Diplomacy Zine -- EP #215 Chapter Seven From: Eric_S_Klien@cup.portal.com Date: Sun, 30 Dec 1990 03:01:33 +0000 Issue #215 of ELECTRONIC PROTOCOL: ************************************************************************* "He's building a woman", Quesada said. "That's what Charley Norton told me. What's he using? A rag, a bone--?" "I gave him some surplus chemicals. Chosen for their color, mainly. He's got some foul green copper compounds and a little bit of ethyl alcohol and six or seven other things, and he collected some soil and threw it in a lot of dead shellfish, and he's sculpting it all into what he claims is female shape and waiting for lighting to strike it." "In other words, he's gone crazy", Barret said. "I think that's a safe assumption." ************************************************************************* Chapter One contains: BAGHDAD, AUSTERLITZ, BLITZKRIEG, KING'S GAMBIT, PASSCHENDAELE, GET SOME, DRAGONS, BLACK OCTOBER And is published by uunet!cti1!rlister or rlister@cti.com/Russ Lister Chapter Two contains: REPUBLIC, BORODINO, JACAL, VERSAILLES, DRESDEN, KHAN And is published by sinhaa@mcmaster.ca/Anand Sinha Chapter Three contains: DAWN PATROL, BERLIN, EL ALAMEIN, SQUALANE, UNGAWE, BRUSILOV OFFENSIVE, CULLODEN, GANDALF'S REVENGE, GOODBYE BLUE SKY And is published by cwekx@htikub5.bitnet/Constantijn Wekx Chapter Four contains: OZARK, DEADLY DAGGERS, YORKTOWN, MONTREUIL-SUR-MER, FIRE WHEN READY, THUNDERDOME And needs a publisher. Chapter Five contains: DEF CON 5, BORDEL, ERIS, MASADA, YALTA And is published by jjcarette@watami.waterloo.edu/David Gibbs Chapter Six contains: TOKUGAWA, BERLIN WALL, HIROSHIMA, GENGHIS KHAN, SEA LION, VIOLENT PEACE And is published by mike@suna.computation.umist.ac.uk/Mike Reddy Chapter Seven contains: HELM'S DEEP, GROUND ZERO, GIBRALTAR, TIBERIUS, BETELGEUSE, IRON CROSS, DEF CON 4, OPERATION DESERT SHIELD ------------- Chapter Seven ------------- Table of Contents: Part three of the article "An Interview With Fred Davis Jr." (Part two was in issue #214) Many items submitted by loeb@nestor.greco-prog.fr/Daniel Loeb. Information about Youngstown submitted by ken@dewey.cac.washington.edu /Ken Lowe. Ideas for variants submitted by tedrick@ernie.berkeley.edu/Tom Tedrick. Answer to Daniel Loeb's quiz by selham@pdc.kth.se/Magnus Selhammar. The following was scribed by Martin Snow/snow@cololasp.bitnet: Taken from Mouth of Sauron Volume Six Number One (English) AN INTERVIEW WITH FRED DAVIS JR. (part two) MN = Mark Nelson, FD = Fred Davis Jr. MN I guess Bruce Linsey would require a few tapes on its own. Can we condense Bruce down to a minute or two? FD Bruce is another person who is a dynamo as regards having energy. He's rather well built, he's about six foot, muscular, and in good physical shape. He has a good loud voice which can be sounded across a room. He seems to have his head tied on very tightly; he knows what he is doing. He has annoyed some people for certain things, he has called people late at night on occasions, he called me once when we're already in bed. When he gets excited and calls he just forgets what time it is! FD What he did with the Runestone Poll is just fantastic, two years ago he got approximately 250 votes, last year (1987) he got 441 people to vote in the ballot. Now there's only about 1000 active players in the whole American Hobby and to get 441 to vote is damn near incredible. Of course, he got almost every publisher to print the ballot both years, that helps. Certain people on the east coast hate him. They won't even mention the Runestone Poll in their zines, but that hasn't hurt the Poll. MN The Martins... FD Dick Martin has certainly worked hard, producing Retaliation and House of Lords. However, he's the kind of a person you cannot get a reaction out of. When you try and argue with him he just sits there and grins at you. It's almost as if he has some secret knowledge you don't have. When you argue with somebody they'll usually raise their voice or flash their eyes, but to have someone just sit there, not responding to you except to grin...this drives me crazy because you want a response of some sort. FD At the 1987 ORIGINS Con in Baltimore out of sheer frustration I said to him that "we'll fight you to the death on this matter of the MNCutC" Everybody else in the world except these three or four East Coast publishers still stick with Fred Hyatt as the legitimate Miller Number Custodian. But I have been told that Julie is actually the boss in the family. She is very, very attractive and has a very strong mind. When she doesn't like somebody she can hold a grudge for a long long time. MN How about the man who started the Hobby and is indirectly responsible for why we are here today, John Boardman? FD I have never traded with Boardman because I don't like the politics that he runs in his zine and because he hates Rod Walker. I sent him a copy of my first issue and I've seen various issues of Graustark at cons but as I don't like his attitude I don't trade. FD I've only seen him face to face about four times. He's in my age bracket, I'm the older and he's getting grey hairs and looks a little older than your average diplomacy player. He goes to conventions. He didn't go to ATLANTICON this year because he had a falling out with Sacks. He used to go to conventions and sit there and let Sacks make a fool of himself, like a monkey pulling the chestnuts out of the fire. While the flack is always directed at Sacks, I personally wonder if some of the suggestions may have come from Boardman. FD The funniest thing I can remember is at one DipCon when Boardman hadn't seen Rod in years and Rod had grown a beard. Rod ran up to Boardman, grasped his hand and shook it while Boardman didn't know who he was. After shaking his hand Rod told Boardman who he was and Boardman pulled his hand away in horror. ((Breaks down into laughter)) MN Perhaps we should finish with yourself, Fred. You've been in this hobby for 20 years. Do you have any last comments? FD I've enjoyed myself and something like Diplomacy was what I had been looking for for half my life. I started out in High School playing Fletcher Pratt's Naval Wargame with model ships on living room floors. Once we had a game similar to Diplomacy, we had a world map with four island continents and about twenty islands around the outside. We would have naval battles over the islands, which provided us with what we would now call supply centers. Although we didn't have that word, they provided the economic wherewithall for our countries. That proke up with the Korean War when two of us went into Service and we lost touch with the old gang. FD For years I hoped to find something like that, when I saw the Diplomacy board I knew from that moment I had found it. I ran around Baltimore the next day to see if anybody had the game, and then I rang Krock's Bookstore in Washington. It arrived on Saturday and I just studied the map for two or three nights, pushing wooden blocks around. I then wrote to Rod Walker and immediately joined one of his games, Imperialism VIIR--a six player game starting in 1937. FD I've enjoyed the friends I've made. Diplomacy is only the hook, only the excuse for publishing really. I've always liked to meet people and now when I travel around the world there's always either a Mensan or Diplomacy person who I can go and visit. This gets me to meet friends whereever I go and that's what I like about the friendships I've made. MN Thanks very much for your time, Fred. Until next time... ((Since this interview Fred has passed on the NAVB, and has re-established the NAVB-EAST. Bushwacker continues and looks set to become only the second zine to have been published without a break for twenty years. A remarkable feat.)) The following was submitted by loeb@nestor.greco-prog.fr/Daniel Loeb: DIPLOMACY PROGRAMMING PROJECT UPDATE ---- NOVEMBER 8, 1990 1) Ken Lowe writes and I reply: Be sure to include my suggested conversation in the progress report. Randy I is not ready to be incorporated into anything useful. I will send you a copy when it is. Are you referring to your random conversation (DMZ GAL ION NTH) or your mute conversation (HUH)? In either case, my reports make it clear that your program plays RANDOMLY, and thus is not really worth talking to. However, feel free to send me any clearification you think is necessary and I will send it to the list. [I am referring here to some sample conversations that Ken wrote. One inwhich RANDY always replies HUH and the other in which Randy choses random messages to send.] Your command syntax with the parenthesis appears to be almost lisp-like, but not quite. I think if it were rigorized a little parsing would be a lot simpler. I haven't looked at anything other than HLO, NOW and DST, which I assume the moderator will send me when the game starts. I expect something that looks like: HLO ENG NOW SPR 101 ( (AUS AMY BUD) (AUS AMY VIE) (AUS FLT TRI) (ENG AMY LVP) (ENG FLT EDI) (ENG FLT LON) (FRE AMY MAR) (FRE AMY PAR) (FRE FLT BRE) (GER AMY BER) (GER AMY MUN) (GER FLT KIE) (ITY AMY ROM) (ITY AMY VEN) (ITY FLT NAP) (RUS AMY MOS) (RUS AMY WAR) (RUS FLT SEV) (RUS FLT (STP STH)) (TUR AMY CON) (TUR AMY SMY) (TUR FLT ANK) ) DST ( (AUS BUD VIE TRI) (ENG LVP EDI LON) (FRE MAR PAR BRE) (GER BER MUN KIE) (ITY ROM VEN NAP) (RUS MOS WAR SEV STP) (TUR CON SMY ANK) ) Actually at the moment DI (Diplomat Interface) only sends out HLO at the beginning of the game. Then ORD and NOW at the end of each turn. Actually, sending out NOW at the beginning of the game is a good idea I'll put that on *my* list. I did not really DST to be used by my program. Perhaps it is a good idea. I could send it out at the beginning of the game at then after each turn (I could have it generated by the program update_sc). Although, perhaps I should give the command another name (for example, SCD (supply center distribution)). [The change to NOW, and the additon of SCD are now in the new version of the protocol. Ask for a new copy if you are interested.] Note that I added a level of parenthesis to the DST command (and perhaps to the NOW command). If anywhere there's a list of a variable number of parameters the list was included in parenthesis parsing would be trivial. Without the extra set of parenthesis on the DST command I was not able to tell when the DST command ended without scanning well ahead. To be more lisp-like the parenthesis should be moved ahead of the command keyword: (DST (AUS BUD VIE TRI) (ENG LVP EDI LON) ... ) and then (HLO ENG). Since the project is not to write a context free parser, using the slightly more rigorous command syntax might speed things up. What do you think? Finding the end of the command is easy. Each message is terminated by an end-of-line character. I would like to make a protocol that can be understood either with parenthesis or without. That would give programs a better choice. [So far no official change on parenthesis has been made to the protocol file. I'd like to get more feedback from all of you, and some more concrete suggestions.] Also, I need a reminder as to why the year is 101 rather than 1901. Is it just because you don't want to parse 4 digit numbers or was there a typo in your protocol spec? In the same vein, why (STP NTH) rather than (STP NC), the latter avoiding the conflict with the abbreviation for the North Sea. I'm trying to have all words in a message be exactly three characters long. This includes the date! This could prove useful for some parsers I imagine. [The new protocol uses NCS ECS and SCS instead of NTH STH and EST in order to avoid conflicting with the North Sea.] I'm also assuming that your seasons SPR SUM FAL AUT WIN are played in that order (it took me a while to distinguish between FAL and AUT, the real rules only refer to spring and fall seasons with multiple phases per season (three in the fall)). The rulebook doesn't use these terms. However, the gamer's guide (published by Avalon Hill), and most diplomacy magazines do use this notation. I think it is a much simpler way of thinking about the game. Yours, Daniel Loeb, DPP "Combinatorics: The Thought that Counts" The following was submitted by loeb@nestor.greco-prog.fr/Daniel Loeb: DIPLOMACY PROGRAMMING PROJECT UPDATE ---- NOVEMBER 9, 1990 1) I will be on vacation in the United States, December 6 through January 1. I would like someone to cover the discuss relating to this project during the interim. Any volunteers? In case, anybody wants to speak to me personally, here is a rough idea of my schedule Dec 06 - 09 New Brunswick, NJ Dec 09 - 10 Murray Hill, NJ Dec 11 - 16 Boston, MA Dec 16 - 17 New York, NY ? Dec 17 - 18 Philadelphia, PA ? Dec 18 - 31 Chicago, IL Finally, my wife expressing some interest in writting a diplomat during her maternity leave starting at the end of February. I'll let you know if anything comes of that. 2) More discussion with Ken Lowe. First his letter: > Finding the end of the command is easy. Each message is terminated by > an end-of-line character. This means that if you want to generate a file full of commands you have to use very long lines. Note that a fully configured NOW command could extend to 472 characters. Most terminals are limited to 80 characters. Many editors have limits of 150 characters per line. Most have limits of 256 characters and I bet there are very few that would allow as many as 1024. I can picture communications where a single message will extend well over the 1024 mark. Even if you can have such long lines, trying to look at a transaction log to see why your program did what it did is going to be a bear. Being able to use the lisp "pretty print" algorithm would help immensely. > I would like to make a protocol that can be understood either with > parenthesis or without. That would give programs a better choice. But I don't want to write such a general purpose parser. You've already got parenthesis up the yinyang in your protocol, I propose making it a rigorous definition such that the input could be fed directly into a lisp list. I would like to generate and receive such input. If you can support that while allowing other people to look for end of lines, fine. I'd like to look at end of lines as just "white space" that is equivalent to spaces and tabs. How about: SND ENG (DST (ENG 0) (FRA 18) DWR) On the subject of long messages, you indicated somewhere that if you want to guarantee that messages be sent in a particular order that you should send them with one SND command. In another message you said that your system is going to be delivering messages in a random order. However, you have the DFM "discard following message" which sort of requires that messages be delivered in a specific order. Perhaps DFM should be stricken since you don't have a syntax of SND or FRM that specifies more than one recipient of a message anyway. On the subject of commands in conversations, I see the need for two more commands: IOU and UOI to be used for bartering. An expected conversation using these commands could go something like: ENG: (SND FRA (UDO ((AMY PIC) SUP (AMY EDI) MTO BEL))) FRA: (SND ENG (IF (IDO ((AMY PIC) SUP (AMY EDI) MTO BEL)) THN (UOI 100))) ENG: (SND FRA (IF (UDO ((AMY PIC) SUP (AMY EDI) MTO BEL)) THN (IOU 50))) FRA: (SND ENG (IF (IDO ((AMY PIC) SUP (AMY EDI) MTO BEL)) THN (UOI 75))) ENG: (SND FRA (IF (UDO ((AMY PIC) SUP (AMY EDI) MTO BEL)) THN (IOU 75))) FRA: (SND ENG (IDO ((AMY PIC) SUP (AMY EDI) MTO BEL))) If France actually comes through and supports Edinburgh to Belgium then France and England will remember that England owes France 75 "points". A certain amount of bartering is required due to differences in what the two powers will consider important as well as exchange rates. England will have to remember that France considers the ownership of a supply center worth 100 points when he considers it worth 50 points. It wouldn't do if France were to turn around and ask England to support him into Belgium and England offer to do it for only 50 points. The value in the IOU and UOI commands would be accumulated debt for the history of the game. Of course at the beginning of the next season after France supported Edinburgh to Belgium the conversation might go: FRA: (SND ENG (UOI 75)) ENG: (SND FRA (IOU 0)) FRA: (SND ENG (IF (UOI 0) THN DWR)) ENG: (SND FRA DWR) Alternatively, a negative number on the IOU command rather than having a UOI command could make processing easier. > I'm trying to have all words in a message be exactly three characters > long. This includes the date! This could prove useful for some parsers > I imagine. Does this mean your alternate format of the DST command and the FOR command should have leading zeros on values less than 100? I think allowing variable length numbers is not a problem. Particularly with the bartering commands I describe above where people might want to use a larger range of integers than 0-999. -Ken 3) Before commenting on Ken's suggestions let me describe what happens when a message is sent to and from my diplomat interface (DI). Let's say the English diplomat (programmed in C) outputs the message (using fprint): SUB (FLT LON MTO NTH) followed by an end of line. This is immediately handed over to LCS by the UNIX operating system which then leaves it in an LCS input stream. The message waits there while the previous messages are handled by DI in order. DI contains an agent devoted to handing messages from this stream to the processing routine. When this messages turn comes around, the agent attempts to create a rendez-vous with the processing routine along the port (out_from#"ENG"). Only after this rendezvous is created and the message is passed, will the agent proceed to the next message (if any). Therefore, input is done on a sequential basis by player. After this rendezvous, the DI routine PROCESS hands the ball over to other agents (in particular SUBMIT) which handle the submitted move and check its validity. After a bit of calculation it is determined this this is a valid move, and the reply THK ((ENG FLT LON MTO NTH) MBV) with an end of line is devised. The SUBMIT agent now demands a rendezvous along the port (inp_for#"ENG"). However, this demand competes with all other simultanous demands for rendezvous (for example France may be attempting to send a message to England). LCS handles these rendezvous requests on a random basis. Thus, output is asynchronous. Eventually, a rendezvous is created with the agent responsible for sending output to England. At this point, two things occur: the routine READY) checks to see if all moves are in, and the English output agent puts the message in an output stream. Every time the English diplomat demands input (using finput since it is a C program), characters are taken from this stream in the order it was put in. If the stream is empty, then the input command will procede as if no input was currently available. 4) And now my response to Ken's letter. **** RE: LINE LENGTH In order to avoid dealing with long lines, several solutions are possible. 1) each diplomat input characters one at a time instead of a line at a time. 2) each diplomat is specified along with its preferred maximum line length. The output routine then chops up long messages into bite sized chunks. I propose that the chopped up messages be delivered with the following syntax: LNG n line1 line2 line3 ... linen EOM when (LNG n) means "look out, a long message of n lines is coming". And (EOM) means "end of message". I would cuts the line between words. Conversely, long input can be delivered according to the same protocol and then be pasted together before being delivered to the PROCESS routine. **** RE: PARENTHESIS I can see how a parenthesis system understandable by lisp would be handy for diplomats written in lisp. Not being an expert on standard lisp itself. I leave the necessary modifications to Ken assuming nobody objects to the changes he makes. I remind people again that we are making a lot of changes now to the protocol, but that once the first "real" diplomat is submitted, we will try to make only "upward-compatable" changes. **** How about: SND ENG (DST (ENG 0) (FRA 18) DWR) The syntax there is a bit wrong. You are sending the message: DST (ENG 0) (FRA 18) DWR to England. What you really want is SND ENG (DST (ENG 0) (FRA 18)) (DWR) which sends the following two messages to England: DST (ENG 0) (FRA 18) DWR **** Fixing the order of messages. My comment on using the SND command with multiple argument to fix the delivery order of messages was wrong. As I explained above, attempts to gain the attention of the output agent are done in a random order. The following procedure should be used to send a series of messages in a fixed order. France sends the first message: SND ENG (msg1) After the SEND agent passes to the output agent for England the message: FRM FRA (msg1) It passes to the output agent for France the message: SNT ENG (msg1) as a receipt. Now, France can send a second message. It might arrive at the output agent for England before the first message is read, but France is guaranteed that it will occupy a latter position in the stream, and thus be read after the first message. Successive messages continue in the same manner. **** IOU and UOI I agree that IOU could be a useful command, but it would have to be classfied in the advanced section since the scaling problems are troublesome. **** Integers Ok, I'll allow integers not necessarily 3 characters long. But commands using such integers will be classified as "advanced." (Most of the above changes are now noted in the newest version of the protocol.) Yours, Daniel Loeb, DPP "Combinatorics: The Thought that Counts" The following was submitted by loeb@nestor.greco-prog.fr/Daniel Loeb: Although Judge has supported the Youngstown variant for a while now, it is only now beginning to offer adequate documentation in terms of rules and maps available to print-out. Here is a status report by Ken Lowe: I need to get some clarifications on the rule changes that Erich Schmidt gave me before I make a formal announcement, but you might want to take a look at what I've got so far (the map is not really on the judge yet). The announcement will probably come out early next week. I haven't actually tried it on an Apple LaserWriter (just a DEC LPS20 and a NeXt). If you can time it I'd be interested in how long it takes for each page to print (I'll also need to know what baud rate your printer is set at). The LPS20 and NeXT printers take about 15 seconds per page or less, I suspect the ALW will be slower. -Ken ---------- The draft announcement: This is a general mailing to all the players and observers signed up via the judge at judge@u.washington.edu. I have finally gotten my hands on a Youngstown variant map (thanks to The Jade Piper). I have transcribed this map to PostScript and it is now available via the judge with the command: get map.youngstown.ps The file is 124Kb, so be sure you have a PostScript printer before you ask for the file. This map can be scaled to print on a single sheet or on multiple sheets which can then be taped together. A 2x2 map on 8.5x11 inch paper looks pretty slick, if I do say so myself. A 4x4 page map would scale to the size of the original Schenectady map (3x4 feet), but that's probably more than you need. If anyone prints anything bigger let me know how it looks at larger scales. I had to take some liberties with Japan in order to make it readable on a single sheet. The result is that Japan is about the same size as Italy, but that should only be a problem to purists. I also redid the offboard boxes so that adjacencies are a little more obvious. I'm sure that the availability of the map will increase the interest in the variant. The judge currently has two Youngstown variant games forming, "Dien" which will be monitored by Danny Loeb and "Young" which will be unmonitored unless there's a volunteer. To sign onto the Dien game, send the judge a message containing: signon ?dien <password> youngstown For more information send the following to the judge: get info.youngstown A description and specific rule changes. get map.youngstown A list of allowable province abbreviations. -Ken ------ The map: %! Youngstown diplomacy % % Copyright 1990, Lowe. % % Based on the original Schenectady Wargamers Association map % by Patrick Stevens and John Fitzpatrick. % % Diplomacy is a trademark of the Avalon Hill Game Company, Baltimore, % Maryland, all rights reserved; used with permission. % % Redistribution and use in source and binary forms are permitted % provided that it is for non-profit purposes, that this and the % above notices are preserved and that due credit is given to the % authors. % % The following parameters can be modified to change the scale of the % map. A single 1x1 map is easy for quick reference, a 2x2 map looks % a whole lot spiffier. A 2x1 (two rows, one column) is significantly % better than a 1x1 map. A 4x4 map will be at the same scale as the % original Schenectady map. It would be very usable, but discrepancies % in the (by hand) digitization may become apparent. % % Specifying multiple copies here rather than attempting to print it % twice will save printer time, particularly with larger scales. Be % sure your printer is in simplex rather than duplex mode if available! /rows 2 def % Number of rows of pages to print /cols 2 def % Number of columns of pages /copies 1 def % Number of copies of each page % The following parameters can be tweaked to adjust the map for various % differences in particular printers. /inch { 72 mul } def /lmarg .25 inch def % Margin at left of physical page /bmarg .25 inch def % Margin at bottom of physical page /pgh 10.5 inch def % Physical height not counting margins /pgw 8.0 inch def % Physical width not counting margins /wgray { .98 setgray } def % Gray scale used for water /mgray { .95 setgray } def % Gray scale used for mountains /tline { .9 setlinewidth } def % Thin line linewidth /fline { 4.5 setlinewidth } def % Fat line linewidth % The following setscreen is required for the NeXT, but doesn't seem to % hurt other things. If you get funny gray tones try commenting it out. A million more lines later... 0 1 xlm 1 sub { /x exch def 0 1 ylm 1 sub { /y exch def gsave xsh x mul sfactor div neg ysh y mul sfactor div neg translate map newpath 0 0 m xdot 0 l xdot ydot l 0 ydot l closepath tline stroke copies { copypage } repeat erasepage grestore } for } for The following was submitted by loeb@nestor.greco-prog.fr/Daniel Loeb: 1) Your signup list still doesn't indicate DIPLOMAT and Ken Lowe's JUDGE and my DIPLOMAT INTERFACE as excellant (and free) programs. I think most of your readers would prefer a free program with the same functionality. 2) You should distinguish between "robot moderated games" and "robot assisted games". The easy way to tell the difference is that a robot assisted game is marked "MODERATED" in judge's listings. Another difference is that the robot assisted games have a human who is willing to intervene, and help you out, etc. Of course, Ken is helpful for all games on the judge. But he is more of a robot manager than a GM assisted by a robot. In acknowledgement of that, games inwhich Ken is the only "master" are marked as unmoderated. (Except for FONTENOY which Ken is going to pay special attention to.) 3) Your signup list of FONTENOY is missing about a dozen players! Send LIST FONTENOY to judge to get more information. 4) Why is my friend not yet listed on the waiting list for French games? 5) You've already published the YOUNGSTOWN rules. 6) While we are waiting for Fontenoy to start, SEND ME PLAYERS for one of my small games on judge at washington...... Why am I not on your list. 7) You can list me as a backup GM (with Judge's help). 8) Are you subscribed to CONTROL so that you can see when new games start? Thanks Yours, Daniel Loeb "Combinatorics: The Thought that Counts" The following was submitted by loeb@nestor.greco-prog.fr/Daniel Loeb: DIPLOMACY PROGRAMMING PROJECT UPDATE ---- NOVEMBER 7, 1990 1) Here is a note from Ken Lowe: "In order to not waste away in this effort to the exclusion of all else, I have budgeted one night per week max to Randy. Randy actually generated some moves this weekend. Not too bright, the units in England spun around in circles while nearly everyone else bounced. That's to be expected though where my selection critera was "The first movement in the adjacencies table". The next step is in evaluating the moves to decide if anything looks better than something else. A coworker appears interested so we will probably end up with some sort of joint project." When he sends me RANDY-1, I'll test it along with my program and show you the results in order to give you an idea of how the protocol works. 2) Michael Hall and I are still working on that article you've seen. Send us your comments. Yours, Daniel Loeb, DPP "Combinatorics: The Thought that Counts" The following was submitted by loeb@nestor.greco-prog.fr/Daniel Loeb: To facilitate distribution of DPP files, I have given the Diplomacy Adjudicator program access to these files. The update version of "table.of.contents" reflects this change: Diplomacy Programming Project table.of.contents Updated: 29 November 1990 Thanks for your interest in the Diplomacy Programming Project. The best way I can tell you more about it, is to send you a sample of some of the files relating to the project.... (The files I didn't send, you can of course request by sending mail to JUDGE@NESTOR.GRECO-PROG.FR containing the line "GET DPP/file" for each file you wish to have sent to you (where "file" is the name of the file). name topic article.tex A better version of "hall.article" (in progress) conversations An appendix to "protocol" giving examples of syntax course How to get credit for your participation dipl.sml The Diplomat Interface progress written in LCS Program is in many parts including: datatypes.sml declarations dipl.bugs list of bugs to fix dipl.comments general comments about the program draw.sml handling of the DRW command endofturn.sml production and distribution of results general.sml a package of general purpose functions init.sml initialisation routines map the standard diplomacy map map.sml the parts of init.sml which depend on the map process.sml the "front-end" of the program rules AH rulebook with commentary regarding DPP strings.sml string "beautification" package submit.sml handling of the SUB command hall.article Article by Michael Hall on programming a diplomat intro First proposal for DPP protocol The language in which Diplomats communicate snake Rules for a simple multiplayer game strategy A simple diplomacy strategy table.of.con... This letter update.sep1,2,3 progress reports update.oct1,2,3 progress reports update.nov1,2,3 progress reports I'm putting you on the mailing list for future reports. I hope you decide to write a program for use with the project. Yours, Daniel Loeb, DPP "Combinatorics: The Thought that Counts" The following was submitted by ken@dewey.cac.washington.edu/Ken Lowe: This is a general mailing to all the players and observers signed up via the judge at judge@u.washington.edu. I have finally gotten my hands on a Youngstown variant map (thanks to The Jade Piper). I have transcribed this map to PostScript and it is now available via the judge with the command: get young.ps The file is 125Kb, so be sure you have a PostScript printer before you ask for the file. Multiple people at the same site should ask for it once and share it. This map can be scaled to print on a single sheet or on multiple sheets which can then be taped together. A 2x2 map on 8.5x11 inch paper looks pretty slick, if I do say so myself. A 4x4 page map would scale to the size of the original Schenectady map (3x4 feet), but that's probably more than you need. If anyone prints anything bigger let me know how it looks at larger scales. I had to take some liberties with Japan in order to make it readable on a single sheet. The result is that Japan is about the same size as Italy, but that should only be a problem to purists. I also redid the offboard boxes so that adjacencies are a little more obvious. However, there is a bit of confusion on the movement rules associated with offboard boxes and Suez. Until I get an official copy of conflicting rules the judge will interpret things as described in the "info.youngstown" file as provided by Jon Monsarrat. The boxes on my map list the provinces and the other offboard boxes that you can move to based on Jon's rules. The conflicting rules prohibit movement from one box to another. I'm skeptical in that the conflicting rules make it sound like an Indian fleet in the South Atlantic offboard box would cut France and anyone else in the northwest off from the Indian Ocean and no one could ever dislodge it. If anyone wishes to clarify this feel free. I'm sure that the availability of the map will increase the interest in the variant. The judge currently has two Youngstown variant games forming, "Dien" which will be monitored by Danny Loeb and "Young" which will be unmonitored unless there's a volunteer. To sign onto the Dien game, send the judge a message containing: signon ?dien <password> youngstown For more information send the following to the judge: get info.youngstown A description and specific rule changes. get map.youngstown A list of allowable province abbreviations. While I've got your attention, I would also like to point out that the "chaos" variant game "fontenoy", where each of the 34 players starts out with a single supply center is waiting for a mere 6 more people to sign on. We're allowing two people from the same site to sign on, but they'll be given centers far away from each other. The NIMR and ERIM folks are conspicuous by their absence (maybe they're the intelligent ones). The first deadline, assuming it fills out earlier, will be January 10th at the earliest. It promises to be an interesting diversion at the least. -Ken The following was submitted by tedrick@ernie.berkeley.edu/Tom Tedrick: One variant you might add to the list is "armored" diplomacy. Players can build tanks at a cost of 2 regular builds. "Tanks" move with force 2. I believe Matt Crawford has the complete rules for tanks. Another game that you might add to the list is "Machiavelli". Haven't been many Machiavelli games lately, and Machiavelli is pretty similar to diplomacy (maybe closest to the "1914" variant). -Tom tedrick@ernie.berkeley.edu The following is the answer to this quiz published in issue #211: From loeb@geocub.greco-prog.fr/Daniel Loeb: "How many different sets of opening orders are there? Such a set of orders must specify a legal movement, support, or hold for each of the 22 units on the board. (Convoys are of course impossible in Spring 1901.) You must consider all possible movements including stupid ones like moving Smy->Syr or Lon->Yor with Edi->Yor. However, you should only count supports which are NOT void. That is to say, the supported unit must be ordered to do what is supported to do. Furthermore, only count useful supports. That is to say, only count supports given by a country "A" for himself or a country "B" which in theory could prevent the success of the move or hold by a country "C" distinct from "A" and "B"." Answer to Daniel Loeb's quiz by Magnus Selhammar/selham@pdc.kth.se I have calculated the number of opening moves, and the the number of opening moves suggested in The Gamers Guide to Diplomacy. In order to forecome Daniel Loeb's next quiz, I have also calculated the number of possible orders for a typical Winter 1901 position, and the number in the sample game at the Spring 1904 positions in The Gamers ... Please take these numbers as approximate, it's very likely that I have made errors. I have counted all possible orders, even orders such as: A Nwy - Fin F Nwg C A Nwy - Den F Nth C (Ger) A Den - Cly F Edi - Cly I have not counted holding supports to empty territories, but all supports to an empty territory from an occupied territory wherefrom an unit can enter from. Even though most of the orders are improbable and bad, (but not illegal), the Winter 1901 position I have calculated on is quite plausible. The number of opening moves are 1.84*10^22. If one use the suggested moves in The Gamers ... the possible opening moves there are 46656. In the typical situation at Winter 1901 there are 4.93*10^35 possibilities, and in the sample game at Spring 1904 there are 6.92*10^44 possible sets of orders. In the "typical" position at Spring 1901 the positons were: (Belgium is still a neutral supply center.) Austria A Tri Vie Bud Ser F Gre England A Nwy F Edi Nwg Nth France A Par Bur Por Mar F Spa(sc) Germany A Ruh Den Mun F Hol Kie Italy A Ven Tun F Nap Ion Russia A StP War Ukr Sev F Swe Rum Turkey A Bul Ank F Con Smy In these situations the convoy possibilities are rather limited. They are the main reason why the number increases so much. With more fleets on the water the number will raise considerable. As an example, in the sample game there are 63 different legal orders for the fleet in MAO. In chess there are most often less than ten "good" moves. In Diplomacy, there are many more, often thousands of them. The problem of writing a program is then quite different and much more difficult than chess, even when one uses parallellism as Michael Hall suggested, and a great challenge indeed. ((Danny's solution follows, I'll post any comments that Danny and Magnus make about their different answers. Danny's answer is about a million times smaller than Magnus's. Although 10^22 and 10^16 are in the same ballpark of big numbers.)) First we must could for each unit the number of neighbors it has and could move to. Obviously, besides supporting, a unit can move to one of these spaces or it could hold. Number of Neighbors Edi Lpl Lon Bre Par Mar Ven Rom Nap Tri Vie Bud 4 4 4 4 4* 5* 6* 4 3 3 5* 5* Con Smy Ank Mun Kie Ber Sev Mos War StP 3 4* 3* 7* 3 4* 3 5 6 3 Next we must consider the map and determine which are the "useful" supports I was referring to. The only "useful" support on nation can give another, you will not is that any of Austria, Italy, and Germany can support a second to prevent the third from taking Tyrolia. In fact, it is possible for two of these countries to so support the third! Useful triples: country units dest France: Par,Mar Bur Germany: Mun,Ber Sil Turkey: Smy,Ank Arm Austria: Vie,Bud Gal Vie,Bud Tri G+A+I: Vie,Mun,Ven Trl Thus, answer is the following product: ANSWER = (3+1)^6 Nap, Tri, Con, Kie, Sev, StP 4096 x (4+1)^5 Edi, Lpl, Lon, Bre, Rom 3125 x (5+1) Mos 6 x (6+1) War 7 x (4+1)(5+1) + 2 (Par*Mar) 32 x (4+1)(3+1) + 2 (Smy*Ank) 22 x A (Mun*Ber*Vie*Bud*Ven) 13174 Where A is the number of possible combinations for Munich, Berlin, Vienna, Budapest, and Venice. We break up A into the possibility that Vie and Bud support each other to Gal or Tri, and if Mun and Ber support. A = ( 4 Vie*Bud 4 x ((6+1)(7+1) + 2)Ven*Mun x58 x (4+1)) Ber x5 = 1160 + ( 4 Vie*Bud 4 x (6+1) Ven x7 x 2) Mun*Ber x2 = 56 + ( (5+1) Bud 6 x 2 Mun*Ber x2 x ((5+1)(6+1) + 2)Vie*Ven) x44 = 528 + ( (5+1) Bud 6 x (4+1) Ber x5 x ((5+1)(6+1)(7+1)Vie*Mun*Ber x(336+12+14+16+3) + 2(5+1) = 11430 + 2(6+1) + 2(7+1) + 3) Thus, the final answer is 4,985,969,049,600,000. At a rate of 1 game per person in the world per minute. It would take about 2 years to try every possible diplomacy opening! Yours, Daniel Loeb "Combinatorics: The Thought that Counts" Publisher comments: Quote is from P. 18 of Hawksbill Station by Robert Silverberg ****************************************************************************** To join in the fun, send your name, home address, home and work phone numbers, and country preferences to Eric_S_Klien@cup.portal.com. ****************************************************************************** Up