[DISCUSSION] Floater / Glider vs Splat Pad - Always immune or current behaviour?

[namida]

I'm wondering whether the current behaviour for floaters and gliders when encountering a splat pad makes sense.

Currently - only if they've started pulling out their umbrella / glider, they'll be safe. Especially for floaters, which fall some distance before pulling out; if they hit a splat pad before this, a splat will occur. See attached level. (The attached level uses floaters. This is equally applicable to gliders, but less noticable because gliders fall only a very short distance before pulling out their hangglider.)

In this level, notice the difference in behaviour between the 7 lemmings:

1. This lemming doesn't move downwards far enough to become a faller, so he does not die.
2, 3, 4. These lemmings do move far enough to become fallers, but not enough to begin floating. They splat. In the case of lemming 2, the difference from lemming 1 (visually, prior to splatting) is almost unnoticable.
5, 6. These lemmings get part way through pulling out the umbrella. They survive. This is consistent with the behaviour that would occur when assigning a floater just before a lemming hits the ground.
7. This lemming pulls out his umbrella completely, and of course, survives.

The case I'm mainly wondering about is the case of lemmings 2 to 4. Should we:
(a) Keep the current behaviour. Although they may have the floater / glider skill, they don't get a chance to activate it, and thus should splat.
(b) Change it so that these lemmings would survive. A lemming surviving a fall via a single frame of attempting to start floating is a well-established, accepted and widely used mechanic; and this is arguably more consistent with that - that it's more about the lemming having the skill. (Indeed, except by changing the max fall distance, you'd never be able to set up a situation where a floater splats due to not having time to pull out the umbrella (except by using a splat pad), even though there is no explicit prohibition on this.)

[ccexplore]

Does NeoLemmix keep the Lemmings floater behavior where, if you assign the floater only at the very end of a long fall, the lemming still survives the fall even though you'd never see the complete parachute-opening sequence?

If it does, then you already have a case that suggests timing of parachute opening has no effect on survivability.  Thus the natural conclusion is that the floater skill alone is sufficient condition for immunity.  In which case it would seem more consistent to extend such blanket immunity to the splat-pad case as well.

[ccexplore]

Does NeoLemmix keep the Lemmings floater behavior where<snip>

Sorry, I just realize you kind of already cover that in 5-6.  Still, I think in my mind, it is already a little strange that a chute that barely started opening (to the point where the falling speed hasn't even reduced yet) is already sufficient to grant immunity.  It feels more natural if we can interpret it simply as a blanket immunity instead.

To put it differently, while technically in the current implementation the parachute actually requires a minimum fall distance to take effect (but no relation to how long the lemming has actually possessed the floater skill relative to when he started falling), because that minimum distance is smaller than the fatal fall distance in the absence of splat pad, this detail of the physics is not easily discoverable from black-box observations.  The only hint is the timing of when the parachute starts opening, but otherwise no obvious clue that it is meant to affect survivability.  Hence this makes it conceivable for someone with no access to the internal code to have an understanding that do not involve a minimum fall distance for floater effectiveness.  So it may be a little confusing for such a detail to suddenly become exposed in the context of the splat pad.

[Simon]

Cull the splat pad entirely? The splat pad makes visual pathing counterintuitive. You can get a similar effect with fire over the walking path to fry floaters.

-- Simon

[namida]

Cull the splat pad entirely? The splat pad makes visual pathing counterintuitive. You can get a similar effect with fire over the walking path to fry floaters.

I think you misunderstand how splat pads work. Essentially, the safe fall distance onto a splat pad is zero (and likewise, onto an antisplat pad is infinite). Floaters and gliders (under current mechanics, assuming they fall far enough to pull out their umbrella / glider) can still survive, because they aren't affected by safe fall distances.

The issue being raised here is - currently, if a lemming (who has one of these two skills) becomes a faller, then lands on a splat pad before they have a chance to pull their device out, they will splat. The question is, should this be changed so that floaters and gliders will always land safely on splatpads.

Lemmings Plus IV, Bumpy 17 "Lem Your Lem Could Smell Like" is an example of a level that relies on this distinction.

[Simon]

I think you misunderstand how splat pads work.

Okay, I rephrase your explanation to make sure I understand.

I assume this behavior for (a): Lem dies on splatpad iff (
    (falls onto it)
    and (doesn't merely step onto it without becoming faller first)
    and ((is not floater and is not glider) or has not opened chute)
).

This is highly unintuitive. You first generate a safe interval (where you don't become faller), then an always-deadly interval (where you don't open the chute), then an interval that is conditional on permanents.

It breaks the mental pathfinding because it was a reasonable assumption in L1 that floaters could never splat.

Now compare with (b): Lem dies on splatpad iff (
    (falls onto it)
    and (doesn't merely step onto it without becoming faller first)
    and (is not floater and is not glider)
).

I don't see why you need such a gadget. This looks like you can always replace it either with a long fall, or fire over a safe walkway.

If you disagree, please post image of level, and show why you can't use either a long fall or fire over the walkway.

The issue being raised here is - currently, if a lemming (who has one of these two skills) becomes a faller, then lands on a splat pad before they have a chance to pull their device out, they will splat.

Yes, and the cull solves this issue.

-- Simon

[namida]

Currently, no level (as far as I know) relies on this distinction in practice, so little if any damage would be done by making the change suggested in original post. Such changes are better made sooner rather than later, before we get to a point where there are levels that rely on it.

The next version's release is too near in the future to cull it for this update.

So under this assumption that what's needed is a decision on how they work, not a keep/cull decision, what one should be used?

[Nepster]

I would keep the current behavior for now, until we have reached some conclusion whether splat pads should exist and how they should work exactly. At that point, I would change them once. This way it is less work for level designers to ensure that their levels still work.

I have no opinion on the falling floaters/gliders issue.

But regarding the general splat pad mechanics, I would like to discuss another detail as well: Currently (anti-)splat pads only work, if their trigger area overlaps the terrain below. If a lemming just falls through them, the pad has no effect at all. If the level designer hasn't been very careful, this creates problems on uneven ground terrain: At one pixel the object works and at the next it doesn't. Without looking at the clear physics mode and knowing about the precise mechanics in the first place, the player won't be able to determine when the pad will work and when not.
Precisely this problem appeared in a level of Wafflem+Flopsy's collaboration level pack, that I tested just now (Inviting 4, for those who have the tester version). There this turned (I believe accidentally) the first anti-splat pad into a very mean trap .

[bsmith]

I was leaning toward floater/glider = always safe landing, but after Nepster's post regarding uneven terrain I think it might be better to remove the splat/anti-splat pads altogether.

[Simon]

Such changes are better made sooner rather than later, before we get to a point where there are levels that rely on it.
The next version's release is too near in the future to cull it for this update.

Hmm, good reason not to cull immediately, thanks.

I don't have a clear like for either (a) or (b). Nepster's case for (a) sounds convincing, have maximum back compat now, then reconsider the issue in medium term. Ask level designers not to rely on the splat pad for now?

-- Simon

[ccexplore]

I don't have any opinion on keep/cull at this time, but I am curious as well as to how splat pads have been used in practice thus far, and potentially the history of how it came about assuming it's more than mere "this is easy to implement and might be interesting to trial out".  It does sound like the only effect of interest is modifying the safe fall distance in specific places in the level.  I do think Simon may be somewhat underestimating the difficulty of replacing the pad with other setups, at least when you want to take visual impact into account (for long falls which takes up a lot of space) or ease of routing around the setup (for fire over walkway, although the splat pad doesn't sound all that much harder to route around to be fair).  Hence why it would be interesting to study how they've been used so far.  With no prior impressions whatsoever on how they've been used, it does feel slightly more gimmicky than genuinely productive puzzle-wise, but frankly I'm not very imaginative with things I haven't personally experimented with so I won't trust my own gut opinion either.

[namida]

The initial reason for implementing them was for compatibility with Cheapo content, but there has been cases that actually use them, primarily in Lemmings Plus IV. Some of them could be worked around with other designs. Bumpy 17 is an example that might be a bit trickier, though probably still not impossible.

Culling antisplat pads is a less severe change, as their effect can be almost perfectly replicated by using updrafts instead.

[namida]

Perhaps more long-term, the issue of whether or not splat pads should be culled altogether should be looked at.

In the meantime, the argument for "keep backwards compatibility" doesn't really hold without an existing level that relies on the current behaviour and wouldn't work if changed, whether that's in the sense of "required as part of the solution" or "prevents a backroute". I feel that the proposed behaviour is more in line with what a player would expect, based on (in particular) the assign-floater-just-before-landing scenario. Is there such a counterexample level already in existance? If not, I'd like to make this change for V1.48n, rather than potentially having a situation where the behaviour is changed after the current one has become more ingrained if the eventual decision is not to cull the splat pads.

[Dullstar]

Considering that floaters/gliders can survive long falls as long as they are given the skill at any time before they become splatters, they should also be able to survive a fall onto the splat pad for the sake of consistency.

I have no problem culling these, however - anti-splat pads are made redundant by updrafts, and I haven't messed with splat pads enough, nor have I played enough levels containing them, to be attached to them.  The anti-splat pad could honestly be kept as a re-skinned updraft for compatibility purposes. It might cause some strange behaviors with gliders, but I doubt it would be level breaking.

[namida]

Off the top of my head (there definitely may be more that I'm not thinking of), the only level that has gliders and contains anti-splat pads is the splatpad / anti-splatpad tutorial level in the NeoLemmix Introduction pack.

Considering that floaters/gliders can survive long falls as long as they are given the skill at any time before they become splatters, they should also be able to survive a fall onto the splat pad for the sake of consistency.

This change has already been made as of V1.48.