I'll be joining (in less than 3 hours from now).

I can definitely recommend it. It's a fun game, and even after many rounds, I still don't know what's the best strategy. You need to adapt to how the other players are playing.

I tried the simplified version (game ends when the pile is depleted, -1 for each remaining card) three weeks ago. They dynamics are definitely different near the end of the game, which is not necessarily a bad thing. What felt unsatisfying to me was that if you have something decent on your hand, you might not get another turn to play it (it's rather unpredictable when the game ends), so you either rip it apart to respond to some other player's sneak, or not play it at all. Counting 0 instead of -1 per card makes makes the latter option slightly less painful, but still not satisfying. But I'm curious how the scoring change affects the dynamics, maybe it simply becomes more predictable when the game ends.

Either way, this simplified version might be a good starting point for new players to get the hang of the game, and the final flaunting round can be introduced after one or two games. (Emphasizing "round": I'm not convinced by simply scoring the flaunts, as e.g. a flush and a straight flush get the same points then. The flaunting round is more strategic than it sounds and often brings interesting surprises as you have no clue what other players have left.)

While I'm not very visible here most of the time, I'm usually lurking and skimming the forum every day.

From now until mid-October, however, I'll be on the road and won't be checking the forum. The situation will probably be similar from mid-November to mid-December.

Thanks to everyone who's played my levels!

Everyone's got the intended solution for Misaligned (I mean there really aren't many options), and it was interesting to read your (different) opinions. It's meant to be a minimalistic level, and will probably appear as part of an ensemble of other levels that similarly focus on one single idea.

For Cycle Decomposition, everyone's got a different solution.
Armani's solution is slightly different from mine, but very much in the spirit of my idea for this level.
Icho's solution cuts out one main idea, but is at least still somewhat in the spirit of the level (I'd like to fix that though).
Kaywhyn's solution is, as you already suspected, an egregious backroute that needs to be fixed.

I want to fix the level at some point, but unfortunately, I'll be on the road for the next two months (and haven't had the time to fix it until now). So the update will probably only come once the contest is over

I think historically, one of the considerations for ditching direct drop in modern engines was that it had the tendency to cause backroutes in custom levels. There isn't really much of a reason for level designers to use direct drop as a feature, so with the backroute consideration it's more of a nuisance.

Regarding your table of where it works, isn't there at least some engine where direct drop only works if the exit trigger area is on terrain (so the lemmings go home instead of splatting, but they would fall through an exit that's in mid-air)? So I think it's more nuanced than "it either works or it doesn't".

Quote2) If falling lems can interact with traps, water, fire and other objects, then they should also be allowed to interact with exits. Any ports/engines/clones which feature one behaviour and not the other are therefore inconsistent.

Comparing permanent traps (like fire and water) to exits in this context feels like comparing apples and oranges. Many triggered trap on the other hand are quite similar.
Should triggered traps allow for direct drop? Arguably that depends on the design of the trap. For example, it makes sense for the ONML rock chameleon to catch lemmings in mid-air. It makes little sense for the 10-ton trap to squish lemmings in mid-air. Even without exits in the mix, you have to make a call which will feel weird for some traps (unless you want to introduce different trap types, which would be a whole new rabbit hole...).

I just saw the updated Pipe Dream in Simon's stream, I think this makes for a really good second level now!

More generally I think out early level progression isn't ideal. Historically we just had a bunch of introductory levels people made which I put in order.

I think in an ideal setting, we would map out a desired learning progression (what are the skills, etc, including reinforcement), and then have a sequence of levels that realize it. If we go by modern puzzle game design, this progression would probably cover at least a full rank, slowly introducing concepts (e.g. skills) and reinforcing them sufficiently along the way, keeping players hooked with interspersed simple puzzles combining the concepts they are already familiar with.

Thanks for playing! Notes on your solutions:

Day Care Center:

Spoiler

Intended, in particular the early jumper. The thing I find fidgety is having to bridge the miner tunnel with a platformer, especially while the miner is still at work and may destroy the bridge. It'd be nice to change that, but I'm too afraid to give a builder (or walker). I believe once you realize the goal (delay the second climber as much as possible), you can't overshoot (i.e. have the climber arrive too late) at least, right?

Undercut:

Spoiler

Close to intended. The 1 px ledge is actually only meant to break the deadly fall from the hatch, and you don't assign a miner there in my solution, so my solution is a bit harder probably but I'm not sure yet whether it bothers me enough to fix it.

Cellar Dwellers:

Spoiler

My intended solution is different, but it's needlessly complicated and harder to execute. I don't know if I want to enforce it or just downrank the level.

Construction Site Conundrum:

Spoiler

Funnily enough, I found this level in my WIP folder, but I believe I didn't have a recording for it, so I wasn't sure if it was solvable. The solution I found is the same as yours. I thought of moving it to the trashy folder, but I think it's unintutive but not hard to pull off, so maybe I keep it.

Btw, if you have suggestions for re-ranking levels, I wasn't confident in my original ordering so I'm happy to shuffle things around a bit.

I just caught up on the stream, most solutions are intended.

In Steady Supply, I'm less efficient at the very beginning (needing 3 bridges), that's why you have one left over. I don't see it as a problem though.

This Level Socks: It's meant to be a bit open-ended. This mirrors one of the solutions I found, with the spare skill.

Lab Map and Missing cubes exactly intended.

The Climber Predicament & The Floater Predicament: I don't actually remember what the intended solutions are, given they are contest levels from a long time ago. I re-solved them recently, but I need to dig deeper to see if I can find some older replays. Either way, it shouldn't be possible to solve both maps the same way, so I'll need to take a closer look at some point.

If you look at Lemmings 2 DOS version, you'll see there's a levels folder with one file for each of the 120 levels plus the 4 practice levels. After decompressing these files, you can read the level layout of each (as documented in the file format description).

The 16x8 tiles in the image dump are in the correct order. So a 1 in the level file would mean the first 16x8 tile, a 2 is the second 16x8 tile, etc, so it should be straightforward. You just need to create the NL tilesets based on those 16x8 tiles (which I assume can also be automated). Then you're sorted for the terrain at least.

Regarding your other question, of course you could also create images of the levels from the level files and the 16x8 tiles, it's a very similar process to the above.

The L2 level/style format has been long known (see here - note that you'll have to decompress the levels/styles first using lem2zip, which you can also find there). The challenge will be that the levels are defined as a grid of 16x8 tiles, while (from what I understand) the existing NL L2 tilesets use pieces that are usually composed of multiple of such tiles. The L2 style definitions do define such pieces in terms of the smaller 16x8 tiles, however, I don't know if the NL L2 tilesets were created in a systematic way reflecting these definitions, or just ad-hoc.

All the L2 graphics (both 16x8 tiles and the composed pieces) have been ripped long time ago: https://www.lemmingsforums.net/index.php?topic=6588.msg102252#new

A slightly ugly way would be to write a script to create 12 new tilesets where all the tiles are of size 16x8 (you could use the ripped graphics as input), and then write another script that converts L2 levels into NL levels using these 12 tilesets, automatically creating the terrain and object placements (assuming the existing NL tilesets are consistent with the L2 ordering of objects, otherwise you either have to manually create a mapping between the two, or extend the first script to also generate the NL objects -- I don't know how NL would deal with fling objects and the chain and cannon though). The skillset you'll have to tweak manually either way as I guess you don't support the full palette of L2 skills.

Thanks for playing so far! A few comments below:

Spoiler

Underground Club and Forever Mine are indeed a bit tricky to execute. A previous version of Forever Mine really frustrated Simon so it's already been improved quite a lot, but I think some of the timing difficulties are unavoidable. I was contemplating putting it in the sketchy rank for that reason, but in the end decided against that. Underground Club has a setup of the cubes which is a bit easier than what you did.
Please Come Home: I don't really understand why this is finicky, especially looking at your solution? Or do you mean you were pursing some red herrings that were precise? There's only the platformer placement where you only have a few pixels of leeway, and it's hard to gauge exactly how high it needs to be, but if you try 2 or 3 time and see where the basher ends up you should be fine. There's also a subtle placement hint of two 1-pixel lines on the rock but maybe it could stand out a bit more.
One is still not enough: I noticed this one is not on your list of solved levels, I'm wondering if you haven't seen the idea yet, or if I should give a tiny bit more leeway.

It's finally time for me to release the Lix levels that I have created over the years. Some of them are brand new, some of them are over a decade old, but better late than never. They are part of the latest Lix download in the level folder single/geoo!

The pack has 5 ranks, which are more like themes. Each rank is independent, and you can play the ranks in any order and freely skip between them. Within a rank, I recommend playing the levels in order.

• wrappy: Levels that make use of the horizontal and/or vertical wrapping. It turns out that even in single player, wrapping is more than a gimmick and gives rise to a lot of interesting ideas! If you only ever play one rank from this pack, this should be the one. The majority of these use 5 skills or less, but even a 3 skill level may stump you for a little bit here. These turn out to be primarily logic puzzles with very easy execution, except for maybe the last few.
• flowy: These levels make use of dynamic crowd control: stalling, redirecting, merging crowds without fully blocking them off. These levels require thinking about when things happen as much as where, and a lot of skill placements will be unintuitive, initially. Due to the nature of this, some levels may require a bit of trial and error to find the placements, but I tried to make execution as easy as possible, and the replay tweaker should help as well. Note: 4 of these levels are also featured in the lemforum pack, but I replicated them here for the purpose of the level progression.
• classy: More traditional and often larger levels with more skills where you need to figure out the skill placements.
• loony: Many of these revolve around some physics quirks or some other quirky idea. A good portion of these keep it very simple, though some of the later ones may end up taking some work to get the execution right.
• sketchy: This is where you find the levels that didn't really work out. Many of these may have some interesting underlying ideas, but somehow these ideas didn't translate into good levels. Feel free to skip this one.

The ranks are not sorted by difficulty, but rather by how interesting and fun I perceive the levels to be. Within wrappy, flowy and loony, the levels are ordered to create a level progression that gradually introduces and explores different ideas. This only loosely corresponds to actual difficulty of the level, but each of these ranks starts out gently. Classy and sketchy are sorted by how I estimate their difficulty, which may be different from how difficult.


I may add and shuffle levels in the future, so posterity, please refer to levels by their name rather than rank/number, as the number within a rank is likely to change in the future.

Thanks to Simon for playtesting, and Ramon for playing some of these during the races with Simon! The levels are part of the official lix release as of version v0.10.21 (for about a month already, so about time I make this topic), and can be found in single/geoo.

I'd be very happy to hear everyone's feedback and see your streams, videos or replays!


Just Mine (wrappy): This map wraps horizontally. Why do you have three miners here when it looks like two are enough? Play this level to find out and wish you had four miners!


Humpy Bump (wrappy): This map wraps horizontally. How can you safely build a bridge when lix are walking around everywhere?


Pommes Schranke (flowy): How do you fit your sausage (that comes with fries, ketchup and mayo) on a plate that's too small for it?


Roots of Life (flowy): Even though this level has only 5 skills, it stumped Ramon and Simon for almost an hour!


Bathroom Break (classy): Hurry before everyone else arrives!


Filter, Map, Fold (loony): The typical (functional) pipeline of reducing data to a single (summary) value. Can you do the same to this crowd of lix?

I think one major part in community history that you haven't mentioned yet is the challenges:

People put a lot of work into trying to save the maximum amount of lemmings or use the minimum amount of skills, sometimes exploiting crazy glitches in the process (like sliding or lingering blocker fields/moved trigger areas). If the discussion touches upon L2, then of course there's the surprising result that you can save every single lemming, the last find being the crazy pausing glitch abuse by ccexplore to save 100% on Sport 1.

I remember having that issue back in the day when I first played Cheapo (which may have even introduced this behavior? Or was it in WinLemm too?) It confused the heck out of me, and it took me a while to understand what was going on. I thought it was a bug and restarted the game in order to restart levels.
At the same time, it's way too useful of a feature to disable by default (thereby making it extremely hard to discover).

I'm surprised that clicking the R does not cancel the replay, having that would be a good start at least (especially if the R sufficiently stands out visually). I think realistically what Lix is doing by displaying some text is a decent solution. It's slightly inelegant, but at least should leave no room for confusion.