Open questions

What is not known, said as what it is. Every byte of the cartridge is assigned to code or data and the listing reassembles byte for byte, but that does not mean everything is understood.

Which distance band is the near one

Creatures pick one of their four drawings from the high byte of ix+6 (0x60, 0x78, 0x90 and 0xA8, p09:AA86), and on the sheet they go from smallest to biggest. Which band means close has not been measured: the dumps do not settle it because the most common creatures jump, and their row does not tell the distance.

The Game Master header puts the stage where the Game Master expects lives

The second header, the one at 0x4010, declares thirteen stages and gives the addresses of the things the Konami Game Master knows how to touch. One of them lands where the Game Master expects the lives counter, and what is there is the stage. It may be deliberate —so that the Game Master lets you change the stage from its MODIFY MODE— or it may be a slip by Konami. There is no way to decide from this binary alone.

What four of the machine's six symbols are

The cherry and the bunch of grapes are unmistakable in the drawing taken from the ROM. The other four are not: there is one in tan tones with green underneath, a red one, a blue one and the jackpot, which is white on transparent. They are published as they come out and readers can make up their own minds.

The exact rhythm of the slot machine trick

The three reels are drawn in the same loop in the same frame, with an identical piece of code between one and the next, so their indices are r, r+k and r+2k with k fixed. And stopping a reel makes the loop do one turn fewer, so the frame takes slightly less and R advances differently.

Which is why pressing at a fixed rhythm gives you not randomness but always the same pattern. What is not measured is what that k is modulo 16, because it depends on how many instructions the CPU fetches inside the loop, counting the BIOS WRTVRM routine, and that differs between MSX models. In other words: the trick could work on one machine and not on another.

Why it does not start on a Philips VG-8020

It is measured that on that machine the cartridge sits on the BASIC boot screen —0xE003 does not move and 0xF0F7 does not go to zero— and that with C-BIOS_MSX1_EU it starts without trouble. Why has not been worked out. The header is correct, the AB and the INIT at 0x406A read back fine at 0x4000 from the emulator's own console, and even so the BIOS never calls it.

The tables that only cover thirteen stages

p00:4105 and p00:4113 index two tables with the stage number, and both have thirteen entries when there are twenty-four stages. With a stage from 14 onwards they read bytes past the end of the table. It has not been checked in motion what comes out of there, nor whether the game ever reaches that path.

The sound, from the inside

The scores are fully declared and the language is understood —FE 00 changes the +0x0E marker and the score switches from the effect language to the music language— but no player has been written to get them out of the cartridge. Until there is one, what there is is a description, not a proof.

What goes by on the sides

p01:6AA8 works by frames, not by distance, and the road check skips its cells. What comes out on the sides is neither drawn nor checked.