[DISC.][PLAYER] Jumper physics

[WillLem]

On the "jumper hits wall" scenario - I do get the feeling many people are only considering the case in the left in the attached pic. However, the case on the right is likely where the difference could lead to some interesting situations - so don't forget to consider that too.

The scenario on the left produces these two differences:

A: Jumper hits the wall, turns, lands, then walks away from the wall.
 i) Direction-wise, no difference has occurred here than if the lem had just walked into the wall without jumping.
 ii) If there was a pickup skill at jump height, then the jump would be useful.
iii) There is no further opportunity to interact with the wall.
 iv) If no further action takes place, the end result is that the lem walks away from the wall.

B: Jumper hits the wall, slides down, then walks towards the wall for e.g. 1 frame before turning around.
 i) Direction-wise, no difference has occurred here than if the lem had just walked into the wall without jumping.
 ii) If there was a pickup skill at jump height, then the jump would be useful.
iii) The lem has a 1-frame opportunity to further interact with the wall, e.g. by climbing, bashing, etc.
 iv) If no further action takes place, the end result is that the lem walks away from the wall.

The scenario on the right produces these two differences:

A: Jumper hits the wall, turns, lands, then walks away from the wall.
i) The lem walks back the way it came, meaning that the jumper has ultimately functioned in place of a walker
 ii) A walker/other skills would be required to turn the lem around if needs be.
 iii) If there was a pickup skill at jump height, then the jump would be useful.
iv) If no further action takes place, the end result is that the lem walks back the way it came.

B: Jumper hits the wall, slides down, then walks forwards (underneath the wall).
i) The lem continues to walk forwards.
 ii) A walker/other skills would be required to turn the lem around if needs be.
 iii) If there was a pickup skill at jump height, then the jump would be useful.
iv) If no further action takes place, the end result is that the lem continues to walk forwards.

Therefore, the second scenario does indeed produce more differences. However, whether one is more preferable than the other depends entirely on the layout of the rest of the level; they do not carry any particular advantages either way in and of themselves.

The first scenario, however, produces a difference which does carry a specific advantage: i.e. the opportunity to further interact with the wall.

[namida]

Another special case to consider: Jumping while in an updraft.

Updrafts are of course most notable for their effects on gliders, but they do make fallers and floaters fall more slowly too. Should they also have an effect on Jumpers? I'm leaning towards yes - perhaps on the order of gaining one extra pixel of height (or as applicable, losing one less) per frame. This is fairly comparable to the effects it has on floaters / fallers.

[Proxima]

Therefore, the second scenario does indeed produce more differences. However, whether one is more preferable than the other depends entirely on the layout of the rest of the level; they do not carry any particular advantages either way in and of themselves.

The first scenario, however, produces a difference which does carry a specific advantage: i.e. the opportunity to further interact with the wall.

As I said earlier, an advantage to the player may be a disadvantage to the level designer, and vice versa. Also, in the case where the lemming simply lands at the base of the same wall, the player could have gained the same advantage (being able to interact with the wall) by not jumping (or jumping from further back and not hitting the wall, in the case where the jump was necessary to collect a mid-air pick-up skill). Finally, any advantage that depends on gaining a 1-frame window adds to fiddly execution, which it's good to avoid where possible.

The case where walking straight forward would go under the wall is much more interesting, and the player does gain an advantage here: the ability to choose whether the lemming turns around (by jumping) or does not (by not jumping).

Regarding updrafts, my inclination is to say "no" for simplicity, but the altered trajectory would be visible with the skill shadow, so I don't strongly object to it.

[IchoTolot]

The jumper hits wall and turns mid-air case has one inconsistency with other skills though:

Shimmiers also do not turn mid-air when shimming against a wall. They fall down first and then the possible turning happens. I do think we should not make this case inconsistent, as both lems go face first into a wall and should react the same to not confuse players, regardless if the other options producess more level ideas or gets a few more votes.

EDIT: For the climber -> jumper case, after seing it in action I tend a bit more to the allow side now. The parkour options are quite large, even though the combination can be very overpowered. So I slightly tend towards yes there now.

[NieSch]

For the climber -> jumper case, after seing it in action I tend a bit more to the allow side now. The parkour options are quite large, even though the combination can be very overpowered. So I slightly tend towards yes there now.

Yeah, this jumping/climbing "parkour" lemming looks awesome! (It must be great to have some levels where the parkour lemming jumps around like a madman to save the crowd in time.)

[ccexplore]

Shimmiers also do not turn mid-air when shimming against a wall.

But if I recall correctly and understand your case above correctly, NeoLemmix shimmiers only hop up strictly vertically to try to grab onto ceiling.  As such, even if it was standing right next to wall when assigned shimmier, strictly speaking the lemming's movement never try to put it any closer to the wall.  The shimmier lemming hops up in parallel to the wall surface and basically could care less whether there's a wall there or not.

Whereas the jumper does have a horizontal component of movement, so there is truly a collision happening with the wall, the result/effects of which is what we want to decide.

[Proxima]

I assume Icho is talking about when the shimmier bumps a wall ahead of them during shimmying, not the initial reaching phase.

Even so, I think there is an argument why it's not "inconsistent" (I am getting to despise that word!) to have the jumper turn. The shimmier's purpose is to shimmy along a ceiling until it stops, then fall. Hitting something ahead of them is one of two ways the end of a shimmyable ceiling might be marked (the other being a sudden rise). So the lemming would be prepared to fall down at that point. The jumper, by contrast, intends to complete the arc of their jump; hitting terrain is a sudden interruption.

[Strato Incendus]

Again: How does this work in Lemmings 2: The Tribes and in Lix? Bouncing off the wall into the opposite direction always seemed like the most intuitive thing for me. So I'm pretty sure I've seen this before.

Especially in the case on the right of namida's picture, this would allow to turn around a single lemming by having him jump and turn, while everyone else slips through the gap and continues walking to the right.

The consistency here would be - as I've described it in the pre-level text of the very first level in Lemmings World Tour - "Whenever a lemming performs a skill and runs into terrain that skill cannot affect, the lemming will turn around."

--> constructive skills, when performed into terrain of any type, cause the lemming to turn
--> destructive skills can affect regular terrain, but when they hit steel, i.e. terrain they can't affect, the lemming will turn around (exception: Digger)
--> therefore: when the Jumper hits terrain while he is jumping, he should turn around as well

Regarding the Shimmier, I would agree with previous posts that running into a wall is one of the intended ways of stopping a Shimmier, whereas the Jumper is supposed to carry out the full arc of its motion.

[Proxima]

In L2, I believe the jumper turns. I never owned the game and only occasionally played it when visiting friends, so I don't have very clear memories here, but as I said, I am pretty sure I had the insight of using this behaviour to solve a level we were stuck on (it may have been Egyptian 9), so it would surprise me if I was completely wrong about that.

Lix has tumbler behaviour, so the jumper will actually bounce off the wall (and will reverse her facing direction).

[Strato Incendus]

...and even though we technically won't have tumbler physics in NeoLemmix, to my understanding, the engine still needs to be able to calculate a Jumper suddenly completing his jump into the opposite direction - if you clone the Jumper mid-jump. So I guess for this particular example, that wouldn't make too much of a difference, coding-wise.

This raises another interesting question, though:

What happens when you assign a Walker to a Jumper?
Does he stop jumping and fall straight down again?
Does he turn around in the middle of the jump, like a Cloner, just without the original continuing to move in the former direction?

[WillLem]

Constructive and destructive skills turn around upon hitting a wall, sure: but the Jumper is not a constructive or destructive skill. In fact, it's far more comparable to the shimmier (which, as Icho has pointed out, does not turn around).

To take this further to Proxima's point, the lem being "ready" to stop could also apply to a Jumper: if they're jumping towards a wall, they can surely see the wall they're about to hit!

This being the case, why would they then turn around upon hitting the wall? It make far more sense to me that they would grab onto the wall and slide down it; it could even be that climbers could be assigned at this point, and the lem starts climbing.

Or, maybe they bounce off the wall: but they still wouldn't necessarily be facing the opposite direction...

[Strato Incendus]

It make far more sense to me that they would grab onto the wall and slide down it; it could even be that climbers could be assigned at this point, and the lem starts climbing.

That would effectively introduce yet another L2 skill, the Slider. By itself, this would run counter to my expectations, though, because no other lemming falling 1 pixel next to a wall slides down on it; he just becomes a regular Faller.

Or, maybe they bounce off the wall: but they still wouldn't necessarily be facing the opposite direction...

I think this would actually be more difficult to sprite than if the lemming were to face the opposite direction? Meaning, according to your suggestion, the trajectory of the arc of the jump would be reflected by the wall, but the lemming sprite itself wouldn't. If the lemming bounces off the wall AND faces the opposite direction, in contrast, you can simply replace the rightward sprite with the leftward sprite, or vice versa. You don't need further "hybrid" sprites where the lemming "moves to the left, but looks to the right", or "moves to the right, but looks to the left".

[namida]

This raises another interesting question, though:

What happens when you assign a Walker to a Jumper?
Does he stop jumping and fall straight down again?
Does he turn around in the middle of the jump, like a Cloner, just without the original continuing to move in the former direction?

You cannot assign a walker to a jumper. This has already been discussed, and the preference seems to be for "don't allow it". Had it been allowed, the behaviour would've been consistent with any other case of assigning a walker to a lemming who isn't currently walking - the skill is cancelled and the lemming reverts to a walker or faller as appropriate.

[Strato Incendus]

Thanks, I must have overread that. But indeed, this seems like the easiest and most straightforward solution.

[Proxima]

To take this further to Proxima's point, the lem being "ready" to stop could also apply to a Jumper: if they're jumping towards a wall, they can surely see the wall they're about to hit!

In the same way that lemmings can see a fire pit in front of them and avoid walking into it?