The following discussion comes in response to various queries on the Internet, including one administered by The Analytical Engine, the results of which appeared in their issue of Volume 1, Number 3 in January 1994.
Each ship could be rotated clockwise or counterclockwise, fire reaction engines that eventually ran out of fuel, and fire missiles of finite range and finite number. The ship obeyed Newton's laws, accelerating and decelerating under the influence of its engines and of solar gravitation....
But one thing it didn't do was do gravity on the torpedoes, and so there were a LOT of sexy techniques that took advantage of that anomaly. Also, Scott didn't mention the wonderful Starfield. Peter Samson had added it to the original SPACEWAR.
It was a real star field, generated by some incredibly clever code so that it had real constellations in their real positions and they slowly drifted across the background.
Collision of two ships produced a vivid, graphically depicted explosion on screen, and both players were out, whereupon the game restarted.
One should probably note that for folks weaned on 'Star Wars' and Super-VGA PC games, it wasn't really 'vivid'. Kind of a little star-burst. Very nice,and crystal-clear what had happened. But nonetheless fairly simple.
The display was a point plotting, no memory scope. ALL displays were 'animated' since there was no display memory: if you wanted something to persist on the screen, you had to be in a loop constantly redisplaying it. The scope had a neat, special phosphor which displayed green when fired, but then faded for short while in yellow. This made the display flicker a lot less, and it also meant that things left 'trails' as they moved around. It made for some quite wonderful 3-D effects.
-- from Bernie Cosell, via Internet
The SPACEWAR to which I refer was run on a PDP-1/B at BBN in Cambridge. ((one of the) first Timesharing machines -- but that's a different thread)
I believe this was the identical SPACEWAR to that run on the MIT PDP-1; certainly many of the same MIT programmers worked at BBN too.
SPACEWAR at BBN was probably the first "object oriented" program (nothing new under the sun), the program had translate, explode and rotate,etc. generic object functions that would operate on whatever object you fed to them (mostly the two ships).
The DECScope had 64(?) intensities (6 bits?). Peter Sampson programmed a rather remarkable Starfield over Cambridge, that rotated a 24 hour day (in about an hour of play time) as the background. (when last heard from, Peter was at System Concepts who were making -10 -20 clones)
Standard (no bit diddling) SPACEWAR was 2 ships, sun with gravity for ships not torps, 8(?) torps with life about 3/4 screen (about 2-3 inches/second); hyperspace survival probability started at about 75% and decreased per use to about 20% (?), everything wrapped, torps could be shot down by other torps. Torp speed was additive to shooter's speed (by some function) but you could shoot yourself down if you fired ahead whilst going too fast,or you could "leave mines" by firing opposite the way you were going. All this in 4k of 18-bit PDP-1 words (including Peter's Starfield database)
Control was accomplished using testword switches (I/O instruction readable) the 4 on the left for one ship, the right 4 for the other. (order may be wrong, but...) rotate left rotate right torps thrust (Actions a whole lot like Netrek) left ^ right => hyperspace.
Program had patchable locations that controlled number of torps at a time and their spacing, life, and speed. Also, rotation could be "by thruster rocket" or "by gyros" -- Gyro would rotate while the switch was thrown and stop when off, whereas Rocket rotation would start and increase angular momentum while the switch was still thrown. To stop, you had to try to thrust the other way for exactly the same time.
The sun had switchable gravity and I think you may have gotten a choice whether torps were affected by gravity. SPACEWAR ships and torps would wrap, although that may have been another option. For a challenge, there was billiard SPACEWAR; single shot on the screen at a time, no kill counted unless it had wrapped at least twice.
PDP-1 SPACEWAR was the source of the first computer-induced medical problem (well before Carpal Tunnel Syndrome), "Spacewar Elbow." Occasionally Cosell (Bernie) and I would spend a night playing SPACEWAR, only to find that leaning on the elbows for 7 or 8 hours straight would leave us unable to straighten our arms for quite a while.
BBN outlawed SPACEWAR occasionally, mainly because switches died. The life of test word switches was shortened by this game. (it took me 5 hours to replace the first switch that needed fixing; I got it down to 22 minutes by the time the machine left.)
Someone posted that the PDP-1 at the computer Museum was one of the MIT machines, I believe that too. (It has joystick control boxes)
-- from Paul M. Wexelblat, via Internet
I ran into SPACEWAR when I was a freshman at MIT in 1962/63, and added a couple of lines of code to it. (I have no idea what those lines were, nor if they were included in any versions after I left. I do have the source listings (promised to the Computer Museum) but I hadn't yet learned the concept of footprints to mark changed code.)
When I first saw it the user interface to SPACEWAR was a pair of wooden boxes about the size of a small file card box, each of which had two telephone key switches (turn left/right, rocket on, go to hyperspace) and a button (fire torpedo). When we got the second PDP-1 someone went over to Eli Heffron and Sons (motto: "We have Surplus Surplus") and bought a pair of Air Force drone controllers to serve as input devices.
And the PDP-1 we ran it on was really an amazing box: Memory of FOUR K! (well, actually that's 4K words of 18 bits each), and a blazing memory cycle of 5 uS.
The game was so popular that it "signed" the console log itself: You might see entries for an hour or two for the staff programmers, a block here and there for a student, an occasional Big Name (Marvin Minsky, for example), and huge blocks of time merely noted as "Spacewar."
When we got DEC's first drum (Wow! 32 tracks, each of which held exactly one coreload of 4K words!) one of the tracks was instantly dedicated to SPACEWAR, and the console load tape for it shrunk to a bootstrap a couple of feet long.
Spacewar wasn't a video game! Video encoding of data was not used to paint the image on the face of the CRT. Instead, the software drove a pair of ADC's to handle deflection in the X and Y directions, and the software painted each dot on the screen. The resolution was far better than standard video, but the number of dots that could be painted without causing flicker was limited.
(By the way, I have played Spacewar myself on a DDP 224 computer at the University of Michigan -- that machine was a nice 24-bit minicomputer, and their version of Spacewar used two joysticks on the graphics display.)
-- from Doug Jones, via Internet