DotWar
a.k.a. Myrmidon

Ward Cunningham
January 14, 1998

In the mid '70s I distributed this manual for DotWar, a multi-user real-time computer game intended to replace the then extremely popular and recently outlawed game called Star Trek.
 
  
Environment
  
   This prototype simulation takes place in a three
   dimensional, near infinite space.  Each user in the
   system is a zero volume object that is free to propel
   itself about.  Control is via an acceleration vector
   which may be changed to any value (within limits) at
   any time.  The rest is under control of newtonian
   mechanics.  An object named 'earth' has been provided
   as a reference point.  'Earth' has no acceleration.
  
Printout
  
   The printout from each update run provides the
   following information. 1) Elapsed time since object
   creation.  2) Current acceleration vector.  3)
   Acceleration, velocity and position vectors for all
   other objects in the system relative to this object.
   4) List of all pending commands for this object.
  
Input
  
   No commands need be specified.  If specified, commands
   must be of the form:
  
      at <time> move <x accl> <y accl> <z accl>
  
   Each element is seperated by one or more blanks.  If
   <time> is less than the current time, the current time
   will be used.  If any acceleration is greater (in
   magnitude) than 1.0e6 then 1.0e6 will be used.
  
Limitations
  
   Fixed storage allocation is used.  Space is provided
   for 20 users and a total of 40 pending events.  It is
   recomended that each user restrict himself to no more
   than two or three pending commands at any given time.
   Commands delayed more than a day or two should be
   avoided.
  
Example
  
   The following job will print current status and add one      
   command.
  
      12345,abc,cm15000,t4.
      pfiles(get,myr1,id=acd) 
      myr1. 
      &eor
      at 15.5 move -500 0 2.7e5 
 
The idea was to move the game out of the computer and into the leasure-time world of the players. The computer would just keep score. I argued that players wouldn't tie up public teletypes since the timescale of the game wouldn't give online users any advantage. Gordon Letwin thought otherwise. I challenged him to a game of tag.  Here Gordon recounts the rules and claims victory in this email posted at the end of the 2-1/2 month bout. (Abr and acd are university login codes assigned to Gordon and me respectivly.)
 
  
Sent to group myr in id=acd 
mail from abr, 10:11:54  01/31/75 
tcb l174, 66 users
  
     The Battle Of The Century Has Been Decided
  
     Yes, sports fans, as many of you know, last year, in
   debating some fine point in the design of a new myrmidon,
   acd and abr challenged each other to a blood-curdling battle.
   Acd was to start in some position distant from earth, and abr
   was to start at earth. Abr was to produce a printout showing
   himself within 1.e7 miles of acd, making abr the victor. Acd
   was to produce a printout showing him within 1.e7 miles of
   earth, making him the victor.
   
     Around thanksgiving of last year, acd began moving out.
   While acd traveled out to near the orbit of pluto, abr stayed
   near earth, perfecting his battle computer, widely acclaimed
   to be one of the most intellegent, (not to mention polite)
   battle computers around. After matching with a few near-earth
   targets to 'blood' the computer, abr set out after acd just
   before christmas vacation.
   
     Throughout vacation, up until about jan 9, acd was the
   model of nonchalant sophistication. He just cruised out in
   far space, not varying course or in any way deigning to
   notice the approaching abr. Abr became so overconfident at
   what he interpreted as the signs of the acd captain being in
   a drunken stupor, that he manually overrid his battle
   computer (it knew better) to approach acd faster, so fast, in
   fact, that for a 30 hour interval, acd could have returned to
   earth before abr could slow down his rush and return.
   
     As fate would have it, acd saw his oportunity and took it.
   Jan. 11 Saw acd aproaching earth, abr also blasting towards
   earth, but with about 1/4 c of relative velocity towards acd.
   
     Things became closer, and acd and abr were expected to meet
   saturday morning. Space-war analysists all agreed that due to
   the deformed structure of space, acd could dodge abr in the
   3rd dimension without slowing in any way his return to earth.
   
     However, abr (with the aid of his battle computer) was
   alert. During acd's lunch hours, abr would reverse blast, and
   pick up a little bit greater relative velocity to acd, thus
   moving up the meeting timetable.
   
    Finally, when the computer said the signs looked right, abr
   made an outright dash for acd. Acd, not noticing, was
   accelerating towards earth, and thereby right towards abr.
   They would meet in 38 hours.
   
    Early thursday evening, abr and computer came to realize
   that acd had postponed his move to late. In the y and z
   dimensions, all that acd could do would be to postpone the
   meeting from 03:59:12 to 04:05:57. However, a 1 and 1/2 hour
   headstart in the x dimension, due to the warped structure of
   myrspace, would put acd in the clear. Abr checked out the
   situation every 45 minutes, and around 11:30, acd made his
   move. Abr caught it in just 13 minutes, followed with a
   pursuing move. Battle computer calculations showed that
   although acd could increase the passing distance to 1.2e6,
   this was not enough. Nothing he could do to prevent
   destruction at 4:01:12 that morning.
   
     Acd slept on unsuspecting, abr slept smugly. At 03:45:00
   abr's alarm rung, he arose, and fired a salvo of jobs at
   04:00:00.
   
     Acd was dead.
  
     Friends, I can tell you now that this scenario is not
   fictional: it actually happened. Acd was intercepted at
   04:01:00 thursday morning, abr is now blasting victoriously
   homeward on a pillar of fire.
   
*** End of mail  09:46:58  01/31/75 
 
My strategy had been simple. Head out for a month, turn around, and head back in hoping Gordon wouldn't know how to estimate my time of arrival. No such luck. Gordon's battle computer had me pegged.

The battle was waged once more, this time using team tactics. Harry Chesley lead the white team, Gordon led the black. White was routed.

The database was finally corrupted by who knows what mechanism. I was working on the manual for the much more elaborate version two by then so restarting version one seemed inappropriate.
 
  
Sent to group myr in id=acd 
mail from acd, 14:06:17  10/22/75 
tcb l043, 80 users
  
  The end has come.  The universe is no more.  In preperation 
  for the celibration of the second anniversary of the creation 
  of the universe I discovered that pace the myrmidon 
  data base, had been destroyed.  This is truely a
  sad moment, as all of us have vannished with it.  I have
  decided not to recreate the universe as this would be dis-
  respectfull of those that have passed away, and I am sure 
  that a second generation would be unable to raise itself
  to the level of achievement seen in the early days of myr1. 
  
  Those of you who never saw the early battle computers,
  the console's ship, faster than light travel, galactic tag
  the myrmidon forum or the epic battle between good and evil,
  do not despair for there is still talk of a new and better
  universe someday.  One with all those things that were
  lacking in the first prototype like fuel and bombs, 
  secret identities and federation spies, and wealth and
  greed and hate.  Keep your hopes up, we may all live again. 
  
*** End of mail  12:26:23  10/22/75 
 
See also:

  • Fortran Source for DotWar program
  • Unfinished Manual for Version II
  • Wiki Page about DotWar and players

© ward@c2.com