Diplomacy Zine -- EP #191 Chapter Seven

From: Eric_S_Klien@cup.portal.com

Date: Tue, 14 Aug 1990 03:24:30 +0000

Issue #191 of ELECTRONIC PROTOCOL:

************************************************************************
"Logic is a tweeting bird, chirping in your ear.
 Logic is a wreath of pretty flowers that smell *bad*."
************************************************************************

Chapter One contains:

BLITZKRIEG, GETTYSBURG, RED STORM RISING, and PASSCHENDAELE

And is published by daybell@aludra.usc.edu/Donald Daybell

Chapter Two contains:

DRAGONSLAYER, JACAL, MANHATTAN, VERSAILLES, DRESDEN, and KHAN

And is published by tedward@cs.cornell.edu/Ted Fischer

Chapter Three contains:

MULHOUSE, DAWN PATROL, SNIKKEL-2, BERLIN, SNIKKEL-1, EL ALAMEIN,
SQUALANE, UNGAWE, CAPTAIN CAVEMAN

And is published by cwekx@htikub5.bitnet/Constantijn Wekx

Chapter Four contains:

NICKEL, OZARK, DEADLY DAGGERS, YORKTOWN, MONTREUIL-SUR-MER

Chapter Five contains:

ARCHANGEL, BORDEL, ERIS, MASADA, and YALTA

And is published by bmers58!dgibbs@bnrgate/David Gibbs

Chapter Six contains:

TOKUGAWA, BERLIN WALL, HIROSHIMA, GENGHIS KHAN, SEA LION, VIOLENT
PEACE

And is published by sinhaa@mcmaster.ca/Anand Sinha

-------------
Chapter Seven
-------------

No games in this issue.

Publisher comments:

Quote is from the "I, Mudd" episode of Star Trek.
Submitted by by AS365156@Orion.YorkU.CA code named Video Dienstag.

More from loeb@rita.laas.fr/Daniel Loeb:

An extra note to my Diplomacy Programming Project Proposal.
If any of you students would like to participate, I would
be glad to write to your school and tell them that as a
professor at the University of Bordeaux, I am supervising your
work, and that they should give you computer science credit
for your program.

A potential warm up exercise follows in another message.

More from loeb@rita.laas.fr/Daniel Loeb:

I've already gotten quite a bit of response to the Diplomacy
Programming Project. However, I don't know how soon I'll have 7
working programs to run against each other.

In the meantime, could you tell me what house rules you will be using
concerning computer PROGRAMS who want to compete against human being
in the Electronic Protocol.  ((I would have seperate categories for
open to all games.  In general, my zine is racist and discriminates
against non humans.))

I suggest that you allow people signing up for a diplomacy game to
specify "humans-only". The default would be "open to all". In GUNBOAT,
it is not revealed that a computer program is playing. In other
variants, the programmer is encouraged to print a list of vocabulary
the other players should use in communicating with the program.
No program can control more than one country in the same game, nor can
there be two programs in the same game written by programmers at the
same site.

More from loeb@rita.laas.fr/Daniel Loeb:

DPP Update:

1) Richard Kirby has told me that it should be possible to have all of
the programs written in the language of the players choice using UNIX
ports to exchange information. In this light, my choice of LCS as the
language was somewhat arbitrary. Michael Hall suggests that some
programmers would prefer a rule-oriented language.

2) Until we have 7 programs working, I suggest we run games involving
human players as well. I have asked Eric Klien for permission. If we
end up with more than 7 programs, than we can devise complicated
tournaments.

3) Should programs know which other programs control the other
countries asks Jack Bennetto, so that they could learn each others
habits or hold grudges. And if so, given the fact that these programs
will be frequently changed, when will a program be considered new
(from the point of view of grudges). My suggestion that is that
programs should be allowed to remember other programs for the length
of any one tournament.

4) Any suggestions on how to find an adjudicating program with which
to hook all the 7 programs together? I've asked Ken Lowe for help in
that category.

5) I am now trying a Prisoner's Dilemna tournament by way of warming
up for a Diplomacy tournament. Tell me if you want to enter.

6) Here is a sample protocol programs can use to talk to each other.
Any comments:

MESSAGE PROTOCOL:

Message			Meaning (example)
TRY msgs		I do not understand the message you sent, following
			is a list of messages I can undestand.
			(TRY WAR PCE UNO UDO)

WAR			I declare war on you
			(WAR)

PCE			I declare peace on you.
			(PCE)

DMZ prov		Let's declare "prov" a demilitarized zone
			(DMZ BUR)

UNO order		Please do not submit the following order
			(UNO BUILD F STP(NC))

UDO order		Please submit the following order
			(UDO A WAR - PRU)

INO order		I promise not to submit the following order
			(INO A MAR S A VEN - PIE)

IDO order		I promise to submit the following order
			(IDO F ECH - MAO)

EXP [msg]		You broke your promise (indicated)
			(EXP UNO BUILD F STP(NC))

SRY			I am sorry
			(SRY)

FOR turn msg		The following message refers to potential orders
			during the indicated turn
			(FOR 15 UDO BUILD F SEV)

IFF m1 THN m2 (ELS m3)	If you m1 [either UNO or UDO probably] then
			I will m2 or else I will m3
			(IFF FOR 15 UNO BUILD F STP(NC) THN WAR ELS FOR
				16 IDO A SWE S F STP - NWY)

YES (msg)		I agree. Message received.
			(YES)

NOP (msg)		I do not agree.
			(NOP)

RPT			Please repeat message or send a new one.
			(RPT)

WHT unit		Suggest a(nother) move for the indicated unit
			(WHT F NTH)

HOW unit/prov		Suggest a plan of attack
			(HOW MOS)

FRM nation msgs		The following message were sent to me from the
			indicated nation
			(FRM RUSSIA (INO BUILD F STP(NC)))

SND nation msgs		Send the following messages to the indicated nation
			(SND RUSSIA (UNO BUILD F STP(NC)))

END			End of message. Ready to submit orders.
			(OUT)

PPT			Offer to become a puppet
			(PPT)

From Jack Bennetto <bennetto-jack@CS.YALE.EDU>:

A few suggestions on the sample protocol:

Is PCE used to end a war or form an alliance?  I think these should be
seperate.  PCE would propose a non-aggression pact, while something
else (ALY?) would propose an allience.  A more tentative suggestion:
ALY ENGLAND might propose an allience against England, ignoring (say)
Italy.  Not only would this allow more to be communicated, but I think
the message would be easier for the program receiving it to deal with.

There should be some way to deal with long term goals, such as a message
suggesting an eventual supply center distribution.  For instance, after
the alliance against England was formed, Germany might send:

DST GERMANY BEL EDI FRANCE LON YOR

France could then judge the distribution and reply.  Otherwise it would
be difficult for Germany to convince France to support him into Belgium.
A simple alliance is somewhat vague and France has no way to refer
back to previous aid ("I got you Belgium, I should get the next center.")
(This could be added but it might overly complicate things.)
DST might include other players as well, and non-supply spaces:

DST GERMANY BEL EDI NTS FRANCE LON YOR ECH RUSSIA NWY

How do you respond negatively to a HOW message?  Maybe:

NOP(HOW MOS)

meaning no, I don't think we sould take Moscow.  (Or we aren't able to
take Moscow?)

Should EXP include some turn number or message number?  Or even YES,
NOP, and SND?

Should there be some way to predict opponents moves?  Or the success
or faliure of yours?  Maybe:

THK ENGLAND F ECH-BRE (I think England will attack Brest with the EC)

WIN A RUH-BEL (I think that move will succeed)
LUZ F PIC S A RUH-BEL (But he'll cut you're support)

This would also allow one program to judge the skill of another and deceive and take advatage of less intellegent programs, making it
difficult for a program to exist on the skill of its allies.  Supports
and supply cuts could be listed after the attack to imply that the attack
will succeed if thess other things are done.

A SNT message might tell someone the you sent someone a message, which
whould make three-way alliances easier.  Alternatively, messages could
be sent with multiple addresses, guarenteeing that your program sent the
message to the other allies. (leading to, of course, DFM, disregard
following message...)

I'm not sure whether UDO and UNO should be part of IFF statements.
UDO and UNO seem to imply more the promise then the action.  Just
semantics, really.  How would your system say "If a third party does
this then I will do this"?

That's all I can think of for now.  Some might complicate things to much, although you have included the TRY command to allow less sophisticated
negotiators to compete.  I feel pretty strongly about something like
ALY and DST, though.

One way to choose protocols would be to play (or at least start) a test human game using only your orignal ones, which might suggest a better
set of protocols and some ideas for writing programs.

I sense I might end up getting drawn into this if I'm not careful.

Jack Bennetto

More from m.i.nelson@gdr.bath.ac.uk/Mark Nelson:

Dear Eric,
          Work on my ratings article is progressing, and an end of
the actual rating process is in sight.  WHat I need is a list of
UNFINISHED GAMES (particular any from 1989) and the ORIGINAL
PLAYERS.  It would also be useful (butnot essential) to have a
similiar list for unfinished 1990 games. (I rate games on a year
by year basis).  ((I sent them.))

Youhave yet to run any adventurous variants, mainly regular
board with minimal rule changes. I guess a problem is that players
can't get maps fromnew games wioth the zine!  So, snailmai zines
have a slight advantage!  Still, I would expect thatwould be
a sufficient interest in m say, DOWNFALL even without including
maps.  When I got around to start running my variant CHAPTER for you
then I shall try to encourage people to play variants which are on
different maps.  ((I snail mail variant maps, it is no problem.))

More from m.i.nelson@gdr.bath.ac.uk/Mark Nelson:

Dear Eric,
          I cannot agree with Daniel Loeb in that kibitzing would
make a good diplomacy variant on its own.  It is used in RATHER
SILLY DIPLOMACY, but is combined with many other obtions and ideas.
Onit's own I don't think that the idea has sufficient appeal
to establish itself has a `good' variant (whatever one of those
maybe!). Certainly, I wouldn't play in such a game.

In the version suggested (no press exept through the kibitzer)
then I don't believe that the kibitizer would have a very high
chance of `winning' the game. Thelikely outcome is that there
would be NO communication between players, and therefore the
kibitizer would have no influence over the course of the game.
i

The version when the Kibitzer is not allowed to alter
messages wouldprobablynot work, it has a chance if the GM
is only GMing the one game.  SImply it would be too time
consuming for the GM to physically check that the original message
and the kibitzer sent message are exactly the same.

If Daniel could work this idea into a `larger' variant (combining
some other different ideas) then there might be a basis for a new
variant.  As it stands, I don't thinkthatthere is sufficient
interest in the game for it to be anythingother than another
variant gathering dust in someone's files!

SUPER DIPLOMACY.  There is already something similiar to this,
CHAOS I by DOn Miller.  Julie Martin's EVERYBODY PLAYS also has
a similiar theme, although her game takes things to extreme!
In these games there is doubtas to which country you are playingat
any time.

Another similiar idea is 1492 in which the GM draws up an extension
to the starting map (which isn't the same as in regular dip) prior
to the game start. The players then discover their world as they
explore it, moreover you're only told about what you find--the
turn report does not specify what other people have discovered.

The concept of discovering the map during play has been used in
one or two other designs but has not proven to be popular, I think
that most designers considered that it added toomuch luckinto
the game.  Personally, I quitelike the idea.

FInally there is SOMEWHAT DEMIURGIC diplomacy in which players
add new rules as the game progresses, this can make for a very
entertaining\silly game (take your pick) as the more imaginative
players comeup with some silly proposals. Another game that is
a challange to GM, since there are so many extra rules that you
must keep track of. One game has been runto completition in the
UKbysnailmail and severalothers have started.

Danielisunfortunate, in living in France.  I don't think thatthere
is a French Variant Bank, although Jeff Bryant runs a variant bank
in Belgium which stocks games.  I don't have time to type up these
rules onto e-mail format.

One game which I should perhapes mentionis Lew Pulsipher's
SONG OF THE NIGHT (about which there is some debate as to
the playability of the game). In this the players know the map,
but in everyprovince they enter there is a `random' event (perhapes
I should add thatthis random event only occurs when the province
is visited for the first time). Being a fantasy -based game the
random events are normally meeting various kinds of monsters
which either lly with you or attack you (possible preventing you
enering the province). Players have Hero pieces and wizard pieces
(which can cast spells) and there is magic and treasure to be found!
It is a pain to GM (I playedinone game and GMd two others). But
perhapes Daniel wouldbe interested in this game given thathe is
also a role--players?

I think that it would be very difficult to progamme a computer
todiplome, but it would be very interesting to try to program
a computer to play a good TACTICAL game. And a game ran between
seven computers playing purely on tactics would make an interesting
specticule.  I'm interestedin seeing hos this devolpes. AN easier
task (possible) would be to cocentrate on programming justfor
the two-`player' game.

One thing not mentioned by Daniel, how far would programmers be able
to give `book' openings?

Do you intend runnng the dip aricles I mailed you, as opposed
to the letter column/editorial I mailed you? Mentioning the
letter column, you have yet to answer my question to you hidden
away in it!  ((Let me know what it is, and I'll answer it.))

Thanks for mailing me Eric's address, I've been able to book my
copy of the full RS POLL results which I am looking forward
to with baited breath.

It's a pittythat you were not appointed a bNC for e-mailgames.
I would ring/write to Don Williams if I were you to try and get
this position, the present situation is clearly unacceptable.
I think that you are being too harsh on the BNC (who I know has had
`rel-life' problems). As for MNs, whydon't youget these of BRAD
WILSON who I have always foubnd to be prompt.

The prospect of setting up a branch of the DAAS on e--mail is
one that I had not considered, but on that (memory permitting) I
hope to putinto action. I will send you a copy of my classification
system, if you wait until September I will send you a copy of the
next catalogue so you cana look at all the articles that I have
already catalogued.

You mention thatthere is a list of which dippy articles
are in which issues.  How can I get hold of this list?

About the Runestone Poll, if youprint anyof my comments then
pleaseprint my name.

Anything else, to add?  Don't think so.

MARK

PS One of the reasons that EP may be so `impersonal' is that
is is very difficult to keep loccing issues, they come out at
such a fast and furious pace that it is very difficult to keep
levelwith them.  Hence people might not bother loccing at all, or
only very occassionally.

More from loeb@rita.laas.fr/Daniel Loeb:

1) Concerning the remarks by Jack Bennetto: I've changed the protocol
somewhat. A new copy can be found below. I don't think it would be a
good idea to test the protocol among humans too much. Two people would
go crazy try to write using these 3 letter prefixes.
EXP can include a turn number now, but not a message number, since we
might not be able to guarantee the order of delivery. The WIN and LUZ
commands are included in the think command, since THK allows the
return of commented moves.

The multiply addressed messages is an issue we will have to discuss. I
am under the impression that it will complicate the interface
unnecessarily.

We should try to indicate which messages should be acceptable to all
machines, and what substitute messages are available in other cases.

2) Ken Lowe says his Judge program is written in C, and can be FTPed
from milton.u.washington.edu as public/misc/dip.tar.Z

3) Eric Klien says he will allow working programs in some of his
Electronic Protocol games. However, by default, all Electronic
Protocol games are open to humans only.

Now here is my revised protocol:

MESSAGE PROTOCOL:

Notes: All message meanings are for advisory purposes only.
Programmers may devise any message they like, and may interpret
messages as they see fit. However, the message "TRY" should be most
useful in communication between program with differing vocabularies.

All messages should be surrounded in parenthesis and begin with a word
of exactly three letters. Parenthesis should be included as indicated;
however, messages should be readible without the parenthesis.

Brackets denote optional arguments, and braces indicate repeatable
arguments. Brackets and braces normally do not appear in messages. A
choice is indicated by a slash. Unnecessary spaces or characters
should not appear in a message. Lower case characters should not be
used.

Lower case letters are used below to describe arguments as indicated
below:

annmoves = 	(unit move (note))
int = 		integer
move =		HLD / MTO prov / SUP unit {MTO prov} /
		CNV unit {MTO prov} / RTO prov / BLD / REM / DSB
msg =		message (or the key word of a message)
nation =	ENG / FRA / GER / ITA / AUS / RUS / TUR / REF
note = 		FLD / BNC / CUT / (RET {prov})/ NMR / NSO / NSU / DSR
		/ VLD / HUH / FAR
prov = 		name-of-province or ( STP/SPA/BUL NTH/STH/WST/EST )
results = 	{(nation {annmove})}
rule = 		rule number according to Avalon Hill rulebook
syntax =	DPP or another standard
turn =		(year - 1901)*5 + 1or2or3or4or5 depending on movement phase
unit =		( A/F prov )

1) Making contact with the other countries.

Message Syntax: (TRY {msg} )
	Meaning: I do not understand the message you sent, following
is a list of messages I can undestand.
	Example: (TRY DWR PCE UNO UDO),
	Example: (TRY) used by a program which can't communicate
	Reply with: Equivalent command from message list.

Message Syntax: (PSV {msg} )
	Meaning: Following is a list of messages I can understand
although I don't use them.

Message Syntax: (ACT {msg} )
	Meaning: Following is a list of messages I use but don't
understand.

Message Syntax: (HUH (msg))
	Meaning: I understood the general command sent, but not the
specific syntax. Here is the misunderstood message with an asterisk
inserted at the point up to which I understand. (The code (ERR) can be
inserted instead of an asterisk.)
	Example: (HUH (DMZ BUR *ENG RUH)) indicated the
nonacceptance of a nation in a list of provinces in the demilitarized
zone command.
	Reply With: A better version of the same command.

Message Syntax: (ALT {syntax})
	Meaning: Please use (one of) the alternate syntax/protocol
indicated. (This is for use if people create alternate conflicting
protocols.)
	Example: (ALT DPP)
	Reply With: Given more than one choice of protocol reply
positively with ALT or the protocol. Given one choice reply positively
with YES or the protocol. Respond negatively with NOP or TRY or ACT or
PSV.

Message Syntax: (YES [(msg)])
	Meaning: I agree to (or have received) your message (as follows).

Message Syntax: (NOP [(msg)])
	Meaning: Contrary to YES.

Message Syntax: (RPT)
	Meaning: Please repeat or replace message.
	Reply With: Message, or NOP is you are unable to do so.

Message Syntax: (BWX)
	Meaning: No comment. None of your bee's wax.

Message Syntax: (END)
	Meaning: Offer to disconnect. Ready to submit orders assuming
the other replies with a YES.

2) Communicating with the GM.

Message Syntax: (CDD)
	Meaning: The country in question has achieved a fatal error
and is in civil disorder. WARNING: Only the GM is allowed to write
programs which generate this message. These programs will
automatically replace programs which crash.

Message Syntax: (TME int)
	Meaning: You must submit moves soon.
	Example: (TME 3) : you may send 3 more message
	Example: (TME 0) : no more messages may be sent. Send moves
immediately.
	Reply With: Your orders.

Message Syntax: (SUB {moves})
	Meaning: I submit the following orders. Order can be
resubmitted for change if necessary.
	Replies from GM: YES if moves are now complete followed by TME
if this is true for all countries. THK if moves are lacking or contain
errors.

Message Syntax: (ORD turn results)
	Meaning: These are the results of the previous turn. If the
coming move is a movement phase, then begin diplomacizing. Otherwise,
submit moves soon, since (TME 0) will be your only warning. NOTE: Only
the GM should generate this message.

3) Basic terminology

Message Syntax: (DWR)
	Meaning: I declare war on you.

Message Syntax: (PCE)
	Meaning: I propose a general cease fire between our forces.
	Reply With: Negatively NOP or DWR. Positively with YEP or DMZ.
Possibly to be followed by ALY.

Message Syntax: (DMZ [{prov}])
	Meaning: Let's declare a demilitarized zone (as follows):
	Example: (DMZ BUR RUH NTH)
	Reply With: YEP, NOP, or DMZ.

Message Syntax: (ALY [WTH {nation}] [VSS {nation}])
	Meaning:I propose an alliance between me and you (and the
indicated country(ies)) (versus the indicated country(ies))
	Example: (ALY WTH FRA VSS GER RUS)

Message Syntax: DST {(nation [{prov}]/int )}
	Meaning: We will divide the spoils as follows. (Avoid listing
nations who won't be getting spoils.) Temporary distributions of
spoils to eliminate inequities can be assigned using FOR command.
	Example: (DST (ENG BRE SPA POR) (GER BEL PAR) (ITA MAR))
	Example: (DST (ENG 3) (GER 2) (ITA 1))
	Reply With: DST with provinces in order to reply positively to
DST with integers.

Message Syntax: ( UNO/UDO/INO/IDO {(order)})
	Message Syntax: ( XNO/XDO nation {(order)})
	Meaning: I/you/he will/won't/should/shouldn't make the
following order(s). (Specify a time with FOR. Otherwise, default is
the earliest time such an order might be appropriate.)
	Example: (UNO ((FLT (STP NTH)) BLD))
	Example: (UDO ((AMY WAR) MTO PRU) ((AMY SEV) MTO ARM)
	Example: (INO ((AMY MAR) SUP (AMY VEN) MTO PIE))
	Example: (IDO ((FLT ECH) MTO MAO))
	Example: (XDO ENG ((FLT ECH) CNV (AMY EDI) MTO WAL))
	Example: (XNO FRA ((AMY PAR) HLD))
	Reply With: YES or SND.

Message Syntax: EXP [turn] [{(msg)}]
	Meaning: You broke you promise(s) (indicated) (before the turn
indicated.)
	Example: (EXP 12 (UNO ((FLT (STP NTH)) BLD)))
	Reply with: SRY, NOP=DNY, DWR, or YEP (to indicate accusation
is true without being sorry or declaring war).

Message Syntax: (SRY)
	Meaning: I am sorry.

Message Syntax: (DNY)
	Meaning: I deny your accusation.

4) Complex messages

Message Syntax: (FOR turn {(msg)})
	Message Syntax: (FOR [turn]-[turn] {(msg)})
	Meaning: The following messages refer to the turn(s)
indicated.
	Example: (FOR 15 (UDO ((FLT SEV) BLD)) : good for Winter 1903
	Example: (FOR 15- (UDO ((FLT SEV) BLD)) : good starting Winter 1903
	Example: (FOR -15 (UDO ((FLT SEV) BLD)) : good until Winter 1903
	Example: (FOR 15-20 (UDO ((FLT SEV) BLD))) : good Winter 1903 -
Winter 1904.

Message Syntax: (IFF (m1) THN (m2) [ELS (m3)])
	Meaning: If you m1 [either UNO or UDO probably] then I will m2
or else I will m3.
	Example: (IFF (FOR 15 (UNO ((FLT (STP NTH)) BLD))) THN (DWR)
ELS (FOR 16 (IDO ((AMY SWE) SUP (FLT (STP NTH)) MTO NWY)))) : If you
don't build F Stp(nc) in winter 1903 then I will declare war,
otherwise I will support you in Spring 1904 to Norway.

Message Syntax: (WHT [{(unit)}])
	Meaning: Suggest a(nother) move for the indicated unit (or all
my units). In the case of his units, then this is a demand to know
what these units will be ordered to do.
	Example: (WHT (FLT NTH))
	Reply with: Negatively with BWX. Positively with IDO or UDO,
and use WHT for the remaining units. (Try to make forward progress
with these suggestions in order to avoid infinite loops.)

Message Syntax: (HOW unit/prov )
	Meaning: Suggest a plan of attack
	Example: (HOW MOS)
	Reply with: IDO or UNO, or NOP if you don't think it is a good
idea and THK, WIN, LUZ to give the reason.

Message Syntax: (THK results)
	Meaning: I predict the following results. (Also used by the GM
to correct errors.)

Message Syntax: (CTE rule)
	Meaning: I cite rule (by rulenumber) which you may have
overlooked.

Message Syntax: (FRM nation [nation2] {(msg)})
	Meaning: The following message(s) were sent to me (or Nation2)
from the indicated nation.
	Example: (FRM RUS (WAR))
	Example: (FRM RUS ENGLAND (WAR))

Message Syntax: (SND nation {(msg)})
	Meaning: Send the following messages to the indicated nation
	Example: (SND RUSSIA (UNO ((FLT (STP NTH)) BLD)))

Message Syntax: (SNT nation {(msg)})
	Meaning: I have sent following messages to the indicated
nation. (Equivalent constructions are available using FRM, but may be
less believable.)
	Example: (SNT RUSSIA (UNO ((FLT (STP NTH)) BLD)))

Message Syntax: (DFM [int])
	Meaning: Discard the next (so many) message(s). (This command
is only useful if ports are available for speaking to several nations
at a time.)
	Example: (DFM 2)

Message Syntax: (PPT [{ IDO/SND/FRM/DST }] )
	Meaning: I offer to become your pupppet (and I will: make the
move you tell me to, send the messages you tell me to, forward all
messages to you, and accept any distribution of supply centers that
you think of. Any conditions on the Puppetship should be places in a
FRM or IFF.
	Example: (IFF (DST (AUSTRIA VIE BUD TRI)) THN (FOR -80 (PPT
IDO SND FRM))) : If you let me (Austria) keep my home supply centers
then I will do whatever you say until the end of 1908
	Example: (PPT)

Alphabetical Index
ACT	message		active vocabulary
ADR	province	adriatic sea
AEG	province	aegean sea
ALB	province	albania
ALT	message		alternate protocol
ALY	message		ally (see WTH VSS)
AMY	type		army
ANK	province	ankara
ARM	province	armenia
AUS	nation		austria-hungary
BAL	province	baltic sea
BAR	province	barents sea
BEL	province	belgium
BER	province	berlin
BLA	province	black sea
BLD	order		build
BNC	note		move bounced (opposed by an EQUAL force)
BOH	province	bohemia
BRE	province	brest
BUD	province	budapest
BUL	province	bulgaria
BUR	province	burgundy
BWX	message		bee's wax, none of yours = no comment
CDD	GM message	civil disordered nation
CLY	province	clyde
CON	province	constantinople
CNV	order		convoy
CTE	message		cite rule
CUT	note		support cut
DEN	province	denmark
DFM	message		discard following message
DMZ	message		demilitirized zone
DNY	message		deny
DPP	syntax		diplomacy programming project (See ALT)
DSB	order		disband (instead of retreating)
DSR	note		move failed since convoy disrupted
DST	note		distribution of supply centers
DWR	message		declare war
EAS	province	eastern mediterranean
ECH	province	english channel
EDI	province	edinburgh
ELS	argument	else (see IFF)
END	message		end = disconnect
ENG	nation		england
ERR	message		error
EST	coast		east
EXP	message		explain
FAR	note		too far = not adjacent (along the coast)
FIN	province	finland
FLD	note		move failed (defeated by a GREATER force)
FLT	type		fleet
FOR	message		for these turns
FRA	nation		france
FRM	message		received from
GAL	province	galicia
GAS	province	gascony
GER	nation		germany
GOB	province	gulf of bothnia
GOL	province	gulf of lyon
GRE	province	greece
HLD	order		hold
HOL	province	holland
HOW	message		how should we attack
HUH	message		huh - error
IDO	message		i will do
IFF	message		if (see THN ELS)
INO	message		i will not do
ION	province	ionian sea
IRI	province	irish sea
ITA	nation		italy
KIE	province	kiel
LVP	province	liverpool
LVN	province	livonia
MAO	province	mid-atlantic ocean
MAR	province	marseilles
MOS	province	moscow
MTO	message		move to
MUN	province	munich
NAF	province	north africa
NAO	province	north-atlantic ocean
NAP	province	naples
NMR	note		no move received
NOP	message		nope = no
NSO	note		no such order
NSU	note		no such unit
NTH	coast/province	EITHER north sea OR north coast
NWG	province	norwegian sea
NWY	province	norway
ORD	GM message	orders
PAR	province	paris
PCE	message		peace
PIC	province	picardy
PIE	province	piedmont
PSV	message		passive vocabulary
PPT	message		puppet offer
POR	province	portugal
PRU	province	prussia
REF	nation		referee = gamemaster
REM	message		remove (due to lack of supply centers)
RET	note		must retreat
ROM	province	romania
RPT	message		repeat message
RTO	order		retreat to
RUH	province	ruhr
RUM	province	rumania
RUS	nation		russia
SER	province	serbia
SEV	province	sevastapol
SIL	province	silesia
SKA	province	skagerrak
SMY	province	smyrna
SND	message		send to
SNT	message		i sent
SPA	province	spain
SRY	message		sorry
STH	coast		south coast
STP	province	st petersburg
SUB	GM message	submit moves
SUP	order		support
SWE	province	sweden
SWI	province	switzerland (an illegal province)
SYR	province	syria
THK	message		think that
THN	argument	then (see IFF)
TME	GM message	time left
TUN	province	tunis
TUR	nation		turkey
TUS	province	tuscany
TRI	province	trieste
TRY	message		try these commands
TYL	province	tyrolia
TYS	province	tyrrhenian sea
UDO	message		you should do
UKR	message		ukraine
UNO	message		you should not do
VLD	note		valid order
VEN	province	venice
VIE	province	vienna
VSS	argument	versus (see ALY)
WAL	province	wales
WAR	province	warsaw
WES	province	western mediterrnean
WHT	message		what to do with
WST	coast		west coast
WTH	argument	with (see ALY)
XDO	message		he should do
XNO	message		he should not do
YES	message		yes
YOR	province	yorkshire

More from m.i.nelson@gdr.bath.ac.uk/Mark Nelson:

Dear Eric,
          I have just finished reading EP 185.  Perhapes you should
rename you zine CYPRPUNK WEELY-a fanzine just devoted to publishing
new cyperpunk fiction?  COuld Gibson have thought up anything more
terrifying and outrageous than this?

However, the most important stuff inthis issue is the formation
of the CLB. If you want good rights, you don't wait for the
legislatros to draft them...you goout and fight for them on the
asis of excisting laws!   I hope that we will be kept informed of
whatever progress the CLB make, what would be interesting is to
read the abuse given the organization by `the more institutionalized
precincts of the computer industry'.

MARK

More from loeb@rita.laas.fr/Daniel Loeb:

Subject: A Computer Diplomacy Strategy

(Eric, please consider this as an article for publication, and remind
people about the Diplomacy Programming Project).

In 1983, I wrote a program for the Apple ][ to adjudicate diplomacy.
In addition, the program would play the remaining countries following
a choice of three strategies:

(1) Civil Disorder. All units hold. Units disband when forced to
retreat. Builds are waived. Removals are accomplished using a
FIRST-IN/FIRST-OUT principle (as opposed to the shortest route rule
indicated in the rules).

(2) Random. All units move to an adjacent province at random. Legal
retreats are randomly taken when possible. Removals are chosen at
random. Builds are in a random open supply center, and if applicable
the choice of fleet/army and of coast is at random.

(3) Somewhat serious strategy. This strategy was the best I could
write in Basic. [Possible simple extensions are indicated in
brackets.]

Drawbacks: no communication with allies, no convoys, no long term
strategy, no stabs, possible cuts not taken into account, no draws, no
concessions

Allies: The program would sort the other countries according to
their "trustworthiness". The initial sorting is at random. The program
considers copies of itself (and strategies 1 and 2) as being more
trustworthy than humans. Other than that the choice is random.
[EXTENSION: the program uses previous games' data to construct list.]
During the game any country which attempts to move into a supply
center I own is moved to the bottom of the list. [EXTENSION: Allies
can attempt to recover their home supply centers if I own them.]

The first two countries in the list are considered allies. The bottom
two are considered enemies and the middle two are considered neutrals.

1901: As a direct function of the trustworthiness list, the program
choses among a set of BOOK openings in Spring 1901. These are what I
consider the best moves except that the English openings leading to
convoys are avoided since the program doesn't know how to convoy.
[EXTENSION: Use book openings in Fall 1901 as well.]

Afterwards: In the following turns, the analysis preceeds in several
stages, namely:
	a - Analysis of vulnerability of each space on board
	b - Valuation of each space on board
	c - Targetting of units
	d - Assignment of moves
	e - Reassignment of holds

(a) For each space, we count the number of own forces, allied forces,
neutral forces, and enemy forces which could move this turn to that
space.
[EXTENSION: Since a unit is unlikely to move to a space when a
unit can move to somewhere much more important, we should recalculate
(a) after (b) taking such units into count as an appropriate fraction
of a unit. (a) and (b) could then be repeated until the results
converge.]
[EXTENSION: An ally's claim that he will move or support
somewhere should be taken into account in these figures---as is to a
lesser extent any rumor about or statements by the enemy and neutrals.]

(b) Each space is given a value of its importance in function of:
	1. the number of each type of unit which can move there this turn,
	2. the kind of space (neutral supply center, my home supply
	center, province, enemy home supply center, etc.),
 	3. its owner, and
	4. a small random factor in order to help the program not be
	"outguessed" by players well aware of the computer's strategy.
Plus, a certain small fraction of the value of the neighboring spaces
is added to the value of each space in order to encourage movement in
"the right direction" and towards key spaces such as Tyrolia and the
North Sea with many important neighbors.
[EXTENSION: The importance of a space as claimed by an ally should be
taken into account somewhat here.]
[EXTENSION: the inland neighbors, bicoastal neighbors, and sea
neighbors should be kept track of separately in order to help decide
if to send in an army or a fleet and if so then on what coast.]

(c) The units are assigned one at a time as their "target" the space
nearest them with the highest importance. The number of units assigned
to each target is kept track of, and the importance of the space is
lowered when the necessary number of units are targetted there. (For
example, it is only necessary to send 1 unit to an undefended space,
but it is necessary to target 3 units to defend a space against a
potential force of 3 units.)
[EXTENSION: Moves left over to be made by ally should be communicated
to him, and moves the ally says he will make should be adjust the
necessary number of units.]

(d) Then for each space targetted, we check if there is a unit in the
space targetted for the space, if so then the others support him, and
he is provisionally ordered to hold.

If not, then we check if we already have a unit in the space targetted
for elsewhere, if so then we attempt to deal with that other space
first. Such a deferment can not result in an infinite loop.

Otherwise, we continue by chosing one of the targetted units to move
into the space while the others support. The moving unit will be the
one coming from the least important space.
[EXTENSION: The moving unit will be the one whose support would
otherwise be cut.]

[EXTENSION: Targettings involving allied units should involve
discussion with ally.]

(e) For units provisionally assigned to hold, we take a last look to
see if there is anything they can do. That is, check if there is a
neighboring targetted space which can be supported.
[EXTENSION: Perform this cocurrently with (d).]
[EXTENSION: If this fails then ask ally for suggestions)
[EXTENSION: Also consider all possible convoys through the space.]

-----------

The program never disbands if not necessary. Instead it retreats
according to the previously established importances. Removals and
builds are also accomplished according to this table.
[EXTENSION: This table could be recomputed according to recent
developments, and in order to avoid having retreats bounce.]

-----------

My feeling is that this outline should provide a good core to any
diplomacy program. What do you all think? Please send in your
suggestions.

Yours, Daniel Loeb

More from m.i.nelson@gdr.bath.ac.uk/Mark Nelson:

Dear Eric,
          I am enclosing a short article I have written on the
Runestone Poll Results which will be published in several UK
zines.  If you think there is anything that I have missed out then
let me know and I will add it to the final copy. (I have not deleted
formatting controls because you have all ready covered the RP in EP, so
there wwould be no point in you running the article.).

Yours mark.

\begin{center} {\Huge {\bf CUTTIN' OUT 7: AUGUST 1990}} \end{center}

\begin{center} {\LARGE {\bf The 1990 Runestone Poll}} \end{center}

\underline{Cuttin'} \underline{out} is a roving subzine which appears
at irregular interviews.  It is produced by {\em Mark Nelson},
112 Huntley Avenue, Spondon, DERBY. DE2 7DU. This issue may be
reprinted in any zine, just so long as I get a copy of the reprint
issue!

The Runestone Poll is the American equivilant of our own Zine Poll,
and was run this year by Eric Brosius.  As well as covering zines,
there are seperate polls for subzines and GMs.

The aim of this article is to review the top American zines, and
to encourage readers to take the plunge to see an American zine.
The following reviews emphasis that there are now a number of
well writte and well-produced zines. In fact there are
probably more such zines in the states than over here!
If you are interested in subbing to one of the zines reviewed below,
then you can send your subscription to: Doug Rowling, 228 Kinnell
Avenue, Cardonald, GLASGOW. G52 3 RU.

You should also order the \underline{Zine} \underline{Register},
and order a selection of zines from the North American Zine Bank
(\$5).

The top ten zines were:

1. NORTHERN FLAME ({\em Cal White})
2. Upstart        ({\em Gene Schenck})
3. The Zine Register ({\em Tom Nash})
4. Perelandra     ({\em Pete Gaughan})
5. Penguin Dip    ({\em Steve Dorneman})
6. Kathy's Korner ({\em Kathy Caruso})
7. Been There, Done That ({\em Tom Nash})
8. House Of Lords ( {\em Dick \& Julie Martin}) -- {\sf folded} --
9. The Boob Report ({\em Jim Burges})
10. ark ({\em Eric Brosius})

{\bf NORTHERN FLAME} Cal is an `old--timer' who was active in the
North American Hobby in the late 1970's before dropping out.  He
returned in 1987 after a lengthy `retirement'.  Ten years older, and
ten years wiser Cal runs a very relaxed zine and takes a poor view
of people who take the hobby too seriously.  The center piece of the
zines is a very long and friendly letter--column.  With six--weekly
deadlines, it is very easy to both contribute to the zine and play
in it; although with a cost of /0.15p per side + postage it is a
little expensive. A refreshing change from the multitude of  zines
(on both sides of the Atlantic) which take things too seriously.

I have only seen a couple of issues of {\bf Upstart} which is now in
its third year of production.  A more or less standard diplomacy zine
with an interest in naval trivia. \$7.50 for ten issues.

{\bf The Zine Register} is the American equivilant of
\underline{Mission} \underline{From} \underline{God}, but being
laser--printed on nice white paper it's looks much better.  Reviews
of all the North American zines (often by two or three people) and
many non-North American zines. If you're interested in seeing a
North American zine you should certainly make this your first port
of call.  Cost \$ 1.50.

{\bf Perelandra} has been described by no less an authority as
Iain Bowen as being {\em ``A US \underline{Denver Glont} without the
mania or the word `fuck'\,''.  A zine that always does well in
American Polls and has a focus on literature. (Yup, another one that
I don't see.) \$ 1.50 an issue.

This article is trying to encourage people to take a look at
North American zines, to break down barriers. {\bf Penguin Dip} is an
 example of barrier breaking between fandoms. It features the normal
diplomacy stuff (games and articles) written by diplomacy fans and the
`usual' (letters, articles, reviews) written by sf fans.
 It's a melting pot between the two fandoms, is extremely well
written and quite excellently produced. At \$20 for a years sub, this
is excellent value for money.

{\bf Kathy's Korner} is a very induvidualistic zine, and very difficult
to review.  There is no--other zine like it in the world!
In spirit, NERTZ comes nearest to it although the two zines are
poles apart.  A reader participation zine, in which the casual
reader will sink without trace.
 Not a zine for the newcomer to the North American
Hobby to see, but once you've found your feet perhapes worth a
look.  Kathy is one of only a handful of people who have won three
games as Italy in postal dip.  \$1 an issue for one of the
longest running North American zines (into the 160's).

{\bf Been There, Done That} is a fairly new zine that Iain Bowen
describes as ``One of the most British of the US zines''.  The
zine has a large letter--column discussing a variety of subjects,
and Tom is one of the growing number of Americans who has a
passion for Railway Rivals.  Perhapes the best place for Brits to
start their exploration of the North American Hobby. At \$ 1 an issue,
certainly a cheap entrance place into the North American Hobby.

{\bf House of Lords} has folded, but it was the top zine for
hobby discussion.

{\bf The Boob Report}  is free to players, traders and regular
contributors and has a letter column that covers all kinds of
music.  However, at three-weekly probably too efficient for
British contributions.

{\bf ark} is devoted to running Empire Builder.  Nuff said.

\begin{center} {\sf Some Other Noticable Zines} \end{center}

13 {\bf Carolina Command \& Commentary} (Michael Lowrey) is produced
by those nice people at Chapel Hill (site of World Dip Con II). A
zine for people interested in American politics and American
foreign policy $\backslash$
international relations.  However, the mainstay of
the zine are all far--right Republicans. \$1 an issue.

14 {\bf Excelsior} (Bruce McIntyre) I don't
know way, but there is a
definite different feeling to Canadian zines.  There are all
friendly and easy going, more fannish (?), than
their American counterparts.  This is another well
written zine (on a par with anything produced over here), but is
produced on a very `relaxed' schedule -- 8--12 week intervals
beetween issues being common.  An unashamed perzine.  Certainly a
zine worth seeing, but possible too slow or those who want to
become active in the North American hobby.

15 {\bf benzene} (mark lew) is not really a diplomacy zine, it's a
perzine with a slant on economics/politics which features
serious letters  by diplomacy fans.  mark is very articulate, as
are many of the contributors.  no overseas price quoted, but
expensive. mark is currently engaged in a feud against capital
letters.

20 {\bf Graustark} (John Boardman)  You wondered why Boardman numbers
were called Boardman numbers?  Well they were named after the first
custodian--John  Boardman who amazingly enough is still around.  But
what is even more amazing is that this is the original diplomacy zine,
which ran the first ever game of diplomacy in the Summer of 1963
and is slowly edging its way towards Issue 600.  \$10 for 12 issues.

More from loeb@rita.laas.fr/Daniel Loeb:

1) Ken Lowe gives more information about getting the source code for
Judge. You can send the command
	get uudip.tar
to judge (judge@milton.u.washington.edu) for the source without
conference maps. (160,826 bytes).

2) Richard Kirby gives more detail about how UNIX sockets work:
Sorry I havent got back to you sooner. Anyway thanks for the messages. As to
your point about parallelism, well UNIX provides multi-processing which is
psuedo-parallel. Back to the idea of a PhD, what I should have said is that I
have read several papers on diplomacy, in the game, and in real life. Also,
with six, seven or more other people doing similiar work it would be hard to
prove that it was my own work. Anyway, I will have a word with my boss about it
and see what he says.

Before I can send some code to you (actually I can do a C program that should
be reasonably portable) showing how sockets would work, I really need to know
if you have a Sun, and what Common Lisps you have, since there is no standard
for inter-process communications.

Basically, how it will work is that each player would be a seperate UNIX
program, and there would be one more which would be the GM. Each program would
have a socket for communicating through. When a player first starts, it
connects with the GM, and gets given a country, and then creates its own socket
with the countries name. Whenever the player feels like talking to another
country, it just connects to the socket with the country's name. (UNIX sockets
are just files which act as rondeavous points). I have the socket code for KCL
(a freebie Common Lisp), and I think that some Common Lisp's provide this stuff
anyway. The point is though, that using the seperate UNIX processes idea, we
can all use exactly what we want as the language, as long as there is the
facility to talk on sockets.

When do you expect to start this? From what I have seen, I think that some
people might not realise how difficult it will be to build a decent diplomat,
but it will certainly be very exciting to try :-)

3) Here are two minor extensions I should have mentioned in my letter
on computer strategies:

First, in addition to being lowered whenever they attack you....
[EXTENSION: programs are moved to the bottom of the list if caught
lying.]

Second, one of the factors used in calculating the importance of a
space is ...
[extension: 5. whether it is now a fall or a spring turn.]
During Fall turns the emphasis will be on supply centers, and in
spring turns the emphasis will be on strategic spaces.

4)
******************************************************************************
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.
******************************************************************************