[OLD] Player Bugs / Suggestions

[GigaLem]

im having problems with replays when every ichotolot sends me one one for a neolemmix level i gives me a "doesn't appear to be a neolemmix replay 0020" error

[namida]

From the V1.35n changelog:

NOTE: V1.35n replays will not work on previous versions of NeoLemmix (previous version replays
      will work on V1.35n).

If your levels were made / edited with a version of the editor older than V1.35n-A, you might get a warning when trying to load the replay (saying it's for the wrong level), but it'll still load, and if it is actually for the right level, it will work fine.

[Crane]

I'm not sure how this is best addressed, but I don't like how you can dig into down arrows or one-way arrows that are against the lemming's current direction, and then bash to stop him (or mine if against one-way arrows).  In the original Lemmings, the action would fail completely because he was inside the one-way arrows object.  This is probably more of a feature than a bug, but I find it a little irritating from a design perspective because it means the player can stop the digger without a blocker, builder or a steel plate and make a pit to trap lemmings.

Personally, I'd check the status of the pixel that is in front of the lemming... what would be the wall of the digger's trench, one above the floor (so it doesn't interfere with mining into one-way arrows to turn a lemming around, or bashing through a regular wall that sits on top of one-way arrows).  If it is against the lemming's direction, then fail (don't do anything except maybe play the "chink" sound) if the player tries to assign a basher or miner.  This also has the nice effect of making it clear to a beginner that you can't bash against the arrows on, say, Fun 11 ("Keep your hair on, Mr. Lemming").

[Wafflem]

Three new glitches:

1. It seems that if a terrain is close to the top of the level and there is about a 1-pixel gap between the terrain and the top of the level, and a lemming climbs up to the top, instead of the lemming walking through, they turn around and fall again. I've attached a level; the trapdoor is pre-assigned climbers.

2. There seems to be another object order issue. If a lemming falls into a terrain that has radiation and splat pad at the same time, the lemming doesn't splat, but instead receives the radiation. I've also attached a level.

3. Minor graphical glitch, but it seems that when a lemming falls down and splats, they move left by one pixel. I've attached an image for this one.

[namida]

1 & 3: I'll look into these.

2: Without using ridiculous amounts of either memory or processing time (especially on larger levels), there's a limit to how much this kind of allowance for overlapping objects can happen. The priority isn't exactly "radiation has priority over splatpad" as such; it's just whichever object is last in the object order has priority. To fix this with anything resembling the current way of working would need as much memory as 64KB per pixel, or around 3GB for just a single screen. I have one idea of how it could be improved somewhat, although this would still only allow overlapping of object types that the individual object does not need to be tracked (only the type does) plus one object per pixel where it does, so for example, you could overlap a triggered trap, a fire object and radiation, but not a triggered trap and a pickup skill (as both of these need to track the individual object).

Of course, eventually, the ideal approach would be to allow for overlapping of any number of objects, but I'm not entirely sure how best to achieve this.


EDIT: Re: point 1, in terms of that they fall down and turn around, this is intended (it will even happen *two* pixels from the edge, and will also happen with walkers). However, there are other issues here that should be fixed, see attached level.

[Wafflem]

If a lemming falls a high distance into a teleporter, and the receiver's trigger area is over the top of the level, when the lemming exits the receiver, they splat. Attached is a level to illustrate this.

[namida]

I'm not sure what you think should happen there? The lemming ends up outside the level, so perhaps there's an argument for removing it entirely (without a splat), but this isn't so much a glitch, as a level design set up to allow the lemming to get stuck in the infinite wall outside the level. The same would happen if you placed the receiver inside the level boundaries, but embedded in terrain; or for that matter if with a more typical teleporter setup, the lemming fell that kind of distance into a teleporter then hit ground under the receiver.

[Wafflem]

It just seems weird for a lemming to splat over the top level boundary, but I agree with your points for the other reasons lemmings splat upon exiting the receiver (splatting in receiver embedded interrain, etc.).

Found another glitch: bomber particles appear on the upper left of the screen instead of the lemming bombing itself. This occurs when the lemming bombs, then I press "R" to replay, then press Space Bar a couple of time to get to the point where the lemming exploded.

Also:

> Lemmings move left one pixel when splatting (possibly only when facing right?)

I can also confirm that they move left one pixel when they splat while facing left.

[namida]

Okay. Before I put out the V1.35n-C update, I'd like to encourage people to report any bugs they've noticed - no matter how insignificant or how unlikely I am to consider them important - because ideally, I'd like to make this the absolute last update before working on V2.00n. This only applies to the player; no need to worry so much about the editor at this stage, unless the issue lies in player-editor integration. Anything relating to the Flexi Toolkit, let me know just in case - though it may be a toolkit issue, not a player issue, I'm probably more able to determine where the issue lies than the average user would be and either way it'll be helpful to know.

For the same reasons, I'll probably get some people who've been particularly good at spotting bugs to test V1.35n-C before I actually release it, even though I normally don't do that with NeoLemmix updates.

Basically, I only want to do a V1.35n-D update if there's either very serious glitches, or glitches newly introduced by the V1.35n-C update. Otherwise, V1.35n-C will be the final release of the V1.xxn branch.

[mobius]

I have player version 1.33-nB.

apoligies if this was covered, I did a quick search and couldn't find anything. The miner safe/deadly fall glitch still seems to be present. I attach a level to

I will PM you a level and replay with a demonstration. [I don't want to post here for spoileryies!!
Glad I thought of that in time.

[namida]

This actually seems to be the inverse of the L1 glitch; miners are surviving a fall that other lemmings cannot (in L1, sometimes miners would not survive a fall that other lemmings would). I'm going to guess that the fall in question in your replay is exactly 1 pixel too far for the normal lemmings (after testing: yes, this is indeed the case), which may mean the miner moving downwards one pixel during its normal movement isn't counted towards the fall distance. That's not to say that should happen; just that it's what is happening. I'll be sure to more closely investigate and fix that.

[Wafflem]

Seems that if I type in the cheat code to enable Cheat Mode (i.e. press "1" to skip levels, unlock all levels), but then when I exit the player and open it again, Cheat Mode is turned off automatically, and I have to re-enter the cheat code to re-enable Cheat Mode.

This happens in Flexi Players such as GigaLems, Lemmings Reunion and the Cheapo Copycat Lemmings players, but does not from players that are built from source code (e.g. Lemmings Plus series).

[namida]

I've noticed this myself. Previously I'd only seen this in other packs, and concluded the designers must have chosen the options that result in this; but I noticed it also happens in the NeoLemmix Introduction Pack (which is Flexi-based, not made-from-source).

[mobius]

I've been checking secret triggers and noticed something;

if I tell a secret trigger to point to a level in a rank of 6 or above it gives a "list index out-of-bounds" message and crash.

[namida]

^ Just to verify, you do have a rank 6 in your pack, right? (In particular, it's not a NeoCustLemmix pack?) If this is the case, I'll need to look into it - LP3B has no issues with secret triggers pointing to rank 6, but this could be a problem exclusive to Flexi.