[SUG][PLAYER] Fix misleading traps!

[Proxima]

I'm sick of having to design around what is an obvious misdesign in the original graphics sets. Lemmings' "pin" (the pixel that determines where the lemming "really" is) is one pixel below the floor, okay. So traps' trigger areas should extend one pixel below the floor as well. Especially traps that have a visible part to show where the trigger area is. This should sit on the floor; it shouldn't have to be buried. This is 2018 and we do not mislead the player! It is the golden rule of modern level design!

Here's a run-through of all traps in the original and ONML tilesets:

Dirt: Bear trap and rock trap have extended trigger areas. 10-ton weight has a trigger area that stops above ground so it has to be buried 2 pixels
Fire: Firepit's trigger area stops above ground.
Pillar: Rope trap has a "blob" of terrain that's meant to be buried, but even the metal support has to be buried to get the trigger area to work.
Crystal: Trap 1 is meant to be buried, enough remains visible. Trap 2 is okay. Trap 3 has to be buried.
Marble: Squisher has to be buried.
Brick: Wheel has an extended trigger area. Squisher has no indicator.
Rock: Weed trap has an extended trigger area.
Snow: Icicle has no indicator. (But note the the trigger area extends below the rectangle of the trap. It's almost as if the designers realised this was a problem and fixed it in the ONML sets! All the more reason we should fix the Orig sets to match.)
Bubble: Suction pipe has no indicator; again trigger area extends below the trap rectangle.

[Nepster]

I would welcome better trigger areas for these traps. But given the huge discussions about whether to allow multiple teleporters map to the same receiver (which compared to this is an extremely minor issue!), I don't have the energy to really go into this discussion as well.
So: If you manage to convince the community to do this change, then great! But don't expect me to put a lot of energy in this discussion.

[Simon]

Traps in L1 sets should kill exactly as Proxima wants.

Pin in the ground is a terrible convention. Unsure whether to fix.

Related: Display trigger areas offset by 1 pixel; if pin stays in the ground, then at least this should be accepted because the terrible pin in the ground confuses newb and experienced tileset designer alike.

-- Simon

[Nepster]

Simon, I understand that the 1-pixel-offset is your pet peeve, but this thread is not the place to discuss it, especially as the discussed changes to the L1 traps are totally independant from your issues! This is about:
1) Whether to modify the game pyhsics of the L1 traps
2) Whether to modify the visual appearance of the L1 traps
So we are discussing the question "How can we make the traps clearer for players?", not the question "How can we display trigger areas better for level designers?".

[Flopsy]

I personally never had an issue with the current trigger areas, I understand that if a Lemming walks along a trigger area and there is no trigger area in the pixel below its feet which is where its trigger area is then the trigger won't activate.

I always made sure when designing triggers for my tilesets that I always made the trigger area go below the actual object by a pixel or two to allow for this. I adapted and I feel so can other people.

[IchoTolot]

Triggers that end too far upwards:
I can see extending the triggers by a few pixels downwards can be beneficial, but I don't see the slight adjustments you have to make in placing as a big problem. I don't think this would affect many levels and if it does that would be mostly ones where you need to go under traps.

Triggers that extent a few pixels into the ground:
I don't mind them and I am against a fix there. I don't see a problem with lems getting hit 1-2 pixels under the trap and I am generally in favor of a bit more "aggressive traps" that cover a bit more than the pin.

Extra indicators:
I am against the indicators on the icicle, brick squisher and suction pipe trap though. They often make excellent decoration objects on ceilings and even mid-air traps to catchbuilders - an indicator would destroy that option and ruin quite a few decorations.

So, extending triggers by a few pixels downwards I would call ok, but I am against trigger shortening and extra indicators. You still can see when these traps hit you + we've got true physics.

In general I don't mind the problems listed here to be honest.

Another idea:

We could also adjust the graphic itself rather than the trigger area - that would still lead to physics consistancy and only small visual things need to be checked.

[Proxima]

I always made sure when designing triggers for my tilesets that I always made the trigger area go below the actual object by a pixel or two to allow for this.

The issue is precisely that if it has become a convention to extend trigger areas to avoid misleading, then all traps should do this. Otherwise the player can't tell (without going into clear physics mode) which traps mislead and which don't.

And while it's great that CPM exists as a backup, it should never be required to work out what a level design is actually communicating to the player. The level itself should contain all the information needed for the player to solve it.

I don't see the slight adjustments you have to make in placing as a big problem

I should add, the issue that provoked me to abandon editing for today and post this topic was that the squisher trap in Marble doesn't work without being moved down, which means it no longer conforms to the grid and has a weird air pixel above it instead of hanging nicely from the ceiling.

Extra indicators

I wasn't suggesting extra indicators, just that traps that have indicators should have trigger areas that match what the indicator is communicating.

[IchoTolot]

I should add, the issue that provoked me to abandon editing for today and post this topic was that the squisher trap in Marble doesn't work without being moved one pixel down, which means it no longer conforms to the grid and has a weird air pixel above it instead of hanging nicely from the ceiling.

L1 actually has a grid? I think all the pieces are gridless anway, with a lot of weird seize numbers.

And having 1 piece in a set that needs grid adjustment I wouldn't call a huge deal to be honest, that would need significant changes. I would simply solve your problem with moving the ceiling down 1 pixel with it or extend the ceiling.

Again trigger changes with the extenion of 1-2 pixels rows below the current trigger or a slight graphic change I would call ok.

[namida]

I set most of the NL trigger areas in official sets, so let me chime in here: Mostly, if the trigger area seems off, it's based on (my judgement of) how the trap looks visually. Using the ten-ton trap example, I felt that it would look weird with the plate exactly in the ground. Maybe this should've been discussed. It mostly happened during the early era of NeoLemmix, back when there was little interest, so most decisions ended up being based purely off my own judgement.

Regarding going forward, best not to change physics if not absolutely necessary. Can graphical changes (most likely, for the most part it'd just be offsetting) address the issue?

[Proxima]

I'm not convinced that changing physics is not the way to go. Firstly, changing numbers in a few text files is a lot easier and quicker than mucking around with the graphics. Secondly, it's true that this would break levels where the architect placed the trap so as to look deadly but actually be safe -- but hopefully there are not many such levels anyway! If the trap is already placed slightly below ground so as to be deadly with existing physics, this would only make the trigger area extend a little further below ground. Some replays may break if a bomber was used to get past the trigger area, but a solving replay could still be produced by placing the bomber more precisely.

[Dullstar]

Under the assumption we're discussing expanding the lower boundary of the trigger area and leaving the upper boundary as-is:

The only possible issue I see with this change is the possibility of breaking existing content, but in this case the "existing content" in question would almost certainly have to be poorly designed. Can anyone think of a situation in a well-designed (I realize this is subjective) level where extending the proposed change would cause a problem? Because, to me, this situation sounds like it would only have the potential to break levels that rely on having traps floating in the air for some probably dumb reason in situations where that doesn't visually make sense, and if I've ever made anything like that, I am perfectly okay with it being broken by this sort of change, because that's just bad design.

I would rather alter these trigger areas than alter the graphics; after all, no level designer should be making levels that require knowledge of the exact size/location of trigger areas in order to solve, so I find it unlikely that any well-designed content will be broken.

I believe the benefits of making this trigger area change greatly outweigh the disadvantages.

Related: Display trigger areas offset by 1 pixel

For the record, I think this is a good idea as well.

[nin10doadict]

I do wonder if shoving the trigger areas down a pixel or two would be the best way to go. What about levels that require bashing/digging/etc just low enough to get under a trap? I feel that modifying the graphics might be better. I could change the spritesheets for those traps in an hour or two, I think. Then again, while this wouldn't break any replays, it might break certain aesthetic choices. I suppose if we change anything we'll have to decide which would be the lesser of two

The whole 'pin in ground' convention is something I've gotten used to, but the operative word there is "gotten." It wasn't how I thought it should work at first. It may be best not to muck with this for now, and perhaps open discussion on it later.

[IchoTolot]

I think a trigger area expansion is the best suited here. If we decide to change something.

Changing the graphic will cause EVERY level to be checked for asthetic errors.

Expanding the trigger 1-2 pixels downwards will likely only break a few replays and maybe some single levels will need slight adaptations.

[Proxima]

I just noticed that in Fun 15, the rope trap has its metal support buried only one pixel -- in NL that's not enough to make it active, the metal bar has to be completely buried. So, how did we deal with this issue in Redux?

We just didn't notice it. The trap is indeed not active in the NL version.

So, add "accuracy to the original game" as another reason in favour of this change.

[nin10doadict]

Ok, yeah, Fun 15 being broken like that is pretty bad. Admittedly if you ever go into that pit you've screwed up, but still.
That little thing has pushed me more toward adjusting the trap triggers. I can think of at least one of my levels that will likely break because of this: "Fall down go boom!"
If the hitbox on the ghostbuster-type trap gets bigger then this level just won't work because...

Spoiler (click to show/hide)

...it involves going underneath the trap and covering the trigger with a couple of stackers. A taller trigger won't get covered.

[Proxima]

Namida gave me permission to bump this, because it was wrongly moved to "Closed" despite no fix being made.

I feel disappointed that we couldn't get a consensus on this. I think I'm right in saying that every other time there's been a discussion of misleading game mechanics (fake and invisible, decoration that looks like terrain, etc) we've come down on the side of "don't mislead the player" even if it meant changes that risked breaking existing content.

Maybe an illustration would help, so here's a snapshot from a RockLems level.



The trap is placed so it sits snugly on the terrain -- which with current physics means it's completely harmless. To make it effective, I would have to bury it in the terrain, which is both bad aesthetically and makes it less easily visible -- especially bad with this specific trap as it's long and thin, so even burying it one pixel loses a lot.

Please let's just fix this.

[IchoTolot]

I think the 2 most basic fixes that would most likely not break anything would be:

1.) Expand the trigger areas 1-2 pixels further below.

That would 100% fix cases like the shown crystal trap, or the 10t masher. As objects are placed on the ground there would only be rare cases where level or replay adjustments would be nessesary.

2.) Adjust the trap's graphic 1-2 pixels higher, so that the trigger area stays the same, but the graphic is now in a more fitting place.

This would lead to no replay adjustments being nessesary, but there would be some visual adjustments nessesary on some levels.

[Turrican]

As a middle solution , I think we could keep the existing traps to keep 100 % backwards compatibility with existing levels , and for the problematic traps we could also add versions of them with adjusted graphics as IchoTolot suggests .

Now , I know that's like adding adittional traps in the game , In that case I think we could even make their graphics a little different , to make it clear that's these are not 100 % the same trap .

In that case I think we could handle the issue Proxima showcases , while maintaining 100 % backwards compatibility with existing levels .

[Proxima]

Sorry, but I think that's a really bad idea. We'd gain something for making new levels, but existing levels would continue to mislead. And players would be faced with two different versions of each trap and would have to learn which one has which trigger area.

My preferred fix is simply to move the trigger areas 1 pixel down (2 in a couple of cases). This would only disrupt levels where the trap is, right now, at the exact height where it looks deadly but is safe. In that case, either the designer intended it to be deadly and we'd be fixing the level, or they intended it to be safe even though it looks deadly. I won't be shedding any tears over such designers having to fix their levels.

[Dullstar]

I agree with the suggestion to extend the trigger area. I don't think breaking a few replays is all that bad, and I imagine few if any levels involve such a degree of precision that a small downward expansion of the trigger area would outright break them, and I would not complain about such levels needing to be fixed, nor would I lose any sleep over potentially breaking levels whose solutions rely on the traps looking deadly but actually being safe.

If we fix it by moving the graphic to better match the existing trigger area, while this would guarantee no existing levels are functionally broken, it could make existing content require aesthetic fixes, and there's probably significantly more levels that have the exact trap placement influenced by aesthetics than there are levels that would be broken by a slight downwards extension of the trap trigger area.

[IchoTolot]

My preferred fix is simply to move the trigger areas 1 pixel down (2 in a couple of cases). This would only disrupt levels where the trap is, right now, at the exact height where it looks deadly but is safe. In that case, either the designer intended it to be deadly and we'd be fixing the level, or they intended it to be safe even though it looks deadly. I won't be shedding any tears over such designers having to fix their levels.

That also would catch the cases where the height of the trap trigger area is used to block lems to a certain degree trying to get over it, or trigger area blocks in general.

An extension like in 1.) still has the least impact on the content as a whole while also fixing the problem.

[namida]

I would likely have to do some testing to say for sure, but just from reading and thinking about the descriptions, I find myself agreeing with IchoTolot here - that extending the trigger area is going to be the best way forwards.

That of course doesn't mean it's set in stone as "that or nothing" - let's keep discussing this, we've got a while until a stable release where such a thing could be changed so no super urgency.

With this being said, there is one thing that I am going to set in stone right now. I want this to be the final discussion on changes to trigger areas for the official styles - if the decision is "no change", that's it, case closed permanently. If the decision is "change", this will be the last time they get changed. So whatever we go with - we need to give good consideration to all objects in official styles, not just traps. This doesn't mean the same changes are needed of course - we might decide "exits are fine, but we do need to adjust traps", or "style A's objects are fine but style B's need changes". All I mean here is - once this discussion is wrapped up and any changes coming from it are made, that's it, the trigger areas for the official styles, both in absolute terms and relative to their graphics, are considered finalized.

[ccexplore]

Maybe an illustration would help, so here's a snapshot from a RockLems level.

Is it possible to see a screenshot of this in clear physics mode?  I'm just curious how the trigger areas has been set currently that led to that setup as harmless.  Is it fair to say maybe that particular trap was more problematic than other ones?

[namida]

Maybe an illustration would help, so here's a snapshot from a RockLems level.

Is it possible to see a screenshot of this in clear physics mode?  I'm just curious how the trigger areas has been set currently that led to that setup as harmless.  Is it fair to say maybe that particular trap was more problematic than other ones?

Here's a screenshot of the same objects at the same heights above the terrain, in clear physics mode.

[namida]

Having thought about this, I recall that last time trigger areas were revised for traps (and I believe I applied this to both the official sets, and all Lemmings Plus sets that existed at the time; and further also applied the same rule to teleporters), in most cases it was based on a certain amount above or below where the actual lemming sprite appears in the first animating frame of the graphic.

That particular trap needing to be fairly buried would be very consistent with such an approach. However - to be clear - I'm just explaining here how the current situation arose, and perhaps it is worthwhile to revise this, at least for some traps.

[ccexplore]

It looks like the official Lemmings 1 levels only used that trap once, in Taxing 28 (POOR WEE CREATURES), and it apparently requires burying trap there just as it does in Proxima's example, by the exact same amount in fact.

So I guess the placement of trigger area for that trap just never had been good; seems like it should extend down at least one more pixel to support the rather natural placement shown by Proxima.

(To be precise, this is tested specifically with DOS Lemmings/Lemmix; I don't know about any other ports.)

[Simon]

I still recommend to change the traps as Proxima wants. Interesting that this went under the radar for a year, but it's a physics change and should be well-implemented on first try.

-- Simon

[Proxima]

So I guess the placement of trigger area for that trap just never had been good; seems like it should extend down at least one more pixel to support the rather natural placement shown by Proxima.

(To be precise, this is tested specifically with DOS Lemmings/Lemmix; I don't know about any other ports.)

I happened to discover an instance just now, browsing the old level review topics. On Genesis Sunsoft 9 (Evacuating a coal mine) the trap is in that exact position relative to flat terrain, making it inert on NeoLemmix. But on the actual Genesis version it is indeed deadly.

[ccexplore]

From Gronkling's speedrun attempts on Genesis Lemmings, we've discovered that some objects' trigger areas in that version are apparently slightly different from DOS Lemmings, in some cases not even rectangular apparently (!) but more irregularly shaped.  Not sure though about the details for the specific trap discussed here, it wasn't talked about on that thread (probably because we never gotten on that thread to any levels making use of it?).

Anyway, I agree that the positioning depicted in your screenshot should be deadly, and it's probably worth at least changing that trap to make it so.  That it impacts one of the official levels (even if from specific ports) just makes it more worthwhile to change.

[Proxima]

Okay, time to make a concrete proposal. I have actually changed my mind to some extent after doing some more investigating, and I'm sorry for taking such a belligerent tone with this topic.

I now think that only two trigger areas need changing: the "ghostbuster" trap in the crystal style (which should be moved down 1 pixel), and the rope trap in the pillar style (which should be moved down 2, or maybe 1 -- and maybe also shrunk a bit).

For comparison purposes, the attached level and screenshot show all the traps at the height that would make them only just deadly, under this proposal (i.e. if they were lifted one pixel they would become safe).

Ghostbuster trap: As just mentioned, this affects an official level (Genesis Sunsoft 9). I want to be able to rest the trap snugly on flat terrain, which is especially important as it's long and thin, and burying it one pixel in terrain loses a lot of visibility. The middle portion would then have 2 pixels showing above ground, which is consistent with other traps (see screenshot and further discussion below).

I would prefer to move the trigger area rather than expand it, as it's already 6 pixels tall, and with the trap being so thin, it looks wrong that lemmings can still be killed when moving high over the visible part of the trap.

(No other trap triggers from the official styles are more than 6 pixels in either dimension. In fact, this is the only trap with a trigger area wider than 4, so maybe it should also be shrunk horizontally? If you look closely then it looks a bit weird, because the lemming is suddenly displaced to the centre for the death animation -- but most of the time this isn't something you would notice.)

Rope trap: This also affects an official level (Fun 15 / Taxing 1). On that level, the bottom metal bar is in fact buried 1 pixel, so moving the trigger area down 1 would be enough to make it deadly. However, I'd like to be able to keep the metal bar unburied (which would require moving the trigger area down 2) as this looks better, at least as regards the support structure. However, there is a counter-argument: if you look closely, you can see a ring of darker pixels on the top 4 rows of the dirt mound. This represents the rope that the lemmings have to step on to trigger the trap. If the metal bar is unburied, then the rope is entirely above ground (there is one pixel in between the top of the terrain and the bottom of the rope). So maybe it should be safe at that height (i.e. move the trigger area only down 1?)

For this one, I feel strongly that the trigger area (currently 6 pixels tall) should be moved and not expanded, and maybe even shrunk. As I just said, what kills the lemmings is stepping on the loop of rope -- it should be entirely safe to pass above it.

* * *

For the other traps, I'll explain why I now believe it's best to leave them as they are. I originally proposed that the marble squasher and 10-ton weight also needed adjustment. I've taken a look through the original levels, and noticed that on levels using these traps (e.g. Fun 24 and Taxing 2), the "pressure plate" (the part that the lemming steps on) is partly or completely buried, so these traps shouldn't be causing any problems with original levels. Also, the plate for the marble squasher trap is 3 pixels tall; that for the 10-ton weight is 4 tall. That's enough that if the entire plate is above ground, it looks as if the lemmings are walking past it and it should be safe. So I now think that I made the wrong call here originally -- though since Simon and others supported my original proposal, I wonder what they think about these traps in particular.

So I now suggest that these traps should be left as they are, which means that if the top 2 pixels of the plate are above ground, the lemming will step on it. As I said earlier, that's consistent with the height of the middle part of the "ghostbuster" trap (and also the bottom platform of the wheel).

[namida]

All looks good to me. A related thought that came up on Discord is whether the falling rock trap should retain the few pixels at the bottom. Between the secondary animation, clear physics mode, and (editor-side) ability to view trigger areas, it seems unnecessary now - I suspect it was originally there to assist level designers with placing the trap.

While this is a purely visual change so not as big a deal to get finalized ASAP, it would be nicer to include it in the same set of changes instead of a separate change later.

Regarding how far to move the rope trap, I'll need to examine it in-game a bit and see how I feel.

[IchoTolot]

I now think that only two trigger areas need changing: the "ghostbuster" trap in the crystal style (which should be moved down 1 pixel), and the rope trap in the pillar style (which should be moved down 2, or maybe 1 -- and maybe also shrunk a bit).

I would agree moving the rope trap by 1 pixel, but not shrink it. For the 2 pixel difference variant I would move it by 1 and extend by another 1. As 6-7 pixels I would still call reasonable.

For the ghostbuster trap I would still tend to extend by 1 rather than to move by 1.

[Simon]

plate for the marble squasher trap is 3 pixels tall;
[plate] for the 10-ton weight is 4 tall.
if the entire plate is above ground, it looks as if the lemmings are walking past it and it should be safe.
Simon and others supported my original proposal, I wonder what they think about these traps in particular.

I'm fine with Proxima's adjustment. I have no strong opinion on trigger areas of plate traps.

Reason: The plate is peaceful by itself. Only when the deadly part of the trap sits on the ground, it should be deadly. But all plate traps have the deadly part hovering in the air.

Instead, in our world where a trap is entirely nonsolid, I consider the thick plate a design problem, and no trigger area would be unarguably good for them. The 10-ton weight in L1 in (Watch Out, There's Traps About) is foreground with hidden terrain behind. The trap looks good like that, but it can't be reasonable to always require hidden terrain with such traps.

In another world, the plate would be earth or steel, not part of the nonsolid trap, but plate and trap together are still a single tile. Such half-terrain-half-trap tiles are impossible in NL, and they would only be useful for these plate traps. Merely an interesting idea for the backburner.

-- Simon

[Simon]

Proxima's image of the 10-ton weight in reply #29 is inconsistent with Proxima's quote from reply #29:

these traps should be left as they are, which means that if the top 2 pixels of the plate are above ground, the lemming will step on it. As I said earlier, that's consistent with the height of the middle part of the "ghostbuster" trap (and also the bottom platform of the wheel).

I recommend that the quote be correct, and that the 10-ton weight/plate may (be raised by 1 pixel and be deadly), but may not (be raised by 2 pixels and still be deadly).



The fire pit (top right in Proxima's image in reply #29), does it come with its own floor? When that own floor is visible by 2 pixels of its thickness (instead of buried in the ground as it is in that image), will the fire pit be peaceful?

-- Simon

[Proxima]

Proxima's image of the 10-ton weight in reply #29 is inconsistent with Proxima's quote from reply #29:

That's because this object's placeholder secondary animation has the plate bobbing up and down, and I happened to screenshot on a frame when it was in between. I don't know if this animation will be kept in the long run.

The fire pit (top right in Proxima's image in reply #29), does it come with its own floor? When that own floor is visible by 2 pixels of its thickness (instead of buried in the ground as it is in that image), will the fire pit be peaceful?

Yes, the fire pit has a 2-thick supporting bar at the bottom as well as at the sides. On the original levels (e.g. Fun 18) this bar is always buried. I'm not entirely sure how I feel about this one.

[mobius]

this isn't about traps but it's a related issue;

I've always disliked the way the button trigger area is located. You can see in this picture what my issue is:

the button must be almost totally buried in order for it to work because the trigger area is high. I get it theory that the lemming has to press the top of the button but then why have the lower part of the button at all? Because if you put buttons on a piece of thin terrain then it'll look weird because most of it is sticking out underneath. I say either move the trigger area or make the whole button graphic smaller.

[namida]

Yes, let's give consideration to all objects in official styles (ie: "orig_*****", "ohno_*****", "default" and "xmas").

[IchoTolot]

I would agree with mobius on the botton there and would propose extending the buttons trigger by ~2 in the downwards direction.

[namida]

Okay so - I didn't look over the topic again, but the changes I've got so far are:
- orig_crystal:trap_03
- orig_pillar:rope_trap

In both of these cases, I've moved the trigger area downwards by 1 pixel. I have not yet fully tested what impact this has on existing content.

I see another change is default:button, with the suggestion to increase the height of the trigger area by 2. Any objections to this?

Any other changes that need to be made?

[namida]

I've implemented the change to default:button. (This change, as well as the others in this topic, will come with V12.7.0, not with the next routine styles update.)

I'm not really looking through these much myself at the moment (I'm focusing on other things), so I'm relying on you guys to let me know if there's any others that need to be changed - so please, take the time to check, especially if this is particularly of interest to you.

[namida]

I've had a look over the remaining objects, and it mostly looks good to me.

The other objects maybe of concern are the splitters in the Brick and Bubble styles (should the trigger area cover more vertical area?) and the unlock button in the Bubble style (I think this one should be moved downwards a bit, maybe?).

Ideally, I'd like for discussions / etc on this to be wrapped up by the time V12.7.0 RC is released (sometime early-to-mid next month), and there will be no further physics changes to any existing objects in the official styles once V12.7.0 stable is released (except to fix mistakes, if any are made with the update) - so this is very much a "speak now or forever hold your peace" situation if you think anything needs further changes.

[Proxima]

I think the other thing that needs deciding is the question Simon raised about the fire pit object:



At the moment, the bottom supporting bar needs to be buried for the fire pit to be deadly. Does that make sense or should we change it?

I don't have a very strong opinion on this one. The bar looks so similar to terrain that if it's not buried, it looks like the object is raised above where the lemmings are walking. Then again, it makes sense to take damage from being that close to fire!

[namida]

If we look at the original game, in almost all use cases of this object, it's buried in terrain. Usually, this terrain goes far enough that it fills in the gaps between the bottom of the object and the flames, even.

The only exceptions I can find are:
- Mayhem 2. In this case, the flames are used in a similar manner to water along the bottom of the level, and have no terrain under them in the first place - though they are placed high enough that their trigger area does not reach to the bottom of the level.
- Sunsoft 22. This isn't a true exception as such, but rather, two fire objects are placed above the terrain, with extra terrain in turn placed along the bottom of the fire objects. There's a third fire object that's directly buried in normal terrain.

(This includes Budget Amiga and Genesis levels. It does not include the PSP levels, as these were fan-remade by hand, so are not useful information here.)

Basically, the official games always bury it - of the above exceptions, one isn't really an exception as much as a "this could look like an exception if you don't look closely" case, and the other is a very unusual use case. Now of course - that doesn't necesserially outright serve as a reason to keep things as-is (after all - maybe they're like this *because* of the trigger area, although I'd have to wonder - at least for DOS / Amiga - why they didn't change the trigger area instead of designing around it, unless this was intentional), it's just a point to consider.

[IchoTolot]

The fire pit only has a deadly flame while the pit edges are not deadly.

I think you can be tricked both ways here:

- Why did the non burning edges kill me? (if we change it)

- Why can I walk under the pit? (if we don't change it)

As both ways have the chance to be misleading, I would vote for no change as people are more likely to have encountered this already in NL or the original game and therefore the to be tricked part is a bit less.

Furthermore, in the case of under the pit walking, I would consider it to be more expected by players as it is common knowledge that lemmings can squeeze through 1 pixel horizontal lines inside the terrain, therefore squeezing under the pit becomes more likely.

At last, I think in terms of fire objects I think a smaller more forgiving trigger area is better. Why?
I would assume players get angrier if their lemmings die unexpected and less angry if they survive, especially if surviving leads to an unexpected solving of the level. Of course an unexpected way can also be irritating and hide a solution, but then this also implies that there was really bad level desgin at work here and the level creater should make things 100% clear anway.

[Proxima]

- Why can I walk under the pit? (if we don't change it)

Furthermore, in the case of under the pit walking, I would consider it to be more expected by players as it is common knowledge that lemmings can squeeze through 1 pixel horizontal lines inside the terrain, therefore squeezing under the pit becomes more likely.

I don't agree with this. If the fire object were raised so there's a 1-pixel gap between the supporting bar and the terrain, then yes, players who have observed lemmings squeezing through 1-pixel gaps might expect to get through. But that argument doesn't apply when there's no gap. I would consider it very bad design if a level used a fire object perched on terrain, like the lower-right object in my picture, and expected the player to guess that it's safe for lemmings.

In fact, because of your argument against changing this object, I'm now much more convinced that we should change it

[IchoTolot]

I would consider it very bad design if a level used a fire object perched on terrain, like the lower-right object in my picture, and expected the player to guess that it's safe for lemmings.

Well, burning them is even less safe than walking through. I would be way more furios if I lose lemmings where I thought it would be safe.

Yes, it's bad design, but here I think the core problem is with the object itself and not the trigger. It's a fire hazard with a solid like frame. I don't think enlarging the trigger would solve the problem here, it would only be irritating in another way and if a change won't solve the problem then why change at all.

Proposal:

- Change the graphic to better represent the trigger-area. An animation on the frame can make it clear that it is indeed not solid. (Like glowing lighting effects for example)

[ccexplore]

If the solid-looking metal frame has a kind of glow animation (ie. oscillating slightly in lighting and hue), maybe that can help convey deadliness if we were to expand the trigger area accordingly?  I'm picturing a red-hot iron rod basically (but maybe a bit more subtle than that in the proposed animation).

[IchoTolot]

I still think conveying non-deadliness is the way to go, as surviving is less rage inducing than dieing if errors are made.

Also it would ensure no side-effects in terms of broken content are occuring what is likely with a trigger-area incease by 2 pixels in 3 directions.

[namida]

Perhaps we need to look at this from a different direction: Instead of expanding the trigger area, could we remove the sides from the graphic? (Without actually shrinking the object graphic's size; ie: it would just have a few pixels of empty space around it.) Or even move the sides inwards a bit?

If there's an agreement "we should keep the trigger area as is" but the graphic remains undecided, I'm okay with not making the graphic change just yet. As it would only have a visual impact, not a physics impact, on levels, it's less of a problem to change this later.

[IchoTolot]

Perhaps we need to look at this from a different direction: Instead of expanding the trigger area, could we remove the sides from the graphic? (Without actually shrinking the object graphic's size; ie: it would just have a few pixels of empty space around it.) Or even move the sides inwards a bit?

Well, the frame is quite often used for decoration. Just the flame by itself looks quite weird without anything around it. 

Moving the sides inwards and then apply ccexplore's proposal would also be a good enough solution in my book (Trigger stays the same and the frame is menacingly glowing).

EDIT:

After a bit of a discussion in discord we gathered the following:

- For a deadly frame would either have to change physics+animation or the size of the general visuals+animation which would alter the apperance of levels. So here we have either a visual or physics change for sure.
- For a non-deadly frame only a animation change would be needed. Like cold coal falling out of the sides of the pit over the frame. So no 100% need for a visual or physics change here.

In both cases we would need an animator for the fire and blue fire graphic though to make the case clearer.

[namida]

In both cases we would need an animator for the fire and blue fire graphic though to make the case clearer.

It should be easy enough to do some copy-pasting magic - or at worst, some code magic - to copy the new animation from one to the other.

Secondary animations might be useful here, by the way - this way the coals animation and the burning animation don't need the same number of frames. This could even be done as a secondary animation that's identically copied to both. (Yes, the common use case is to have a constant and triggered animations on the same object, or to show different graphics at different times; but the system most certianly supports - intentionally - being able to just have multiple standard animations like this, with different frame counts or to easily copy between objects, too!)

[namida]

I don't want to hold up 12.7 for this, so I'm declaring an ultimatum here: If we don't have consensus, or if there is consensus but not an animation to go with it, I'm going to make the calls myself as follows:

- Once 12.7.0 RC is released (planned for Friday evening / Saturday morning), the trigger area will no longer be up for discussion. If no solid agreement to change it has been reached, it will be permanently remaining as it currently is, and this will not be reconsidered again in the future, ever. If the agreement is "change it, but we need a new graphic to go with it" and no one is offering to make this new graphic, then I will treat this as "no agreement" (because it relies on something we don't have). In the case that I have to fall back to this, I will make a graphic that either removes the sides, or shrinks them inwards - whichever I feel looks better based on some testing.

- Once 12.7.0 stable is released (sometime in November), I'll be much less willing to change the graphic, although this one is not quite to the extent of "it will never happen no matter what".

The statement with regard to trigger areas will apply to all objects in official styles, not just the fire pit. These next couple of days are the last chance to propose any alterations to their trigger areas - from NL 12.7.0 RC onwards, they will be considered final, and only the graphics may change.

[Proxima]

I'm okay with leaving the object as-is. Icho makes a good point that it makes sense for all sides of the frame to be treated the same, so if the fire pit is raised 2 pixels (as in the lower-right of my screenshot, now on the previous page) then there is a gap between the flame itself and the terrain underneath, and it makes sense that lemmings can squeeze through there. I still have reservations, because I'm not sure how easily this would be understood by new players; but that's not something I can know for sure. In any case, the frame is often buried in terrain, so in most cases this distinction doesn't matter.

I'm definitely against removing the frame from the graphic, because this would damage existing content.

It's possible we could animate the frame so it looks less like terrain, but I don't think that's necessary or a good idea. Given how often this object is placed in a dip inside terrain (e.g. Fun 28/Mayhem 5), this would hide the animation or make it look strange. Also, this isn't a problem unique to this object -- most traps have parts that look solid but are not, e.g. the gun in the bubble tileset. New players just have to get used to the rule "part of the object is deadly = entire object is non-solid", and there's nothing we can do to change that.

[Simon]

The bottom-right fire, I still perceive this as deadly to lemmings walking on the top terrain bar. It even looks as if the bottom-right fire is the way to place this tile on even ground. Rarely does a one-piece trap sit so nicely and flat on terrain. (10-ton weight is 2 pieces.)

If the bottom-right trap were one pixel higher, I'd accept lems walking through unharmed.

(As a kid, I expected the sides to be solid. Really, all 3 sides of the frame are problematic design. Not a good argument here, thus, back to the matter.)

If we treat all 3 sides the same, then the frame should be deadly. Or do you want to put these firepits side-by-side, overlapping neighboring frames even, and have lems land unharmed in the 2-pixel-wide vertical frame piece?

Hmm, lems burning in the grey left/right sides is also not perfect. Still, if you stick these pits next to each other, it really looks like the TA should span all the horizontal space.

I don't buy the argument that all 3 frame parts are created equal in the first place. Lems touching the sides will have no overlap with flames. Lems walking underneath have the entire body covering the flames.

Nasty decision either way. Increasing the TA by 4 pixels horizontally and 2 pixels vertically is no easy decision. Keeping the frame peaceful is wrrrrrr. Or enlarge the TA downwards until bottom-right firepit (in Proxima's image) becomes deadly, at the expense of treating all 3 frames the same, and allow lems falling through the vertical bars unharmed. Nan nan nan, tough.

It's possible we could animate the frame so it looks less like terrain, but I don't think that's necessary or a good idea.

We can make the bottom bar look less solid. Maybe like a gridiron, or like feet. Then it wouldn't look like sitting squarely on the ground anymore. Hmm.

Whatever we do, the general problem is that the lem's pin is in the ground, but TAs often don't take that into account. If the one-piece trap sits squarely on the ground, it should be deadly.

-- Simon

[namida]

Here's some screenshots from in-game with the sides shrunken inwards.

In the case of Fun 18, I have screenshots with and without terrain modifications. The extra terrain that has been removed isn't necessary for the trigger area to be reachable (the objects were buried enough for that), but rather, seemed to only serve the function of filling in the transparent area.

Mayhem 2 looks a bit iffy. The rest of these look fine to me, as does a couple of others I checked but didn't feel added anything of value to the comparison.

I haven't compared on any custom levels yet. I've been holding off on preparing my official-style-using levels for V12.7 until this topic concluded so theoretically that would offer a chance to compare some - but I've used this object in exactly three levels - two of which are early/repeat versions of each other ("Falling Away From Everything" and "Die The Death Of The Damned" from LPI; with the third level being "The Troublesome Trio Go To Hell" from LPA). I'll include this graphic with the V12.7.0 RC release though, so it'll be possible to compare on any level - if we get a better animation, we can replace it for the actual stable release.

[Dullstar]

For what it's worth, I think the bottom right case where the object is completely resting on the ground should be deadly. Expanding the trigger area to make all of the bars deadly is then probably a good idea for consistency. I don't think the object's visual size should be altered, but I am not against animating them to make them look hotter.

[namida]

One basically-negligable other change I've made: The hanging lights in the Xmas style will now be horizontally resizable. This was mostly done because there's literally no reason not to do this.

[Dullstar]

If the solid-looking metal frame has a kind of glow animation (ie. oscillating slightly in lighting and hue), maybe that can help convey deadliness if we were to expand the trigger area accordingly?  I'm picturing a red-hot iron rod basically (but maybe a bit more subtle than that in the proposed animation).

If we want to do this, I made a mockup of the idea. I can definitely do better with the colors (specifically I think the areas where I used the yellow color on the right side need adjusted), but I figured for a mockup it's fine. What do you think of the effect?

My process to create it was drawing a red version of the bars as a new layer on top of this sprite, than adjusting the opacity of the layer on each frame. I do think it would likely look better as a secondary animation with a different number of frames. I can probably have a finalized version done by Thursday of next week. It likely won't take too long, but I will not be available to do it until sometime Tuesday or Wednesday night due to travel.

[Proxima]

With all respect to namida as creator and maintainer of NL, I think it's abundantly clear we need more time to decide on this. We are as far from a consensus as ever, and several different ideas are being developed.

[ccexplore]

I like Dullstar's modified fire pit, maybe a slight bit too subtle to convey the dangers of the frame, but at the same time the animation looks very natural and I think may be worth doing for pure visual reasons, even if we later decided against making the frame deadly.

[Dullstar]

With all respect to namida as creator and maintainer of NL, I think it's abundantly clear we need more time to decide on this. We are as far from a consensus as ever, and several different ideas are being developed.

I have to agree here. I don't think requiring a consensus before [deadline] or [default option] becomes permanent is a good way to go about this decision. While I agree it's best we try to decide what the object should do before we take action so we don't have to make several potentially content-damaging changes to the same object in a row, I don't think it's necessary to declare that "All modifications to existing objects we make must be done at the same time." I think it's okay if we defer a few trigger areas for later versions so we know that we have a clear consensus regarding whatever decision gets made.

Maybe it's because I'm not really good enough to solve the really difficult custom levels, but I can't think of any I've played that are so dependent on the exact details of the trigger area that a reasonably small expansion of said area would break them, so I expect the content damage from these physics changes will be very small.

[namida]

With all respect to namida as creator and maintainer of NL, I think it's abundantly clear we need more time to decide on this. We are as far from a consensus as ever, and several different ideas are being developed.

I do understand where you're coming from, but official style triggers have already been modified a few times, and there's always the risk of content breaking when it happens. That's why I want to set this as an absolute final time for it to happen - if something, by this point, hasn't got universal agreement, it probably never will and it comes down to having to make an "executive decision" so to speak. Likewise, if something hasn't been raised throughout all of these sets of changes, it's probably not important enough to worry about and thus it can stay as it currently is.

And while "change some objects now, some later" may not raise issues on the same levels, it will mean that the same packs must be updated multiple times, for each of these changes. While sometimes, pack updates will become necessary due to changes, we have a very clear option to avoid it here, by making all changes to the official styles at the same time. I think if anything - the sensible alternative would be "hold off on all changes for now", but I'd really rather get this matter over and done with, rather than holding it up for the sake of a single object that (maybe due to the issue not being that big in practice, maybe due to the object not being used all that frequently) only came up at the last minute. I would also at this stage have to go and seperate the changes out again, as I've at this point started preparing for V12.7.0 RC's release and as such these have been merged into the main NL branch now. Not yet to the extent that there couldn't be further changes to the trap's trigger area, but to the extent that I'd rather not have to roll back these changes as a whole to delay them for a future release.

So I stand by what I've said - V12.7.0 RC will be the last release that may change trigger areas of official objects (the only exception being if a change that was supposed to go in got excluded by accident), and ideally (though not "no exceptions") I'd like animations to be finalized for V12.7.0 stable.

[Dullstar]

It also seems to me that some of the reason for the lack of consensus is that there are essentially two issues at play: First, should the bottom pixel on this trap be deadly? If yes, then should this be done by changing the trigger area or the graphic? I suggest we should get a consensus on the first issue, then if we decide it should be deadly, only then should we move onto the second.

The bars seem to be creating some disagreement as to whether or not they should be deadly, which is why I will admit I don't think namida's proposed graphic change really solves the problem - to me, if I'm understanding it correctly, it looks like it shrinks the object around the trigger area, but it doesn't really resolve the conceptual disagreement surrounding whether this part of the object should, in fact, be deadly, which means the resulting object is no more clear, but it does have the potential to create situations where aesthetically guided object placements, such as multiple of the traps placed side-by-side, may break, which I suspect to be more common than situations relying on the pixel-precise details of the trigger area. Of course, if you've seen enough of my posts on these sorts of issues, you may have noticed I tend to oppose changes I believe will negatively impact the aesthetics of existing levels, valuing these sorts of levels over breaking precision levels.

Incidentally, I also think this object might be a good candidate to make resizable in the horizontal direction (depending on how exactly the graphics for resizable objects work, as I haven't worked much with them), although it may take some work to create a tiling animation for it.

[IchoTolot]

I was already pming with Willem to gather the animations (already got 2 but they have a different slightly widths as I forgot to mention that should stay constant - asked for an adaptation though) and we are at Step 1 again.

I agree with namida that this should soon come to a closure soon as having possible trigger and animation changes in the most used sets in the game every now and then is just a bad time to maintain. We all see things differently and there are bound to be things and objects that won't fit for every person. There needs to be an end point at some point.

I still stand with the following:

- For a deadly frame would either have to change physics+animation or the size of the general visuals+animation which would alter the apperance of levels. So here we have either a visual or physics change for sure.
- For a non-deadly frame only a animation change would be needed. Like cold coal falling out of the sides of the pit over the frame. So no 100% need for a visual or physics change here.

In both cases we would need an animator for the fire and blue fire graphic though to make the case clearer.

And therefore I am still for a non-deadly frame with an animation change.

I will post the new animations from Willem when I get them (I already asked for the adaptation before I read the recent posts to clarify).

[Simon]

Icho, your position on the firepit condenses to (deadly frame is a physics change or a graphics change, therefore it's worse than a pure graphics change).

You make no statement on what feels correct with the fire pit re misleading. For consistency, this trap should be treated with the same rules as other traps.

Please go over other people's opinions and attack some specific ideas. I'd be really interested in your counterarguments to replies #52, #53, and #57.

-- Simon

[namida]

Incidentally, I also think this object might be a good candidate to make resizable in the horizontal direction (depending on how exactly the graphics for resizable objects work, as I haven't worked much with them), although it may take some work to create a tiling animation for it.

Resizable objects are simply tiled. Since V12.6.0, it has been possible to use "nine-slicing" (I've also seen this referred to as "nine-pieceing", "3x3 resizing" - I don't know if there is a single standard name for it; but basically, there's a center region that's tiled, then the borders are only tiled in one direction, while the corners don't get tiled at all - see attachment). So basically, what would be needed is some portion of the object in the middle that tiles nicely.

I agree - if animation suits, this object should become horizontally resizable.

[IchoTolot]

You make no statement on what feels correct with the fire pit re misleading. For consistency, this trap should be treated with the same rules as other traps.

I have made the statement in past posts that I rather have my lemmings survive than die when something unexpected happens. I would be much more annoyed by a deadly frame in the current state. Also from the current state of the object I would assume it is non-deadly as there is no flame.
Therefore change the animation to make the frame look more harmless as it#s not part of the flame and don't make the frame deadly.

Icho, your position on the firepit condenses to (deadly frame is a physics change or a graphics change, therefore it's worse than a pure graphics change).

And that's my other point.

Please go over other people's opinions and attack some specific ideas. I'd be really interested in your counterarguments to replies #52, #53, and #57.

You just want more and more popcorn! I've got you, admit it.

No, but seriously the 2 points above is my viewpoint on the topic and I agree with namida to draw a line under the topic soon and don't have it drag on for eternity and dug up again and again.

[namida]

As V12.7.0-RC has arrived with no consensus reached, the trigger area for this object will not be changing - case closed.

V12.7.0-RC has shipped with the shrunken frame animation, as demonstrated above. However, I am still open to changing the appearance of the object.

We can look at the horizontal resizability later. As such a change would not in any way harm existing content, not even in theory, I don't have nearly the same level of objection to making this change at a later date.

[Simon]

trigger area for this object will not be changing
still open to changing the appearance of the object.

Ah, we're in for the true popcorn. x_X

-- Simon

[namida]

Here's an idea that came up thanks to something Crane mentioned on Discord:

How do people feel about shrinking the existing object, as is currently the case, then creating a new object that has the old graphic, and an enlarged trigger area to suit? Levels that primarily want the visual aspect would switch to this object; those where physics is the main concern would keep using the existing one.

To be clear - the one that would be used for all existing levels (unless modified) would have the original trigger area, with the shrunk graphic. The new, alternative option would be the original graphic, with a larger trigger area. Since this would be adding a new object, not changing an existing one, I don't feel it violates the "no further changes to trigger areas" decision.

[IchoTolot]

I think that would still cause a ton of confusion and checking and switching objects would also still be mandatory.
If we would go with 2 objects I think the new one should be the shrunken one, the visual breakage is much higher this time as the physical. I would rather take the trigger area change than the size change.

I still think a simple animation change is the way to go though.

[namida]

If we would go with 2 objects I think the new one should be the shrunken one, the visual breakage is much higher this time as the physical. I would rather take the trigger area change than the size change.

Are there many examples of levels where the visual breaks that are not misleading? As far as I can see - almost any existing level either has the object placed such that it looks to be deadly but isn't; or has it buried enough that at worst, the new graphic will leave the firepit sitting on top of terrain instead of buried in it.

The only exception that comes to mind is Mayhem 2, as mentioned before.

[IchoTolot]

From the original levels not that much. I am referring to custom content here. And I also count visual breaks as those where pits are flying in mid air now.

Some examples that would need visual fixes due to either pits not standing on ground anymore or creating unwanted gaps in between objects:

Reunion: 

5 12, 4 19, 4 14, 4 04, 3 19, 3 15, 3 09, 2 24, 2 19, 2 15,  1 24,  1 02 , 1 01

This is Reunion alone.

Now the possible list for checking Reunion where the trigger extend could cause a break:

5 12, 4 02

I can expand the list for more packs if you want.

So you see that the trigger fix is much less work in content terms than the shrinking. Therefore if we have 2 objects the shrunken object should be the new one.

[namida]

If a pit is not standing on ground anymore, preserving that pit is not important, because it was misleading in that what looked like it should be either solid or deadly was not. As such - not only do I disagree that this breakage is important to consider, I disagree that this is actually breaking them at all - rather, it is fixing them, so the player is clearly aware that there is a safe space to pass under them.

I do notice that in some of your examples, for example, 0101 from Reunion, accuracy to physics isn't particularly critical - one fire object is impossible to reach, and the other two are simply a case of whether you fall into them and burn, or fall between them and drop out of the level; the lemming is dead either way. This would be less severe than a level that places such a fire pit along the intended path, or somewhere that could be reached during a failed attempt or backroute.

However, ultimately, this is an object with an effect, not a purely decorative piece. Concerns about decorative use come second to concerns about a good visual-physics correspondence.

[IchoTolot]

However, ultimately, this is an object with an effect, not a purely decorative piece. Concerns about decorative use come second to concerns about a good visual-physics correspondence.

Yes, but it's still a lot more extra work. I still consider a level with visual flaws a flawed level that critically needs fixing. It doesn't matter to me if it's only decorational it's still flawed.

As such - not only do I disagree that this breakage is important to consider, I disagree that this is actually breaking them at all - rather, it is fixing them, so the player is clearly aware that there is a safe space to pass under them.

As the objects were unreachable this is anything but not a fix. The change did not affect the level gameplay (no fixing) and broke the visual aspect.

The workload is a critical thing to consider after all.

If the 2 object approach is chosen I would still consider to let the shrunken version be the new one superior as the possible workload is way smaller this way.

[IchoTolot]

Alright, here is a new graphic made by Willem that should make the outer frame appear less solid, it is a little change though.

The frame is now more rough with cold black coal edges on the left and right. In the middle we have grey/red glowing coal. These should appear way less solid.

I think in both cases (deadly/non-deadly) it should do a better job of appearing non-solid.

If alterations are desired those are totally possible!

[namida]

I have to say that personally, I'm not a huge fan of that animation as part of orig_fire (this last part is important - I do like the graphic itself, I just don't think it fits into orig_fire as a replacement for the firepit very well). I also would agree that this graphic solves the issue for the left and right sides, but if I were to encounter that graphic sitting on top of terrain, I would at no point even consider the possibility that lemmings can walk past underneath it - if anything, I think this new graphic is worse at communicating that. Thus, I don't feel that that animation in any way solves the issue at hand here (however, I do feel it's a good animation for use in a new, custom object, with an accurate trigger area).

Any graphic proposed needs to (a) very clearly fit the existing trigger area (if changing that was still on the table, then solution: change it and keep the animation exactly as-is), and (b) fit into the orig_fire style in terms of color selection etc. Beyond what's needed to achieve those, it should change as little as possible.

As the objects were unreachable this is anything but not a fix. The change did not affect the level gameplay (no fixing) and broke the visual aspect.

This level previously had a fire object that looked like it should be deadly, but could actually be walked under. It now has a fire object that's just as oddly placed as it was before (namely, placed such that if a lemming were to reach it, the lemming could safely walk past it), but now has visuals that reflect that. In this particular level, it's trivial to see that one fire object is unreachable and the other two aren't important to have accurate triggers (because if the lemmings miss the fire trigger, they fall out of the level anyway instead).

Let's look at some of the others. 0419, for example, every fire pit is reachable. 0414, which you mentioned as an example that breaks, actually does the exact opposite - the fire pit now sits perfectly on the ground instead of being slightly buried in it. 0319, the fire pit is definitely reachable and at least at a quick glance I would not immediately rule out the possibility of going near it; this change has made it clear that should you reach the area it's in, you can safely walk past it. 0315, most of the fire pits are accessible and clearly intended to be obstacles - and again, these don't really break as such, though one does look slightly awkward.

I didn't look at every example level you mentioned, and there are some where I do agree that ultimately it didn't provide any benefit - such as 0404 - but I think you are very much overestimating the caused problems vs solved them ratio of this change. Ultimately, I still feel just as much as I did before that, at worst, this change will trade a physics clarity problem for a visual oddity one, which is less severe by far.

[namida]

Here's a possible animation I came up with. This moves the fire pit up 2 pixels, and puts small support beams under it so existing levels don't have gaps. (Attached image can be dropped in to replace the existing PNG in either stable or V12.7.0-RC NeoLemmix; no NXMO change needed.)

I don't know how I feel about the trigger area re: the top of the object, though... Honestly, personally, I prefer the current (shrunken) one. I feel the sides are less critical; I'm okay with keeping the full width here.

Screenshots (click to show/hide)

From editor:

IchoTolot suggested on Discord combining this with Dullstar's glowing frame idea. I like the concept, I'm not sure how well I feel it fits the style (especially in regard to limited colors), but maybe it can be done. I do feel that, unless the object still gets shrunken inwards horizontally, only the bottom part of the frame should be glowing; this would clearly distinguish the deadly bottom from the non-deadly sides.

[IchoTolot]

IchoTolot suggested on Discord combining this with Dullstar's glowing frame idea. I like the concept, I'm not sure how well I feel it fits the style (especially in regard to limited colors), but maybe it can be done. I do feel that, unless the object still gets shrunken inwards horizontally, only the bottom part of the frame should be glowing; this would clearly distinguish the deadly bottom from the non-deadly sides.

That's correct. Maybe if compared to Dullstar's posted example the colors are slightly more red it fits the style better.

It also would add another option designwise: Burry it in the ground or let it stand on the support beams. So this would lead to more possible desgin choices.

[Dullstar]

For clarification, is the support beam example the same size as the original object horizontally? If so, I think it's a good solution.

I agree with this statement by Icho:

Yes, but it's still a lot more extra work. I still consider a level with visual flaws a flawed level that critically needs fixing. It doesn't matter to me if it's only decorational it's still flawed.

Resizing the object horizontally would definitely create visual flaws. The support beams would alter visuals, but I think the levels would still look fine and wouldn't need any critical fixes.

I also reiterate my suspicion that the visual breaks from multiple placed in rows is likely higher than the number that would be broken by the exact placement of the trigger area. If minor alterations break it, the level probably required too much precision to begin with, and I hate pixel-precise solutions.

[namida]

For clarification, is the support beam example the same size as the original object horizontally? If so, I think it's a good solution.

You can literally see how it looks a few posts up. It's two little vertical supports at each end, not a horizontal beam.

[namida]

Is there anyone who would prefer one of the previous proposals (or another specific graphic, not just "something else in general") over the one present in the latest V12.7.0-RC's? I'd like to wrap this topic up soon.

(Horizontal resize will be considered later. As that's a 100% backwards-compatible change, I see no issue with doing that at a later date.)

[namida]

Seems not, so let's close this now.