In Search of "Dojin" by Hiroshi Suzuki (1980)

John Szczepaniak emailed me with a suggestion for a possible "recoding" project (He'd noticed I do such projects) regarding a lost game from the early period of Japanese Game development.  The game "Dojin" was inspired by Manbiki Shounen.  It then helped inspire the game Nostromo. Here are some notes of my discussion of the program with my son Charlie, who watched the video transcript of Szczpaniak's interview and was able to use his knowledge of Japanese to provide some further context about the conversation going on.

From the descriptions Charlie heard in the interview and from the transcript and from the two diagrams, one by the Hiroshi Suzuki on the top left and another by a fellow programmer, Masakuni Mitsuhashi.  Both were youthful programmers and game enthusiasts who had been contracted by Taito to create games and show them to the company as a kind of game idea incubator/computing club.  Mitsuhashi had also played the game. I was able to rough together two MC-10 variations of the game screens based on their different sketches and descriptions:

Mr. Suzuki's Sketch

Mr. Masakuni Mitsuhashi's sketch (Thanks Charlie for helping create this one):

As John Szczepaniak pointed out in his email the two sketches and two descriptions suggest quite different games. The first one would involve what I felt was a fairly straightforward 8-bit BASIC text game using standard tracking algorithm for the enemies moving along the cyan colored hallways. The orange item/obstacle moves from top to bottom (or bottom to top, or both) and the player must lure the enemies via manipulation of their tracking behaviour to "cross the road" and be crushed by the orange moving obstacle (like a "car" on a road is mentioned, or a "piston" or "pendulum"). There is also an objstacle in Masakuni's version but it is not shown in the screenshot above.  In Masakuni's description the player navigates between closed rooms or cells (both descriptions mention a "forest" or "trees") by breaking the walls between them. Masakuni suggests that the enemies have a harder time breaking the walls and so move slower than the player.  But the overall goal is the same: Lure the enemies onto the "road" and time it so that they are crushed by the orange object as it moves up and down.  

According Szcepaniak, Dojin was a forerunner (inspiration of) Nostromo: 

JS: Nostromo was also inspired by Mr Suzuki's unreleased game, called Dojin?

AT: Yes, certainly. I created Nostromo on a PET CBM. Later Mitsuhashi-san ported it to PC-6001. My idea was different from Dojin - actually, Dojin was a very interesting and playable game. But the enemy was always viewable by the player. So because I loved [the film] Alien, as well as Star Wars, I wanted the enemy invisible to the player. Visible only when it is viewable by your character. So that's the idea of Nostromo.

MC-10 Nostromo based on NEC PC-6001 version

In terms of game dynamic Nostromo has a large red "niche" in which the player can shelter from the Alien (the Alien will often pace back and forth in front of it before randomly veering off down another hallway).  That niche reminds me of the little "randomly distributed" niches in Mr. Suzuki's diagram.  But Nostromo also has a dynamic of breaking the walls of the 8 "lockers" that the player must access to retrieve the randomly distributed items needed to clear screens.  So I'm in a real quandary about what the game was actually like.  I am pretty sure I can make a game similar to another Japanese game called Heiankyo Alien, which also has enemies that track you through a linear grid (the programmers actually mention this game on page on p. 400).  That game inspired the game Space Panic (mother of all platform games) and involves digging holes and luring the aliens into them, then quickly refilling the holes before they can escape.  I could see using Suzuki's map to do a very similar thing, except you lure the enemies to the central road area to be crushed.  Here is a video of Heiankyo Alien:

The main problem I can see would be in how the enemies react when they reach the road.  My thought is that they should always proceed across to the other side.  That way they won't just end up wandering in to the central road zone to be killed.  The player would then have to use careful timing and reading of the enemies' tracking behaviour to lure them onto the road for a sure death.

But that suggests that there might be something to the "walls and wall breaking" of the second sketch.  Perhaps the dynamic that prevents the aliens from simply wandering to their deaths on the road as they track the player is that they are compartmentalized.  If they break walls at at a slower pace, then that might give the player time to create paths to lure specific ones to an exit to the road that the player creates.  That dynamic would have aspects like the wall breaking in Nostromo, which requires 2 or 3 shots before the player can blast through the outside wall of the 8 lockers/rooms.

Szczepaniak seems to agree that the wall breaking dynamic has some advantages.  In his email he says:

"I probably won't attempt [to recode] Suzuki's description, since the way he described the random forest, and hiding in the gaps, sounds difficult to implement."

It might not be that difficult to implement, but it is unclear how the tracking enemies will interact with the road and its continuously moving obstacle and whether the resulting dynamic would provide an interesting game or just a exercise in temporary avoidance until the enemies wander to their deaths.  But the second variation also has unknow dynamics.  If the aliens are randomly placed then they will begin tracking towards to the player.  Perhaps they need several attempts before walls breach for them, but will there really be enough time for the player to create trails and lure specific ones into following to the road?  Should the player move slightly faster?  Once on the road how quick (what rate) should the obstacle move?

As Charlie pointed out the obstacle's movement rate (the ratio of obstacle's movement rate from top to bottom or vice versa and the player's movement rate) can't be so quick that the player can't make it across or up or down and back into the "forest". And for both variations, can the enemies get swept before the obstacle (as it moves either up or down) but still move right and left to possibly escape back into the forest before being "crushed" at top or bottom wall edge as mentioned in the interview transcript?  That would certainly increase the need for timing one's luring.

And for an MC-10 re-code (same screen as the NEC-PC6001) would the reduction from the 40 by 25 screen of the PET to a 32 by 16 screen disrupt the original dynamic regarding these ratios?  So many unknowns...  In the end, I will simply have to code each dynamic, do some tweaking of the relative movement rates and see if some kind of playable game "emerges."  And then perhaps send some video to the original author and players to see if either variation comes close.

Here is an actual page from S's book with part of the transcript:

Any comments or suggestions would be much appreciated.