Concurrent Tentacle Instructions

I mentioned three phases in the tentacle walk. Using kayaking terminology these would be entry (tentacle makes contact with the surface), pull (player moves forward), and recovery (tentacle moves forward to prepare to make contact again).

Here’s the walking animation from yesterday.

I want the entry/contact phase and the recovery phase to overlap. While the tipward parts of the tentacle are moving forward for the recovery phase, the baseward parts can start making contact with the ground.

A problem I’ll probably see is that the second instruction will drag the tipward segments and then the first instruction will try to rotate the segment it’s working on. These might work at odds and/or just look silly.

One solution to this might be using relative angles instead of global angles. If each segment is trying to match the previous segment’s angle instead of a global angle then the results will probably be better. Probably..?

I ran into another problem before that, though. The second instruction can catch up to and pass the first instruction. Each tentacle instruction rotates its current segment until it’s pointing in the desired direction and then it moves onto the next one. So if the difference between the starting angle and the target angle is very large for the first instruction and then very small for the second one, the second one will overtake the first.

Here’s a demonstration.

Changing from the first target angle to the second target angle happens almost immediately for those first four segments. Then after that I think both instructions work on the segments at the same rate since they’re both doing a 180° rotation. In this case, there’s a bend at the fifth segment because it looks like second instruction completes its rotation first (it’s closer to the starting angle so that makes sense) and then the first instruction completes it. So we’re left with the angle from the first instruction.

How to deal with this.

• Instructions move at a fixed rate down the tentacle. This would probably mean I need to separate the rotation of the segment from the segment index of the instruction. Segments can continue to rotate after the instruction has moved onto other segments.
• Instructions over-write each other. If a newer instruction crosses paths with an older one the old instruction is dropped.

Thinking the first one through a bit, I don’t think it’ll work. Even one instruction could result in multiple segments all rotating at the same time and the frontier (the tentacles that are all still rotating) will end up really wide. I’ll lose the nice effect of the currently rotating tentacle segment dragging the other segments behind it.

Instructions overwriting makes sense. At least for these instructions to point in a particular direction.

Question then. Should I just never have more than one instruction running? …no. Having the two running at once is still necessary to combine the recovery and contact phases. I still want the rest of the tentacle to form a line in the direction of motion after the … hmm, well, that’s not true is it? Theno, I think it is true. Lol. Okay, sticking with multiple instructions at once.

Here’s the updated behaviour with canceling older instructions when their current segment is overtaken. Only one angle.