"Lemmings Redux": A "remake" of official games for NeoLemmix.

[namida]

I think it would be best to work on any needed adjustments to levels before we put too much focus on their order. We should decide the order based on how the levels turn out post-adjustment. In some cases this may not make a huge difference, in others it will probably make a lot.

Levels that may need anti-backroute adjustments might be a more controversial matter (since we don't know for sure what is the actual intended solution), so I think we should first start with simply ensuring the levels conform to NeoLemmix conventions. This would mean:
1. Making hidden traps visible (and other similar fixes to misleading designs)
2. Reducing the number of lemmings where suitable
3. Removing time limits on levels that don't need them

#1 will be the most straightforward. There should be very few cases that need discussion, and it will be more a matter of simply going through and making the changes.

#2 will be a bit trickier, though should still be straightforward. The way I see it, the ideal approach would be to simply use some kind of formula or conversion table, which could be applied by a bit of quickly-written code to the entire collection of levels. The levels where this should not happen, I believe, would be in the minority, so these could be reversed manually. (Off the top of my head, the only example I can think of is Cascade, though I'm sure there's more.) Two ideas that have come up are "Halve the count on any levels with more than 30 lemmings", or to use a system like "Leave <=30 as is, 31-50 change to 30, 51-80 change to 40, 81-100 change to 50". Of course, any such formula can be overridden for specific levels - for example, "With A Twist Of Lemming Please" we probably want to consider something even lower than this. Likewise, we might also want to change the NES-origin levels with 14 lemmings to instead have 10 or 20, or maybe even 15 - the reason for 14 there is not because 14 is the optimal amount, but because 14 is the maximum the NES version supports.

#3 may be a bit trickier. Some levels should clearly keep their time limits, for example, "Just A Minute". Some very clearly should not, such as "Save Me". But there are cases that may be a bit more ambiguous, such as "It's Hero Time" - sure, the time limit adds to the challenge, but take it away and you've still got a good, challenging level. Do also keep in mind that we have the option of simply changing the time limit, too - for levels that do need one, but the existing one is a bit too tight. One major advantage here is that whereas the official games only allowed whole minutes, NeoLemmix time limits can be adjusted by seconds.

[Proxima]

Sure, that makes sense -- although those adjustments don't really affect difficulty, except for backroute fixing; but it's still good to focus on one job at a time.

#1 I'm fine with that. I don't think there are any hidden trap levels left in, other than those mentioned in bsmith's post here. The NXP in that post contains possible fixes.

#2 I can think of a few more levels that should keep their lemming count: "Go for it!", "Just a Minute" (both parts), "No time for a detour". Levels with a splat trapdoor should be treated separately instead of by formula: "POOR WEE CREATURES" and "No hurry, Relax."

#3 bsmith's post here has a preliminary list of levels that should keep their time limits.

For #2 and #3, I'll have a look through the pack myself to see if I can find any others.

There may be other minor adjustments to make -- I mentioned enlarging "Fall and no life", and we also need to decide which version to use for each of the levels that appears in both Genesis and ONML.

[namida]

#2 I can think of a few more levels that should keep their lemming count: "Go for it!", "Just a Minute" (both parts), "No time for a detour".

Agree on all four of these. I'd most likely approach this by doing the automated parts first, then looking at each level one at a time during which I'd see these things.

and we also need to decide which version to use for each of the levels that appears in both Genesis and ONML.

I'd say ONML versions in all cases. There'll be less use of the ONML tilesets.

But on the subject of tilesets, one thing that does come to mind - the Xmas levels, should we perhaps convert them to Snow levels? I'm not sure if it's kind of weird to be having the Santa lemmings, and two slight variations of what's otherwise the same graphic set? (This would be as simple as manually replacing the objects, while the terrain can be left as-is.)

Re: bsmith's time limit list, here's the list. Anything I haven't commented on, I agree with.

2-27 Just A Quicky
2-28 Speed Trap
3-15 No Time for a Detour
3-21 Four Play
4-8 Five Alive - Would have to see how this level plays when debackrouted before I could comment.
4-13 Synchronized Lemmings
4-14 Just A Minute
4-17 The Silence of the Lemmings - Disagree. My first attempt at solving this level just now (with zero memory of the solution I've used in the past) had 1:30 left on the clock. I can't see how the time limit really plays a part in this level.
4-24 Onward and Upward - Unsure. I had 30 seconds left on this one, but my solution did use a somewhat obscure trick. Anyone else have any thoughts on this one?
5-1 Its Hero Time - Unsure here. I believe there are a few solutions that would work if it weren't for the time limit, but perhaps it's better to backroute-fix those in other ways and remove the time limit?
5-6 Lemming About Town - Disagree. The time limit may be tight, but it doesn't really serve a purpose beyond annoyance.
5-12 Just A Minute (Part Two)
5-18 Scaling the Heights - This is one where I'd be in favor of loosening it up a little bit, maybe 15 seconds. But yes, overall it should stay.
5-24 Hello John Got A New Lemming
5-26 Oh No! Squish

[namida]

Regarding trap unfairness fix for "Worra Lorra Old Blocks"... do the traps actually need to be there? The only effective way to deal with them is to build up to the steel, so is there any reason that the steel block as a whole - as well as the traps it's hiding - couldn't simply be removed, and the builder count reduced by one?

Found another level that I think should keep the time limit - 4-18 "The Race Against Cliches".

[namida]

Here's a first draft NXP with lemming counts reduced, time limits removed and traps made visible (may have missed some, let me know if I have).

Lemming counts were determined by the 31~50 -> 30, 51~80 -> 40, 81 ~ 100 -> 50 rule, with manual adjustments on certain levels. In the two NES levels that had 14, I changed one of them to 10 and the other to 20. The decision for 10 in the case of "Hello John! Got A New Lemming?" was because, similarly to "With A Twist Of Lemming Please", it has mass-assigning floaters at the start.

In most cases, the save requirement was determined by the following rules:
1. If the original save requirement is 20% or less of the original lemming count, leave it as-is. (Eg: A level that required 10 out of 100 would now require 10 out of 50.)
2. If the original save requirement is between 21% and 80% of the original lemming count, keep it at the same percentage of the lemming count. (Eg: A level that required 50 out of 100 would now require 25 out of 50.)
3. If the original save requirement is over 80% of the original lemming count, keep the number of lemmings that may be lost the same. (Eg: A level that required 97 out of 100 would now require 47 out of 50.)
This worked very well for most levels, though a few cases (such as "Poor Wee Creatures") needed manual changing. In cases where it didn't seem that it would significantly affect the level, I've also adjusted them to nicer numbers in some cases (eg, a lot that came out 24 of 30 (from an original 40 of 50) I adjusted to either 20 or 25 required.

Time limits were kept for most of the levels on bsmith's list, as well as "The Race Against Cliches". In the case of "It's Hero Time", I made an adjustment to the layout that prevents the only solution I know of that fails solely due to time and cannot be tweaked to meet the time limit. I know of a few that are very borderline on time (to the point where tweaking them a bit can make them pass), so I decided to go with no time limit here (we might want to consider similar changes in other cases like this - the key factor being that some solutions can go either way depending on how optimized they are). In the case of "Scaling the Heights", I added an extra 20 seconds.

I've also switched all the Xmas levels to Snow levels. I kept this in a different commit than the other fixes, so it can easily be changed back if we decide that way.

The order still remains unchanged.

Let me know if I've missed any hidden traps, or if there's lemming counts / save requirements / time limits that need further adjusting (as well as more input on the disputed cases for time limits).

[Proxima]

Well, I think we need to slow down again. Having got a final level selection, after taking so long, is great and feels like we've really gotten things moving, but we still need to discuss changes before implementing them. In this case, you've barely allowed any time for people to chime in on whether we should reduce lemming counts and remove time limits at all, never mind discuss which levels.

Right now, I'm looking over the previous NXP to compile statistics on lemming counts and tileset usage, so that I can make a post with informed views. This will take some time.

In the meantime, here are the statistics for lemming counts in Rank 1 only:

1 - 1 level
10 - 4 levels
14 - 1 level
20 - 1 level
40 - 1 level
50 - 12 levels
60 - 2 levels (including LEMTRIS, which we've considered up-ranking)
80 - 3 levels
100 - 5 levels

It follows that 50 can be seen as a kind of "default" value that's the most commonly chosen when there's no reason to choose any particular value. (It remains to be seen whether this will persist over the whole pack.) So reducing the lemming counts by formula will result in another value, most likely 25 or 30, becoming the most common and being seen as the default. That's something we should bear in mind.

[Proxima]

Okay, my thoughts.

Firstly, I'm not sure about the whole idea of "NeoLemmix conventions", because it's never really been made clear what these are -- namida's preferences as NL's owner; an overview of what's actually produced by the community; or community-agreed guidelines for new content? If it's the last, then there ought to be a discussion (separately from the Redux pack) of whether the community actually does agree on some of the guidelines, including reduced lemming counts. So far, all I know is that namida and Nepster are in favour, IchoTolot is against (he has said he likes seeing crowds of lemmings walk to the exit); and a 2-1 majority is hardly conclusive of anything. (I don't count Simon's view as known, since the reduction of lix counts in D Lix is a separate issue, resulting from the lack of player-controlled release rate.)

Anyway, I've looked over the 150 levels we've decided to include, with the following results:

<10 - 15 levels
10 - 6 levels
20 - 9 levels
50 - 54 levels
60 - 8 levels
80 - 17 levels
100 - 25 levels
Intermediate values - 16 levels in total

In short, there are "peaks" at 50, 80 and 100, with minor peaks at 20 and 60, other values being rare. Namida's proposal would result in similar peaks at 30, 40 and 50. The alternative of halving lemming counts would be very similar, but peaks would be at 25, 40 and 50, and the 60-lemming levels would retain their distinct identity as a 30-lemming peak instead of joining the 40-lemming peak.

However, there are two disadvantages to the halving system. 25 is an odd number, so save requirements would be a bit more awkward. There are no levels in the set with 25 lemmings pre-reduction, so aesthetically it might seem weird to suddenly make this the most common value. (Or maybe this is a good thing, as it makes a clean separation between the old and new systems?)

So, my vote is to go with namida's system, but only if the community agrees that reducing lemming counts is a good idea.

One more level where I'd suggest keeping the original lemming count is "Patience". It's the first level with RR 99, and this is made a lot less stressful for the player by the fact that you can lose as many as 40 lemmings.

* * *

I haven't looked at the time limits issue yet. Maybe we should concentrate on one issue at a time so that each can get fully discussed?

* * *

Regarding the Genesis/ONML levels, I counted how many levels use each tileset:

Dirt - 9
Pillar - 23
Marble - 29
Crystal - 13
Fire - 17
Bubble - 6
Brick - 26 (interestingly, there are very few Brick levels in the first two ranks)
Rock - 7
Snow - 15 (including Xmas)

There are nine levels in ONML that have a Genesis clone, and the Redux pack retains six of them:

Dolly Dimple (Brick / Fire)
The Stack (Snow / Crystal)
And now, the end is near (Brick / Pillar)
ROCKY VI (Rock / Dirt)
Time waits for no lemming (Brick / Marble)
Worra load of old blocks! (Brick / Marble)

So, except for "Dolly Dimple", in every case the choice is between two styles that are very close in usage count. And in the case of "Dolly Dimple", the principle of selecting the less used style would actually favour Fire, not Brick. However, Fire is still the fourth most used style, so I wouldn't weight this consideration very highly.

In short, I think we should select on an individual basis, prioritising which version of the level looks better rather than which style is most used.

[Nepster]

Although I am in favor of reducing the number of lemmings, I don't agree with those blanket reductions based on just their original numbers. Here are some reasons:
- Some reductions may add backroutes, e.g. in "No added colors and lemmings".
- Some reductions may prevent some solutions, e.g. in "I have a cunning plan".
- In some levels, the number of lemmings in the hatch after finishing the route to the exit might still be considerable.
- Some reductions may now allow to ignore one or more hatches, where previously lemmings from all hatches have to be saved.
(Note that I haven't looked at namida's version whether the last one occurs or not.)

I just played through the whole Redux pack in the version namida attached to post #253. Here is a list of lemmings number, that I would personally change, together with the number of lemmings spawned after I finished building the route to the exit. The numbers are displayed as "save requirement / total number of lemmings".

Big list (click to show/hide)

Fun  3: 5/20 instead of 5/50
Fun  5: 5/20 instead of 5/50
Fun  6: 5/20 instead of 10/50
Fun  7: 10/20 instead of 25/50
Fun  8: 10/20 instead of 25/50
Fun  9: 10/20 instead of 25/50
Fun 10: 10/20 instead of 25/50
Fun 11: 15/20 instead of 75/80
Fun 12: 30/40 instead of 50/60
Fun 13: 18/20 instead of 48/50 (spawns all)
Fun 14: 12/20 or 22/30 instead of 42/50
Fun 15: 20/50 instead of 20/100 (allow at least 20 lems to die)
Fun 16: 30/50 instead of 80/100 (spawns all)
Fun 18: 15/30 instead of 60/100
Fun 19: 10/20 instead of 25/50
Fun 20: 15/20 instead of 45/50 (spawns ~30)
Fun 21: 29/30 instead of 99/100
Fun 23: 15/20 instead of 75/80
Fun 24: 20/30 instead of 40/50 (spawns ~40)
Fun 25: 16/20 instead of 56/60 (spawns ~30)
Fun 27: 20/20 instead of 40/40 (spawns ~25)
Fun 28: 20/30 instead of 90/100

Tricky  2: 30/50 instead of 80/100 (spawns ~50)
Tricky  4: 20/20 instead of 80/80 (why need any more??)
Tricky  6: 20/20 instead of 50/50
Tricky  8: 10/10 instead of 80/80 (spawns ~7)
Tricky 10: 20/20 instead of 50/50 (spawns ~30)
Tricky 11: 60/60 instead of 100/100 (one solution requires ~50 lemmings)
Tricky 12: 15/20 or 20/25 instead of 45/50 (spawns ~25)
Tricky 14: 15/20 or 25/30 instead of 75/80 (spawns ~40)
Tricky 15: 15/20 instead of 45/50 (spawns ~25)
Tricky 16: 20/30 instead of 70/80 (spawns all)
Tricky 17: 10/20 or 20/30 instead of 50/60 (spawns ~40 until digging)
Tricky 21: 20/30 instead of 90/100 (spawns ~35)
Tricky 23: 33/40 instead of 53/60 (spawns ~10, but needs more lems to avoid saving only some of the hatches)
Tricky 24: 20/20 instead of 80/80 (spawns ~30)
Tricky 25: 20/20 instead of 80/80 (spawns ~15)
Tricky 26: 15/20 instead of 45/50 (spawns ~25)
Tricky 27: 20/30 instead of 40/50 (spawns ~35)
Tricky 30: 40/50 instead of 90/100 (spawns ~20, but needs more lems to avoid saving only some of the hatches)

Wild  1: 20/30 instead of 90/100 (spawns ~65)
Wild  2: 20/20 instead of 50/50 (spawns ~15)
Wild  5: 20/20 instead of 50/50 (spawns ~25)
Wild  8:  5/30 instead of 25/50 (spawns ~25, but need 25 to lose)
Wild  9:  5/10 instead of 45/50 (spawns ~7)
Wild 10: 20/20 instead of 50/50 (spawns ~25)
Wild 11: 19/20 instead of 99/100 (spawns ~20)
Wild 12: 20/20 instead of 50/50 (spawns ~40)
Wild 14: 29/30 instead of 49/50 (spawns ~25)
Wild 16: 30/50 instead of 80/100 (spawns all)
Wild 17: 18/20 instead of 58/60 (spawns ~40)
Wild 18: 90/90 instead of 100/100 (to fill the upper loop completely)
Wild 20: 19/20 instead of 49/50 (needs last lemming)
Wild 22: 17/20 instead of 47/50 (spawns ~10)
Wild 26: 29/30 instead of 99/100 (spawns ~60)
Wild 27: 14/20 instead of 44/50 (spawns ~20)
Wild 28: 20/20 instead of 50/50 (spawns ~30)
Wild 29: 18/20 instead of 48/50 (spawns ~40)

Mayhem  1: 20/20 instead of 50/50 (spawns ~30)
Mayhem  2: 20/20 instead of 50/50 (spawns ~30)
Mayhem  3: 20/20 instead of 100/100 (spawns ~60)
Mayhem  4: 20/20 instead of 75/75 (spawns all)
Mayhem  5: 20/20 instead of 50/50 (spawns ~20)
Mayhem  6: 30/30 instead of 50/50 (spawns ~30)
Mayhem  7: 20/20 instead of 70/70 (spawns ~50)
Mayhem  9: 10/20 instead of 40/50 (spawns all)
Mayhem 10: 10/20 instead of 40/50 (spawns ~30)
Mayhem 11: 16/20 instead of 76/80 (spawns ~20)
Mayhem 12: 20/20 instead of 85/85 (needs last lemming)
Mayhem 13: 20/20 instead of 80/80 (spawns ~20 without RR change)
Mayhem 15: 20/20 instead of 100/100 (spawns ~30)
Mayhem 16: 20/30 instead of 85/100 (spawns all)
Mayhem 17: 20/30 instead of 40/50 (spawns ~20)
Mayhem 19: 15/15 instead of 80/80 (spawns ~12)
Mayhem 20: 13/40 instead of 83/100 (spawns ~40)
Mayhem 21: 10/20 instead of 40/50 (spawns all)
Mayhem 22: 18/30 instead of 48/60 (spawns ~10)
Mayhem 23: 19/20 instead of 49/50 (spawns ~30)
Mayhem 24: 15/20 instead of 75/80 (spawns ~40). I guess this reduction allows saving one more lemmings than in the original version by a more precise version of the standard solution.
Mayhem 26:  9/10 instead of 29/30 (spawns ~10)
Mayhem 28: 20/20 instead of 50/50 (spawns all)
Mayhem 29: 20/20 instead of 50/50 (spawns ~35)
Mayhem 30: 20/50 instead of 70/100 (spawns ~40)

Havoc  4: 15/20 instead of 45/50 (spawns all)
Havoc  5: 19/20 instead of 79/80 (spawns ~30)
Havoc  6: 15/20 instead of 45/50 (spawns all)
Havoc  8: 20/40 instead of 60/80 (spawns all)
Havoc  9: 20/20 instead of 60/60 (spawns ~30)
Havoc 10:  8/10 instead of 68/70 (spawns ~10)
Havoc 13: 19/20 instead of 49/50 (spawns ~30)
Havoc 14:  6/10 instead of 46/50 (spawns ~6)
Havoc 15: 20/20 instead of 50/50 (spawns ~20)
Havoc 17: 20/20 instead of 60/60
Havoc 18: 15/20 instead of 45/50 (spawns all)
Havoc 19:  9/10 instead of 29/30 (spawns ~15)
Havoc 20: 20/20 instead of 100/100 (spawns all)
Havoc 21: unsure whether to change at all, perhaps to 34/40?
Havoc 22: 19/20 instead of 49/60 (spawns ~25)
Havoc 23: 28/30 instead of 68/70 (spawns all, but 30 are sufficient to fill to crowd area)
Havoc 25: 30/30 instead of 80/80 (spawns ~40)
Havoc 26: 20/20 instead of 80/80 (spawns ~8)
Havoc 28: 59/60 instead of 99/100 (spawns ~50)
Havoc 29: 20/40 instead of 80/100 (spawns 90)
Havoc 30: 20/20 instead of 50/50 (spawns all)

And now regarding time limits. I generally would remove even more time limits than namida:

Another big list (click to show/hide)

Tricky 8 "Just A Quicky": If we reduce the number of lemmings, then the current time limit does not pose any problems at all. Can anyone show me solutions that are only possible if we remove this time limit? If not, then I prefer to remove the time limit here.
Tricky 28 "Speed Trap": I agree - keep this time limit.
Wild 15 "No Time for a Detour": I agree - keep this time limit.
Wild 21 "Four Play": I agree - keep this time limit.
Mayhem 8 "Five Alive": I don't see any way to solve the level using more than one minute. So remove the time limit.
Mayhem 13 "Synchronized Lemmings": I agree - keep this time limit.
Mayhem 14 "Just A Minute": I agree - keep this time limit.
Mayhem 17 "The Silence of the Lemmings": I, too, don't know what function the time limit has in this level. So I vote to remove.
Mayhem 18 "The race against cliches": I would add this to the list of levels that keep their time limit. Otherwise there is no race at all in this level: The top lemming finishes his work and then the bottom lemmings start...
Mayhem 24 "Onward and Upward": I don't see any reason to keep the time limit here, especially if we reduce the number of lemmings so that the player doesn't has to set the RR to 99. My solution doesn't use any obscure tricks.
Havoc 1 "Its Hero Time": I can see a few additional ways to solve the level without the time limit. Fixing all those would be a major task. Moreover I always thought the title of this level to have a double meaning, one referring to the very tight time limit. So keep this time limit.
Havoc 6 "Lemming About Town": Remove the time limit here. I don't know any reason to keep it.
Havoc 12 "Just A Minute (Part Two)": I agree - keep this time limit.
Havoc 18 "Scaling the Heights": I had roughly one minute left on the clock and don't know of any solutions requiring more time than given. So what is the reason the time limit should stay?
Havoc 24 "Hello John Got A New Lemming": The time limit is very tight, but what function does it serve beyond anoying the player? I vote for removing this time limit.
Havoc 26 "Oh No! Squish": This is another level where reducing the number of lemmings makes the time limit pointless. Currently one has to think about setting the RR to 99 early enough. But is there any solution that requires more time than the given two minutes?

Finally some general comments and questions:
- Fun 9 has lots of space on the left-hand side and the whole playing are is a single screen on the right edge. Should we crop the level a bit?
- Fun 22 has a screen start at the right edge, away from the hatch and exit. This should be changed. Moreover I would prefer enlarging the OWWs to cover the whole terrain piece and not only the middle part. This should just be an aesthetic change.
- Tricky 17 has a wrong screen start position, too.
- Tricky 28: Same problem with the screen start position.
- Havoc 17: Is is intended that one can start digging directly from the starting platform?
- Havoc 22: Give another builder to adapt to stricter builder terrain checks? Currently the level is solvable, but I don't know any nice and elegant solution.
- Havoc 24: Why the deadly fall out of both hatches? What do you think about moving them down and giving only one floater?
- Havoc 30: Has anyone solved this level in NeoLemmix? My old solution with building from the basher tunnel into the miner tunnel no longer works.
- There are a lot of levels where either the hatch or the exit are not set to no-overwrite, e.g. Fun 30. This creates problems when one wants to build in that area.

[bsmith]

I was just thinking of doing what Nepster did, playing the levels and seeing how many lemmings are spawned when I am effectively done.  I hope to get through all the levels through the weekend, then I can post my numbers and see how close I am to Nepster.

[Proxima]

In that case we should definitely postpone the question of lemming counts until people have had more time to play through the pack.

Meanwhile, I've looked through to decide how I feel about time limits. Other than the 16 levels already mentioned, I don't see any that would benefit from retaining a time limit, so I think we can concentrate on just those 16.

These levels should definitely keep the time limit, as it's the main point of the level: "No time for a detour", "Four Play", "Synchronized Lemming", "Just a Minute" (both parts), "The race against cliches".

All other levels could do without one. However, some levels certainly had a short time limit chosen for effect. I believe "It's hero time" is one of these, rather than the time limit being intentionally chosen to prevent backroutes (although it's true that the time limit does prevent some solutions that would otherwise work).

My inclination is to keep one-minute time limits (Speed Trap, Five Alive, It's hero time) and time limits that provide a psychological effect of being pressed for time (Scaling the Heights, Hello John). On these levels, it feels that the time limit should not be much annoyance, as it's hard to carry out the correct solution without being in time.

(Note that the title of "Speed Trap" seems not to refer to the time limit, but to having to hurry at the start to keep the crowd under control. So I don't think this is a level where the time limit is important to keep, though I would prefer to keep it.)

There are several 2-minute levels in the pack: Just a Quicky, Twins, Let's get together, The Lemming Funhouse, Watch right or left (Part 2), Scaling the Heights, Oh No! Squish, I am A. T. On most of these, the level actually requires considerably less time so the time limit doesn't really do anything. (This is true of Just a Quicky even with its original lemming count.)

The title of "Just a Quicky" does suggest a short time limit, so an alternative would be to reduce it to one minute. This wouldn't really make it any harder (especially with a reduced lemming count) and it would be a good level for a new player's first introduction to time limits.

Spoilers for ONWARD AND UPWARD:

Spoiler (click to show/hide)

The "obscure trick" namida referred to is to send up two climbers close together, and bash the thin pole with the first one, starting a step back so its first stroke ends in the middle of the pole, turning the other climber. On the original game, this is the only way to achieve lose-3 (not possible on NL since it also uses the left edge of the level). This trick isn't necessary to finish in time, although without using the trick, I was down to the last second. I don't think the intention was to require this trick, so I would prefer to remove the time limit.

[namida]

Screen start issues should be fixed in the newer version.

Regarding that we might be moving too fast - these changes can easily be reverted if need be. Likewise, if any further cases need manual adjustment rather than the defaults, that can be done too.

[namida]

Going to have a proper look through the feedback now:

One more level where I'd suggest keeping the original lemming count is "Patience". It's the first level with RR 99, and this is made a lot less stressful for the player by the fact that you can lose as many as 40 lemmings.

Not completely against this, but I think lose 20 is still very generous. Personally, on my first playthrough, I didn't find this level stressful at all - and this was on the Sega Master System version, where this level only allows losing 10 lemmings (and yes, unlike some other levels (even including this level's repeat), this level remains RR99 on that version). But yeah - I'm okay to go either way on this one.

I haven't looked at the time limits issue yet. Maybe we should concentrate on one issue at a time so that each can get fully discussed?

Removal of time limits is, beyond any doubt, an established NL convention. So it'd be really weird not to do so for this pack. The reason I did them at the same time was because, with it being automated - as well as needing a manual looking through to reverse some cases - it was much easier to do everything in one go to come up with a starting point, from which any individual cases can be reverted as nessecary. (If we do decide against these changes altogether - which is unlikely for the "no hidden traps" or "remove unneeded time limits", maybe a bit more debatable for the "reduce lemming counts" - I can just throw together another bit of code to reverse them.)

(regarding styles)

Hm, that changes things then. We may indeed want to consider the Genesis variants of these levels. The only counter-argument I could think of is that the ONML variants may be a bit more recognizable to some people.

- Some reductions may add backroutes, e.g. in "No added colors and lemmings".
- Some reductions may prevent some solutions, e.g. in "I have a cunning plan".
- In some levels, the number of lemmings in the hatch after finishing the route to the exit might still be considerable.
- Some reductions may now allow to ignore one or more hatches, where previously lemmings from all hatches have to be saved.

We can always reverse the count reduction (or even just reduce it by a lesser amount) on levels where it has these issues.
However, I'd be interested to know which solutions you're referring to in the first two points?
For "No Added Colors Or Lemmings", the obvious one may be to instant RR99, bash with the earliest lemming that would result in all lemmings still turning around, then build to cross the gap before the crowd gets back. The most optimal execution of this I can manage on the current version of the level (30 lemmings) still ends up with 5 lemmings falling down. You can reduce this to four by using the usual trick for the end of that level, or to three by simply using the now-spare climbers to delay some of the crowd lemmings, but neither is enough to complete the level.
For "I Have A Cunning Plan", I assume you're referring to the solution using the Medi 23 trick? This solution requires at least 42 lemmings; the modified-lemming-count version of this level has 50.

[namida]

- Fun 9 has lots of space on the left-hand side and the whole playing are is a single screen on the right edge. Should we crop the level a bit?

I'm okay either way on this one. What does everyone else think?

- Havoc 17: Is is intended that one can start digging directly from the starting platform?

Not possible on the DOS / Lemmix version (and thus presumably not Amiga either), so this probably should be fixed.

- Havoc 22: Give another builder to adapt to stricter builder terrain checks? Currently the level is solvable, but I don't know any nice and elegant solution.

When it was first brought up that the direct conversion of this level was impossible, I considered the options to make it possible again, and IIRC I found that an extra builder either made it too easy or added a backroute, so I opted for an extra basher instead.

- Havoc 24: Why the deadly fall out of both hatches? What do you think about moving them down and giving only one floater?

Only read if you've solved the level. (click to show/hide)

I thought that too at first. But after I solved the level, I realised that only giving one floater would make the solution a bit less hidden. By requiring them all at the start, it doesn't imply "the worker needs a floater at some point", which hides the solution a bit more.

The time limit here actually works similarly - it prevents a solution where you let the worker continue going then dig down to the left-exit, thus penalizing the player if they bash the OWW at the top too early. This is more subtle, but similar to, how the one at the bottom also shouldn't be bashed too early.

- Havoc 30: Has anyone solved this level in NeoLemmix? My old solution with building from the basher tunnel into the miner tunnel no longer works.

Can confirm it's solvable on V10.13.16 exps, haven't tested on stable but I doubt any of the changes will affect this level.

- There are a lot of levels where either the hatch or the exit are not set to no-overwrite, e.g. Fun 30. This creates problems when one wants to build in that area.

Yes, this should be fixed.

(Spoilery stuff about Onward And Upward)

I actually couldn't think of any other way to solve the level, to be honest.



By the way, to avoid confusion (especially if we change the order later) we may want to either include titles when referring to levels, or refer to them by their original game positions rather than Redux positions. Former is probably a bit more convenient while working on Redux.

[Proxima]

Removal of time limits is, beyond any doubt, an established NL convention. So it'd be really weird not to do so for this pack.

I know; but my point was that (as you can see above) there's still a lot to discuss on this issue, and this is made harder if we're discussing several issues at once so that we have to look a long way back in the topic for everyone's thoughts on a particular issue. Anyhow, I've given my thoughts on time limits now, so I'll wait to see if we reach a consensus.

Regarding the particular levels that have been mentioned:

Snuggle up to a lemming: I would favour mobius' suggestion of moving the exit so as to use more of the level area. If not, then I believe we should keep the decorative terrain, as the original games make good use of decorative terrain to produce memorable levels. (We are keeping the decorative terrain in "Turn around young lemmings", aren't we? [Goes to check] Yes.)

Steel Block Party: Certainly, this should be fixed, as it trivialises the level.

Hello John: Agreed with namida's reasoning for keeping the splat start.

ONWARD AND UPWARD: The other solution is to send up two climbers, block one on the thin pillar to turn the other, also block the front lemming of the crowd. This feels more intended, as it uses every given skill except the second basher. (I was wrong about lose-3 being impossible on NL; you do lose the climber-miner when he splats after climbing the left wall, but this is still lose-3. So that could be a talisman.)

No added colors: Even if the backroute doesn't work with 30 lemmings, the player is more likely to waste time attempting it, so I would keep the original lemming count for that reason.

One more level to keep the original lemming count: The gate trap lemmings. It feels like having to avoid the stragglers from the left hatch splatting in the digger pit was intended to be part of the level.

* * *

As for the confusion when referring to levels by rank and number, this is going to remain an issue when the pack is published, so we should come up with unique rank names for Redux, preferably a system of adjectives gradually increasing in intensity like the original games, but different from any rating names already used.

My suggestions: Gentle, Quirky, Zany, Manic, Lunatic. "Quirky" has been used by the Lix pack, but no NL content.

[Nepster]

For "No Added Colors Or Lemmings", the obvious one may be to instant RR99, bash with the earliest lemming that would result in all lemmings still turning around, then build to cross the gap before the crowd gets back. The most optimal execution of this I can manage on the current version of the level (30 lemmings) still ends up with 5 lemmings falling down. You can reduce this to four by using the usual trick for the end of that level, or to three by simply using the now-spare climbers to delay some of the crowd lemmings, but neither is enough to complete the level.

Yes, I thought one could save even more lemmings this way. Note however that delaying two lemmings by letting them climb on the left-hand side allows even saving 27 lemmings.
Anyway all of this means that while the backroute does not quite work, this is by no means obvious: It would work if the gap was one pixel smaller or if one could delay three more lemmings. So I would really prefer to keep the original 50 lemmings just to make it obvious from the very beginning that this appraoch won't work.
Edit: Proxima was faster

For "I Have A Cunning Plan", I assume you're referring to the solution using the Medi 23 trick? This solution requires at least 42 lemmings; the modified-lemming-count version of this level has 50.

Yes, I refer to the Medi 23 trick, assuming this means

Spoiler (click to show/hide)

blocking directly below the hatch and then free the blocker with a newly spawned lemming.

Again I would give a little bit more lemmings than the bare minimum, so that less experienced players have no problem executing this solution. As the idea is by no means obvious, I think we should honor them for the idea and not make the execution harder than necessary. So I still vote for 60 instead of 50 lemmings, especially as the level has RR99, i.e. becomes only 3 seconds longer with this change.

- Havoc 24 "Hello John": Why the deadly fall out of both hatches? What do you think about moving them down and giving only one floater?

Only read if you've solved the level. (click to show/hide)

I thought that too at first. But after I solved the level, I realised that only giving one floater would make the solution a bit less hidden. By requiring them all at the start, it doesn't imply "the worker needs a floater at some point", which hides the solution a bit more.

The time limit here actually works similarly - it prevents a solution where you let the worker continue going then dig down to the left-exit, thus penalizing the player if they bash the OWW at the top too early. This is more subtle, but similar to, how the one at the bottom also shouldn't be bashed too early.

Mild spoilers (click to show/hide)

I don't think giving only one floater is really a big hint.
Regarding the time limit: Fair enough that it prevents one solution, but with the current time limit I cannot first assign all the floaters and then start with the climber, because for this I need at least 5 seconds more. So I already have to assign the climber while lemmings are still spawning, which I dislike. And it prevents the other solution only barely. I attached the screenshot of the last frame when I tried (an optimized version) of the alternative approach.
So overall I am more in favor of changing the level slightly in other ways to make sure that the alternatice approach doesn't work (e.g. removing one basher and one digger from the skill set).

- Havoc 30 "Lemming's Ark": Has anyone solved this level in NeoLemmix? My old solution with building from the basher tunnel into the miner tunnel no longer works.

Can confirm it's solvable on V10.13.16 exps, haven't tested on stable but I doubt any of the changes will affect this level.

Ok, I managed to solve the current level in the stable version, but...

Spoiler (click to show/hide)

The basher needs to be placed at exactly the correct height to prevent bashing further than one block. So I would stongly suggest enlarging the thin horizontal gap to 4 pixels height instead of only 2 pixels.