For me it's a simple concept: The ranks display the difficulty setting.

I doubt every player will play other games expecting to beat everything on hard difficulty. No, most players stick to easy or normal.

So I see it as the following most players stick to the early and mid ranks that represent easy and normal difficulty. They always have the ability to check into the later ranks (hard difficulty) to improve themselves even more and expand their limits.

Already experienced players will most likely rush through easy and normal difficulty only moderately get challenged (most levels won't leave a big mark on them), while the real interesting parts and the core challenge lies in the hard difficulty for them.

Talismans hide away hard difficulty and water down the experience for those players. Like they get reheated food (levels in this case). With my method everyone gets fresh food and everyone is free to try everything and see what tastes best for them.

We finally reached the last group of this phase!

Round 1 is open for voting.

Anyone?

I think it's on Nepster's fix list. Still he has to come up with a solution, it could be that he still works on that.

Expect a solution/fix with the next patch.

A while ago I found a presentation from a Game Develoer Conference and I was just remembering it now and thought: Why haven't I linked it here?

It's ~ 1hour long and it's called: Empowering the Player: Level Design in N++
As you can tell from the name N++ is the example here and the developers who created thousands of levels for it are talking about game and level design.

Now not all the points they bring up can 100% be projected on Lemmings or Lix levels in my opinion. For example the point: Let players choreograph, not just execute.

Of course only having precise trick levels is not very good, but especially on higher difficulties levels Lemmings level are a bit more about one or a small group of solutions. On the lower ranks this point fits way better, but in higher difficulty ranks you often want to enforce certain patterns or parts and I think finding the right mixture of restricted and a bit more open levels is the best way.

Other points they bring up:

- Minimalism on the core level idea + expanding that idea with more and more details for higher difficulties. Also the danger of overdoing it with adding too much.

- Level creation process - from core structure to the finished level.

- The nessesarity of culling stuff. Scrapping of levels and ideas - not everything fits and it can be better to cut it out.    - This point I feel needs to be addressed more here in the case of levels. For mechanics: Yes, I know I argued a lot about keeping stuff and features, but I also was in favor of culling a few things (gimmicks as an example). I still believe the right balance is correct and I don't wanna see things that offer a lot of gain and just fit disappear.
Sometimes levels just don't work out that well. I think that can also be expanded to features and machanics. They talked about at some point they added portals in N++, saw that it makes things too complicated in a fast platformer and threw them out again. I think Slowfreeze, Radiation and Gimmicks are the NeoLemmix equivalent here with them adding only unnessesary precision and complexity without offering too much.

- Balance between "boring/calm" and "exciting". Calm parts make exciting ones stick out more.

- Smooth difficulty curves and inclusion of breathers.

- Mastering of skills and levels of high complexity + difficulty.

- Variance in level desgin - one type of level is not enough!  - I especially agree with that point. You don't wanna make only huge/tiny/button or trick levels for example. You wanna have a huge variation of levels that has something for everyone.

- Naming of levels.

- Access to all/a lot of the levels from the getgo to avoid getting stucked completely.      - We solved that with unlock all.

 
So, I think maybe you can take a few things with you from this presentation, even if it's another quite different game they are referring to and not everything can be transferred to Lemmings 1:1.

Annnnd the next round is up.

A related request: currently when the grid is off, ctrl-arrow moves a piece by 8. When the grid is on, ctrl-arrow moves a piece by 8 times the grid size. Much more useful would be, in the latter case, to move a piece by 1.

This would deal with some of the same use cases as the on/off hotkey and would make it less urgent, but I believe it would still be very beneficial to have both.

I agree on everything.

There are 2 things currently feeling not quite right to me about this tileset by looking at the picture:

- The background makes it somewhat hard to see a bunch of stuff and the stones look a bit weird let's say - more like a sewer than a basement. I feel this can still be optimised by quite a bit.

near the top exit

- Is the exit these 2 blue things (flags?) on the wall? I must admit I can't really tell