Project progress: 80% to 85%

Finish documenting DitherScreenBuffer, though that random stuff is still weird (though it's just a way of choosing a random pixel byte from the column buffer).

Back to ApplyTactics part 4.

For this short bit of code we need to work out enemyData8, sub_C1AA7 and enemyData1.

Calls to sub_C1AA7 are controlled by bit 7 of enemyData8, and enemyData1 seems to contain a number rather than flags (as it is affected by DEC and INC elsewhere).

Probably a good idea to analyse sub_C1AA7 for clues here.

sub_C1AA7 iterates through the objects and looks for a tree-on-a-stack or a boulder that's exposed on top of a tile and whose tile can be seen by the enemy.

In other words, a tree or boulder that the enemy can drain.

So:

• sub_C1AA7 -> FindObjectToDrain

And document it, as it's quite straightforward.

Back to ApplyTactics, and finish off part 4 (though I still don't understand enemyData1).

On to ApplyTactics part 5, which we can split into two (part 5 and 6): looking for robots to drain, and rotating.

It's all fairly clear, and as a bonus we have now identified all the game's sound effects, so I can replace all the "???"s in the various sound routines and variables with the details.

Document parts 5 and 6 and move on to part 7.

ApplyTactics part 7 drains energy from the object it's called with.

targetVisibility is put into enemyData7, so rename to enemyVisibility.

It's fairly straightforward, draining energy in the usual way unless the target is partially obscured, in which case it calls sub_C197D, presumably to think about turning a tree into a meanie... but let's have a look.

sub_C197D, looks like enemyData1 stores an object number as we scan from 63 down, looking for a tree that can be turned into a meanie, so:

• enemyData1 -> enemyMeanieScan

And as this is looking for trees to turn into meanies, so:

• sub_C197D -> ScanForMeanieTree

And:

• enemyData3 -> enemyFailTarget
• enemyData4 -> enemyFailCounter

as they keep track of failed meanie scans.

Also, this makes more sense as all the reset data is meanie-related:

• ResetEnemyData -> ResetMeanieScan

Finish ScanForMeanieTree and finish ApplyTactics.

Good progress today!

A quick win here - finishing off the main variable workspace ay &0C00.

Just need to convert L0C40 to L0C44 to polygonPoint, as this is where the (up to) five polygon points get stashed in the polygon routine... and that's the workspace done.

>>> I have now broken through the 85% barrier <<<

A big leap because the workspace is a big one, though in fairness it has been updated slowly throughout the project.

• sub_C191A -> GetPlayerDrain

This works out whether the player is exposed to an enemy and whether they enemy can see the player's tile, and it sets the variables accordingly.

It doesn't do the actual tile visibility calculation, it just uses the values from enemyVisibility etc. so it's pretty simple.

On to the final category - drawing polygons (there are some buffer variables that I haven't explained yet, but they seem to be tied into the drawing routine, so I'll work them out when I get there.

First spot is a quick variable change:

  LDA L54A0,Y
  SEC
  SBC L54A0,X
  STA T
  LDA L0B40,Y
  SBC L0B40,X

So L54A0 is a low byte and L0B40 is a high byte from a 16-bit value, so let's rename:

• L0B40 -> L0B40Hi
• L54A0 -> L0B40Lo

DrawPolygon is documented apart from the two calls to sub_C2D36 and sub_C2299, and some logic around the values of L002C (for left row buffer) and L002D (for right row buffer).

Here's the hierarchy:

  DrawPolygon
    sub_C2D36
      sub_C2EAE
        Absolute16Bit
        sub_C2EAE
        sub_C316E
    sub_C2299

Let's look at sub_C2D36 first, which returns the C flag to control the call to sub_C2299... but this will have to wait until next time.