Quote from: JawaJuice on November 26, 2025, 11:36:02 PMAs a level designer, do you have any view on alternative solutions you hadn't considered? Or is everything that isn't very close to the intended solution a backroute?

I had another crack at R1, probably another backroute since I didn't use the basher. Gotta say, Crane, yours and Icho's levels are devilishly hard! XD

Backrouted again.  V3 is out.

If a solution is unique, beautiful or more difficult than what's intended, then it might be kept as a viable alternative.  It's very subjective though.

One recent example with my own levels... the original "Ice Cream Cake" was just one level with the "part two" solution as the intended one, but it had 2 bombers that were meant to be a red herring, but Armani playtested it for me and found a solution that was just too beautiful to pass up.  Coincidentally, this was at the same time that the previous contest was being run where you had to make a pair of levels with different skillsets, where one wasn't a subset of the other (so they had completely different solutions and one couldn't be performed in the other), so I withdrew my submission (which was the "Outside the box" levels) and submitted two versions of Ice Cream Cake... one with a bomber removed (to stop Armani's solution) and one with a climber removed (which wasn't required for Armani's solution)!

Sometimes you do get happy accidents.

@JawaJuice
Close enough well done!

@WillLem

QuoteJust because something is "long-standing", and a "standard", doesn't mean that (a) it's the right way to do things, (b) it's the best way to do things, or (c) it shouldn't be subject to scrutiny, review, and potential updating.

In other words, "it's always been that way" shouldn't be a strong argument not to try something different.

What I mean by a long-standing community standard is simply that most people have just always done it that way when uploading level packs. They've always had the option to include or not include replays, but most chose not to. I've given my own reasons, but I assume everyone else has their own reasons as well. My reluctance toward your idea can be summed up in one sentence: "I worry that it could create an undeserved disadvantage for the majority of authors who choose not to include replays." Everything else I wrote was simply a response to your response.

QuoteSurely if we can trust users to make solvable levels, we can trust them not to look at solution replays for spoilers! !

Including replays or not has nothing to do with whether we trust players or not. I enjoy seeing how far players can get without having the answers available from the start, but I also understand that some people don't want to put too much time into a single level and prefer to see the solution and move on. So another main purpose of including replays is to allow players to compare their own solutions with my intended solution without having to upload their replays, wait for me to check them, and then wait for my response. By releasing the replays after some time has passed, I can filter out any obvious backroutes or major design oversights before a self-sufficient "ecosystem" forms where players can check the intended solution on their own without my involvement. It has nothing to do with whether I trust users or not, because my intention from the start is to let players freely compare their solutions with mine whenever they choose. Conversely, during the early stages after releasing a level pack, I actually want players to report any fatal issues so I can fix them. That's another reason why I don't publish the solution immediately.



@namida

QuoteLevel-as-directory in general seems like an interesting idea.

This made me wonder : what if, in a completely new hypothetical engine, we created a dedicated website where users could voluntarily upload all sorts of assets related to each level, including styles, solutions, music, previous versions of the levels and so on so that players could download as much or as little as they want.
In a sense, it would be like an expanded version of NeoLemmix's style downloader extended to other categories. By  keeping a core NXLV format and allowing players to download the level, the required styles, and optionally the replay directly from the forum thread, the traditional method would still remain available.   My workplace actually has a similar internal network: all the artists upload their characters, 3D models, sprites, and all sorts of components to the internal repository, and each department can pull whatever they need whenever they need it. The idea I'm suggesting would work somewhat like that.

@JawaJuice

Your solutions to my R1 and R3 are intended! Great job!

For R1, there's actually two intended ways, since

Spoiler

the basher and miner are interchangeable in the solution It's true that the miner skill is a downwards moving destructive skill, but the basher skill's destruction mask starts at the row of the Lemming's feet and removes several rows of pixels above it too If you do it this way, you need to make sure to mine through the dart in the same way so that he is not delayed, i.e, you need to be right up against the dart when you assign the miner. Then the basher assignment to release the ohnoer is at nearly the same spot as the miner, just an extra pixel forward but the effect in the end is the same: The ohnoer touches the exit trigger on the ground in time before he explodes  

Quote from: JawaJuice on November 26, 2025, 07:56:30 PM

Spoiler

You like your destructive skill extensions via builders don't you?

Heh, I like how you've played levels of mine that happen to require this and you are quick to associate this solution element with me, when in fact only one other level of mine (in one of the previous LDCs, not this one) requires this Yes, I do love this combo in general, but I've only used this as part of my intended solution just last year, 3 years after I first started level designing! This is also only one type and way of doing so. I haven't yet made other levels that involve this with the other skill types!

QuoteIt's of course possible that I might stumble upon a trick I wasn't previously aware of just by ditching prior assumptions and trying many things, but perhaps not probable. Where do you stand on topics that detail advanced tricks? I read through Strato Incendus's Advanced Blocker Tricks topic (nothing on there I didn't already know, but interesting nonetheless, probably helpful for some) and the discussion that ensued makes for interesting reading. Opinions seemed mixed on whether such a topic was a good idea or tantamount to a cheat sheet. After a quick search, that's the only topic documenting tricks that I could find.

Even though I am one of those Lemmings level solvers who doesn't like hints or spoilers of any kind, I personally don't mind such a thing existing. I think they're good to have, if only to serve as reminders for those who are already aware of them, since it can be kind of easy to overlook or forget them from time to time when solving levels. I myself have the tendency to overlook the simple parts of the solution while getting the complicated solution parts quickly If they're hidden behind spoilers like they are in that topic of Strato's, then those who prefer to avoid them can simply not open up what's behind the spoiler tags, while those who want to know can if they like.   

Quote from: Simon on November 26, 2025, 09:40:50 PMA level might be a directory. It should contain the level proper, the intended/accepted solutions that should pass during mass replay verification, and backroute replays that should fail.

I'm not sure how useful the backroute replays would be in a finished level (they could be very useful for a WIP level where the backroute in question has not yet been fixed, and it's there to remind the author what they need to fix), especially without some kind of "auto replay repair" mechanism. Otherwise, the fail could simply be that a lemming is not quite in the right place due to some minor unrelated change, and that if the assignment was tweaked back to the correct position, the replay would pass again. It could perhaps be useful as a collection for a human to verify (but then, they'd need a way to see the backroute work on a vulnerable version of the level, otherwise they might not know exactly what they're trying to fix), a lot of effort though.

Level-as-directory in general seems like an interesting idea. Another advantage would be that "vgaspec" levels could have their unique terrain pieces as part of the level, rather than needing an (author)_special style. Disadvantage would be a bit more hassle when creating / organising packs by hand, as well as the need to ZIP them when sharing - although this could be avoided by still having a core NXLV or similar file that can also be used on its own, assuming no level-specific pieces like the vgaspec idea are needed for the level in question - perhaps by the engine having an "import" folder you drop the level into, and it moves it to an appropriate place in the tree structure with a directory of its own. Another possibility could be to work similar to how styles associate extra files with pieces - the main level file is titled, say "Example.nlxv", and then any files associated with it are "Example_Terrain.png", "Example_Replay1.nxrp", etc; keeping each level in its own folder then becomes a matter of convention (and engine support for a folder that contains a single level, rather than being a rank / pack) rather than obligation, and the NXLV remains useful on its own if no other files are needed. Downside to that would be potential issues arising if one level's filename starts with the same thing as another's full name, eg, if you have "Example.nxlv" and "Example_(Part_II).nxlv" in the same folder, whether the engine might think that "Example_(Part_II)_Solution1.nxrp" is a replay called "(Part_II)_Solution" for "Example". Far, far too late for NL / NLCE at this point (SLX could consider it), but that's a very good idea for any future engines / overhauls.

@IchoTolot

I think we're good with R1 this time

Spoiler

Studying the terrain more carefully, the unusual sloping underside of the first block is a clue - about the same angle as a builder bridge, so suggests the top part of basher rails Neat level!

EDIT: And R2? It meets the criterion of one lemming to each exit, at least!
EDIT: And R3! That was a tough one!

As a level designer, do you have any view on alternative solutions you hadn't considered? Or is everything that isn't very close to the intended solution a backroute?

I had another crack at R1, probably another backroute since I didn't use the basher. Gotta say, Crane, yours and Icho's levels are devilishly hard! XD

As a level designer, backroutes are to be expected.  Obviously we hope we patched them properly, but backroutes happen, and we don't hold it against anyone if they find them.

The third rank of this pack caught me a bit by surprise. Almost all of the talismans were very easy to obtain, despite being in gold and silver. There was one noteable exception - and in this particular level I could not make out any strategy, so I skipped the talisman in the end. Generally spoken, this rank is weaker than the previous two in my opinion. The first and the last level were the ones I enjoyed most.

Spoiler

Treacherous

1 - Rift walkers

Okay level to start off the final rank. The hardest thing was to bring the pioneer back where he came from. Without bashers, I had to dig and interrupt - I think that's intended here.

All my solutions can be found here: https://www.youtube.com/@Swerdis

During play, the level tree is immutable and sacred. Don't modify it.

Putting the intended solution with the level ... Reminds me of my old idea for Lix that I've never realized: A level might be a directory. It should contain the level proper, the intended/accepted solutions that should pass during mass replay verification, and backroute replays that should fail. This would be excellent documentation for the pack maintainer.

For NL/CE, intended-solution-next-to-level feels like too big a break in formats. After all, the established tree structures (levels, replays, ...) are formats, too.

-- Simon