization of both rewriting and resolution and so makes it possible to use both the styles of programming in a unified framework. ------------------------------ Date: Thu, 27 Jun 85 14:14 EST From: "S. Holland" Subject: Seminar - Shape from Function (GE) SHAPE FROM FUNCTION VIA MOTION ANALYSIS with Application to the Automatic Design of Orienting Devices for Vibratory Part Feeders Dr. Tomas Lozano-Perez MIT Artificial Intelligence Lab Cambridge, MA. 02139 Wednesday, August 14, 1985, 11:00 a.m. General Motors Research Laboratories Computer Science Department Warren, Michigan 48090-9057 This talk explores the premise that the function of many devices can be characterized by how they interact with other objects. I suggest a representation of function of these devices in terms of motion constraints. These motion constraints are expressed as a diagram in configuration space. Combinations of these diagrams serve both in describing a device's function and in designing devices with specified behavior. This leads to a view of design as an inverse of the motion planning problem in robotics. In both cases we know the shape of the moving part. In motion planning, we are given the obstacles and we must find a legal path between the specified origin and distination. In this view of design, however, we are given the desired motion (actually a range of possible motions) and are asked to find a legal shape of the obstacle, that is, the device. I illustrate this approach to design with a case study of mechanical part feeders, a class of real devices with an interesting and direct relationship between shape and function. Dr. Lozano-Perez has authored technical articles in the areas of motion planning, robot programming, and model-based object recognition. He has been affiliated with the M.I.T. Artificial Intelligence Laboratory since 1973. ------------------------------ Date: 07/05/85 15:18:19 MEZ From: Christian Bader Subject: Expert systems application in business A workshop on Expert Systems in business will be held November 26/27 1985 in Berlin (West Germany) as a part of the BIG-TECH fair. We are interested to know about business applications of Expert Systems both in Germany and elsewhere. Please let me know if you have an expert systems application that you could present at the workshop. Please contact Christian Bader Technische Universitaet Berlin Sekr. FR 6-7 Franklinstr. 28/29 D-1000 Berlin 10 West Germany phone: (49-30)-314-4903 or (49-30)-314-73260 (leave message) Network address: ARPA : BADER%DB0TUI11.BITNET@WISCVM.ARPA BITNET : BADER at DB0TUI11 CSNET : BADER%DB0TUI11.BITNET@WISCVM.ARPA ------------------------------ Date: Tue, 2 Jul 85 16:03 CST From: Adelsberger%tamu.csnet@csnet-relay.arpa Subject: Conference - Intelligent Simulation Environments CALL FOR: PAPERS, PANELISTS, SESSION COORDINATORS INTELLIGENT SIMULATION ENVIRONMENTS, 1986 SCS MULTICONFERENCE, JAN 23 - 25, SAN DIEGO The Society for Computer Simulatiion is sponsoring a multiconference January 23-25, 1985 in San Diego, California. Solicited are papers in the areas of: * User friendly simulation environments. * Knowledge based simulation sytems. * Artificial intelligence applied to simulation environments. Papers of special interest might describe models that (1) have many symbolic processes, (2) use heuristic search, (3) have a command structure separate from knowledge domain, (4) have expertise built into the model so that decisions by the user would be minimized. AI papers dealing with subjects that are not necessarily directly simulation related but which have a strong time dimension or concern would also be welcome. We are also interested in panel discussions or sessions coordination on a particular aspect of the subject. Detailed abstracts (300 words) of proposed papers and special sessions should be sent direct to the program chairmen not later than July 21, 85. Camera ready copies of accepted papers would be due October 15, 1985. Heimo H. Adelsberger Program Chairmen Texas A&M University Computer Science Department College Station, TX - 77843 Phone: (409) 845-0298 ------------------------------ Date: Mon, 24 Jun 85 15:31:56 cdt From: Don Kraft Subject: Call for Papers -- NAFIPS Meeting CALL FOR PAPERS North American Fuzzy Information Processing Society (NAFIPS) International Meeting Monteleone Hotel New Orleans, Louisiana (In the Heart of the French Quarter) June 1-4, 1986 Papers on all fuzzy topics are encouraged, and wide international participation is expected. Deadlines Notice of intent with a title and abstract 9/1/85 Completed paper (3 copies) 10/15/85 Notification of acceptance 1/15/86 Camera-ready copy due 3/15/86 Proceedings will be distributed during Conference registration. Send all abstracts and papers to: NAFIPS86 Department of Computer Science Florida State University Tallahassee, FL 32306 Abraham Kandel and Wyllis Bandler, Program Committee Co-Chairs Fred Petry and Donald H. Kraft, General Meeting Co-Chairs ------------------------------ End of AIList Digest ******************** From csvpi@vpics1 Mon Jul 8 18:26:19 1985 Date: Mon, 8 Jul 85 18:26:13 edt From: csvpi@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by csnet-relay.arpa id a000845; 7 Jul 85 14:56 EDT Date: Sun 7 Jul 1985 10:51-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #88 To: AIList@SRI-AI Received: from rand-relay by vpi; Sun, 7 Jul 85 21:23 EST AIList Digest Sunday, 7 Jul 1985 Volume 3 : Issue 88 Today's Topics: Queries - Generators in Lisp & Expert Systems for Configuration & Natural Language Processing Software, Robotics - Spatial Reasoning, Games - Learning in Chess Programs, Publications - New IEEE AI Journal, Review - AI Report Vol 2 No 5 & AI Report Vol 2 No 7 ---------------------------------------------------------------------- Date: 3 Jul 1985 14:17-EDT From: Conal.Elliott@CMU-CS-CAD.ARPA Subject: Generators in Lisp query I'd like to implement a simple generator-language (i.e. functions with backtracking and able to return more than once) on top of Common Lisp. It doesn't need to be anything fancy, as it will be for my own use. I would appreciate any hints or pointers. Conal Elliott conal@cmu-cs-cad.arpa ------------------------------ Date: Tue 2 Jul 85 22:29:55-PDT From: Marty Tenenbaum Subject: Expert Systems for Configuration Does anyone know where I might acquire an expert system for solving configuration problems (ala R1/XCON) on a PC. I am interested in such a system, more as a tutorial aid than as a serious application. Jay M. Tenenbaum, Schlumberger Palo Alto Research. Please respond to Tenenbaum@SRI-KL, or call me at 415-496-4699. ------------------------------ Date: Wed, 3 Jul 85 12:24:15 cdt From: Mark Turner Subject: natural language processing TIRA at U Chicago is looking for a robust Natural Language Processing system - actual code - it can obtain and install on a 4.2BSD Unix system. To elaborate: Many faculty members from Departments of Library Science, English, Linguistics, Classics, Romance Languages, etc. at U Chicago who currently work in searching and processing natural language text data bases have now formed the Textual Information Retrieval and Analysis (TIRA) research center. The Department of Computer Science at U Chicago is only a few years old, and although I understand that it would be interested in hiring an Assistant Professor in AI/NLP, it has not yet done so. Consequently, we lack a faculty member who might focus his energies on installing and tuning a Natural Language Processing System. Several of us are familiar with NLP, though, and we have some programmers on staff. So I am beginning to wonder how we might obtain, for academic research purposes, the code and documentation for someone else's NLP system, and install it here with relative ease, to help us with semantic, grammatical, thematic, and morphological parsing, in various Indo-European languages, principally English, French, Greek, Latin, Italian, German, and Spanish. I would appreciate your responses. Mark Turner Department of English U Chicago 60637 >ihnp4!gargoyle!puck!mark ------------------------------ Date: Wed, 3 Jul 1985 14:14 EDT From: Juliana Kraft Subject: Spatial Reasoning Query From: " CUGINI, JOHN" Can anyone suggest a good survey article or textbook that covers AI for spatial reasoning, especially for 3-D? I have in mind things like, "will this refrigerator fit thru that door", etc. Thanks for any help. For 3D you must consider 6 degrees of freedom (3 translational and 3 rotational). I recommend "Motion Planning with Six Degrees of Freedom," by Bruce Donald, (261 pp), MIT AI-TR 791, available from Publications Office MIT AI Laboratory Room NE43-818 545 Tech Square Cambridge, MA 02139 (617) 253-6773. ------------------------------ Date: Wed, 3 Jul 85 17:08 pst From: "furth john%d.mfenet"@LLL-MFE.ARPA Subject: The Best Chess Program I would like to add something to Richard Jennings' words on the responsibility of the author of a chess program for its performance. The most rigid chess program will play chess only as its author would at his/her best. The program that learns has the possibility of doing better. Suppose the author wrote his/her program without any instructions for playing chess but only for learning how to play chess. Then the program could learn and execute maneuvers that the author was unaware of. Now this program learns only as its author learns at his/her best. We may continue this iterative procedure to some arbitrary degree and declare the author's taint to be negligible. In the process, however, we will probably have accumulated some large overhead. The time spent passing information up and down this ladder of learners and the storage required at each rung of the ladder will make the program unusable. To attain an independent and useful intelligence, the learner must be able to discard signifigant portions of the means by which it has arrived at its present level of ability. The original hub of its actions must fall away and a new one be generated. So the adult forgets the involvements of childhood and the state the cares of its early days. With whatever vestiges remain, the organism must take on a whole new orientation to meet new needs with a closer approach to the optimum. It is better to forget the past than to live there. The best chess program will forget most everything its author ever told it to do. John Furth ------------------------------ Date: Sun 7 Jul 1985 11:11-PDT From: Laws@SRI-AI Subject: New IEEE AI Journal >From IEEE Computer, July 1985, p. 101: IEEE Expert is the newest addition to the Computer Society's list of publications, which already includes five magazines. The Computer Society Board of Governors gave its approval to the new quarterly at its May 10 meeting ... David Pessel of Standard Oil of Ohio will serve as acting editor-in- chief and will have the responsibility of preparing for the initial publication in the first quarter of 1986. The magazine is expected to treat such AI areas as knowledge engineering, natural language processing, expert systems, and conceptual modeling. ... ------------------------------ Date: 4 Jul 1985 10:46-EST From: leff%smu.csnet@csnet-relay.arpa Subject: AI Report Vol 2 No 5 Summary Report on Stanford University AI efforts including Knowledge Systems Laboratory (VLSI design, MOLGEN, interpretation of nuclear magnetic resonance data on proteins, computer-aided teaching of diagnostic reasoning, ONCOCIN for administration of medical treatment protocols, lymph node pathology diagnosis system, robotic manufacturing strategy development, financial resource planning). Basic AI research includes non-monotonic reasoning, robotics, mechanical construction of computer programs, design, description and interaction with computer systems, database retrieval research. RADIX [formerly RX] is a project which will use computer programs to examine over 50,000 patient years of accumulated medical data. Report on ESPRIT and ALVEY, AI efforts of the European Economic Community and England respectively: The following are a list of some books mentioned in the report: Artificial Intelligence Applications for Manufacturing Artificial Intellgience Applications for Business Management The 1985 Handbook of Manufacturing Software (all three by SEAI Techical Publications) Machine Vision -- A Summary and Forecast (Tech Tran Corporation) A Practical Guide to Designing Expert Systems by S. Weiss and C. Culikowski William Gevarter: Artificial Intelligence, Expert System, Computer Vision and Natural Language Processing William Gevarter: Artificial Intelligence and Robotics: Five Overviews Mitsubishi Research Institute has inititated a multi-client AI research project Report of work done by the Knowledge Information Research Institute of Computer Services Corporation of Japan Report on AI at Ohio State: medical systems which infer data from broad data descriptions and concepts including a red cell antibody identification system, a system for diagnosing fuel problems in autombile engines, an air cylinder design system. Report on Imperial Chemical Industries which has developed an expert systems shell called Savoir, an agricultural advisor system. Infologics of Stockholm has announced a PROLOG for IBM-PC costing 295 dollars. Automata Design Associates has five versions of PROLOG available for IBM-PC (public domain, educational $29.95, FS Prolog, $49.95, Virtual memory prolog $99.95 and Large virtual model prolog $300). TOPSI is selling an OPS-5 for CP/M and MS-DOS for $400.00. The Automated Reasoning Corporation is selling a fault-diagnosis system. Odetics (the maker of six legged robots) has announced the development of an AI center. Expert Technologies has been developed to sell AI technology to printers and publishers. Report on new shareholders of NCC. Lynn Conway has left Darpa to join University of Michigan ------------------------------ Date: 4 Jul 1985 10:18-EST From: leff%smu.csnet@csnet-relay.arpa Subject: The AI Report Vol 2 No 7 Summary The Artificial Intelligence Report July 1985 Volume 2 No 7 Nippon Telephone and Telegraph Report on Nippon Telephone and Telegraph, the Japanese AT&T, (NTT) includes general description of company and its computer related R&D efforts. In AI, they are working on a Japanese-English translation effort, medical expert systems, systems to recognize handwritten Japanese and Chinese characters, robotics, speech recognition and speech synthesis. They have also developed a Lisp machine using the language Tao, which is a blending of LISP, PROLOG and Smalltalk. It 40 to 50 times faster than ZetaLisp interpreter, 3 times faster than Smalltalk-80 on the Xerox Dolphin and five times faster than the DEC-10 Prolog interpreter. Also Computer Services Corporation (CSK) is completing work on a LISP machine prototype which will run Prolog, LISP, UNIX and process Japanese natural language input. The AI profits discusses interest by new and old companies in AI. based at a Gartner group forum. Reports on Lisp machine vendors, Texas Instruments, Symbolics, Xerox, Lisp Machine Inc (LMI). Symbolics revenues are expected to top 85 million dollars this year and LMI revenues will top 25 million dollars. They predict that Xerox will introduce a 10,000 dollar low-end AI machine. The Gartner's group of Lisp machine sales in 1990 is over one billion dollars. Also discusses expert systems. They feel that natural language understanding will not be as big a seller as expert system tools. IBM has over 300 researchers pursueing AI objectives. Reports on new DEC microvax products, management changes at Lisp Machine Inc, announcement by Radian Corproation of a IBM PC expert system shell, an apple Macintosh OPS5 interpreter, R&D expenditures for various companies, a prediction that a billion transistors will be packed on a single chip. The Institut fur Entscheidungstheorie und Unternehmesforschung at der Universitat Karlsruhe in Germany is conducting an international survey on Expert Systems in business. They review the following: The Fourth Technical Conference of the British Computer Society Speicalist Group on Expert Systems which has been published as "Research and Development in Expert Systems" V. Daniel Hunt's "Smart Robots: A Handbook of Intelligent Robotic Systems" Eugene Charniak and Drew McDermott's "Introduction to Artificial Intelligence" Also report on the Army AI Center which is doing research on systems to field new equipment to the army. The following is a list of some government documents on AI that I found in this report: An Overview of Artificial Intelligence and Robotics, NASA-TM-85836 An Overview of Computer Vision PB83-217554, An Overview of Computer-Based Natural Language Processing PB83-200832, Overview of Expert Systems PB83-217562 and Flexible Manufacturing System Handbook ADA-127927. ------------------------------ End of AIList Digest ******************** From csvpi@vpics1 Mon Jul 8 18:25:59 1985 Date: Mon, 8 Jul 85 18:25:52 edt From: csvpi@vpics1.VPI To: fox@opus (FRANCE,RDJ,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by csnet-relay.arpa id aa00309; 7 Jul 85 20:16 EDT Date: Sun 7 Jul 1985 16:28-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #89 To: AIList@SRI-AI Received: from rand-relay by vpi; Sun, 7 Jul 85 21:26 EST AIList Digest Monday, 8 Jul 1985 Volume 3 : Issue 89 Today's Topics: AI Tools - Lisp vs. C & Interlisp Comments ---------------------------------------------------------------------- Date: Wed, 3 Jul 1985 15:30 EDT From: "Scott E. Fahlman" Subject: Lisp vs. C I won't try to take issue with Richard Jennings's views on Lisp vs. C, except to note that he is only in a position to compare one dialect of Lisp (XLISP) to one dialect of C on some sort of MS-DOS machine -- presumably a tiny one -- for one particular kind of task with one particular virgin programmer who had been trained in in a different way on the two languages. His observations are probably valid for this case, but I wouldn't draw any sweeping generalizations from this. Lisp really requires a full-fledged environment in order to be an attractive language. A lot of people got turned off very badly back in the bad old days when the Lisp environment was primitive and the address space of most machines was too small to hold the kinds of features that we see today on the various Lisp machines. Now we are seeing the same "turn off" among people whose exposure to Lisp consists only of using very small Lisps on machines with only, say, 512K bytes of memory. On such a machine, a language like C (which evolved to fit the PDP-11, a machine whose address space is even smaller) probably is superior for getting real work done. This phase will pass just as soon as machines with adequate virtual memory systems become as common as PC's are today. -- Scott Fahlman ------------------------------ Date: Wed 3 Jul 85 13:55:02-PDT From: Christopher Schmidt Subject: ~= re: Interlisp comments [Forwarded from the Stanford bboard by Laws@SRI-AI.] [...] Unfortunately at least once programming system I know of (Interlisp) actively discourages this by making it painful to put comments in programs. For this reason alone Interlisp disqualifies itself as a serious programming environment. Isn't that a little like saying that Boeing, by having tiny restrooms disqualifies itself as a serious airplane manufacturer? I didn't say that comments are impossible in Interlisp -- merely that it's painful to put them in. Pain is subjective. For my money, tracking down bogus indentation, missing semicolons or unbalanced parentheses or misspelled symbols in a text-editor-oriented language is ever so much more painful than commenting Interlisp. All of these problems associated with the syntactic sugar of most programming languages are non-issues in Interlisp, where the interned form is edited. For the non-Interlisper it should be explained that in DEdit (the Interlisp-D display structure editor) the programmer selects S-expressions with the mouse and operates on them through a menu. Note that the user selects S-expressions; not the textual representation of an S-expression. Eg., a user can't select a parenthesis because it is only an artifact of the pretty-printer. Since parentheses are generated only by the pretty-printer on redisplaying a transformed S-expression, they can never be unbalanced. Since atoms are typically "typed" into a program by buttoning an existing instance and using a menu command which copies the interned pointer (not the characters of the PNAME) spelling errors can't occur. Similarly, indenting is the job of the pretty printer-- not the programmer--and is done per-window, so editing windows may be of different sizes. I am amused to see CADR-sized editor windows on 3600's. For the edification of those who have not had the privilege of being subjected to Interlisp's slavish adherence to the principle that it should constitute an entire programming environment (as opposed to being just another programming language living on a general purpose computer system), one of the concomitant requirements of this philosophy is that all operations, including editing, be done on Lisp objects. Some of us are willing slaves. This means that comments (which are handled by a function called * that does not evaluate its arguments) are a part of the running program. Thus, extreme care is required in the placement of comments. [...] Extreme care must be taken when spelling identifiers in C programs and placing semicolons and parentheses. Nobody said programming would be safe! <:-) [...] In fact, because Lisp is largely a functional language, there are relatively few safe places to put comments. I don't think that it is unfair to ask a programmer to learn the semantics of the programming language he is using. Given an understanding of PROGN, COND, AND, OR, and the interpreter, I don't think it is difficult to comment Interlisp programs at all. Anyway, the compiler catches instances where the programmer has used a comment for its value and warns him with a message like "Warning: value of comment used". Actually, I haven't seen this message in a year or two. When was the last time cc barfed on a semicolon you forgot? [...] Similarly, in the display editor on Interlisp-D, comments are kept as far away from the executable code as possible (on the same line) and displayed in a font which is considerably less readable than that used for non-comments. You have only to rebind COMMENTFONT to a larger/bolder fontclass if you don't like the default. What font does EMACS use for comments in C programs? This is the basis on which I justify my earlier claim that Interlisp "discourages" comments, which I consider an undesirable goal. As I've said too many ways above, I don't think Interlisp discourages misplaced comments any more than C discourages misplaced semicolons. --Christopher Schmidt ------------------------------ Date: Sat 6 Jul 85 00:01:40-PDT From: Peter Karp Subject: Interlisp comments [Forwarded from the Stanford bboard by Laws@SRI-AI.] The Interlisp manual contains roughly 25 chapters which average approximately 25 pages each. These are big pages with closely packed type. The Interlisp environment has been evolving for almost 20 years. This should suggest to you that it is a rich and complex entity. Are we really expected to take seriously the proposition that we shouldn't use this language because it doesn't let you put comments anywhere you please? [...] ------------------------------ Date: Fri 5 Jul 85 16:24:29-PDT From: Christopher Schmidt Subject: ~= re: Interlisp comments [Forwarded from the Stanford bboard by Laws@SRI-AI.] [...] We're not talking about syntactic sugar or ease of use. We're talking about a limitation on the functionality of the programming environment. My mistake then. I think that C comments are the epitome of syntactic sugar; discarded by the cc parser as if they never existed. And rather than Interlisp's being a programming environment that "holds comments in contempt" I would go so far as to claim that it gives them MORE respect than unix/EMACS/C. In the unix/EMACS/C environment let's say that we have an error. unix gives us one of the many helpful error messages from the set {bus error core dumped | segmentation error core dumped}. The user invokes the break package (cdb) (by hand). Where are the comments now? Where is the symbol table? The hacker only gets that if he is lucky! In Interlisp, by contrast, one of >50 error messages is printed, a break package window opens up, a stack trace opens up (any frame can be selected and inspected symbolically in its own window), and a menu of break package commands is available (in addition to the entire programming language). If one invokes the editor from the break package (picking the function from the stack trace with the mouse), the source code is right there; the variables can be evaluated in the stack context of the break; and the comments are still there! I call that esteem, not contempt. Now an Interlisp break is not a post mortem. Within the break package one can change the variable; rewrite a function if desired and continue the program with the new definition, which is now THE definition. Do fixes made in cdb get incorporated into the original source automatically? [...] --Christopher ------------------------------ Date: Fri 5 Jul 85 18:01:35-PDT From: Wade Hennessey Subject: interlisp comments [Forwarded from the Stanford bboard by Laws@SRI-AI.] Chris' comments about the ease of debugging in INTERLISP are true of any lisp system. They are not really relevant to greep's dislike of INTERLISP or liking of UNIX. For example, ZETALISP uses a commenting style similar to C, and yet it provides the same debugging functionality of INTERLISP. There are several useful functions in ZETALISP which get the system to find the source file where a function is defined, and then let you start editing/examining the definition, comments and all. Thus, you needn't make comments part of the code to make them easily accessible at all times. Wade ------------------------------ End of AIList Digest ********************