lemforum, D replay collection

[Nepster]

Here is a version of Trading and Cooperation with a different hatch order, so that I can solve the level. This is mainly meant to have at least some version that certainly works in D-Lix. geoo should have a look and decide what to do with it.

[ccexplore]

One that I'm wondering about is Top Gear (attached).

As I recall, geoo and I talked about it some time ago, and came to agreement that there was another level (I forgot which one) that had a very similar concept but was better in some ways (probably execution-related), so that level was picked over this one for inclusion.  Thus as much as I'd love to have another level of mine included, I don't think it'd be a good choice.   In any case I don't think it'd ranked for Hopeless (though of course there's always the option to insert into a lower ranking and move some other level into Hopeless 1).

[geoo]

The original version of Trading is perfectly solvable. Maybe the level has to be ranked up a bit?

Solutions for those three levels are attached.


If some level need to be replaced, I have one rather hard level that I never released. I think it still needs some backroute checking, but tell me if you're interested.

[Simon]

Solutions for those three levels are attached.

These 3 replays all solve. The entire pack is 100 % covered. I've updated my replay collection in the first post.

Backroute-fixing or execution-easening can break replays again. Let's preserve the perfect coverage by sending proofs of solvability along with every serious change.

About Trading, from IRC:
Simon: geoo, your hatch order is a local optimum? Every hatch is exactly as you want it, or could you make execution easier still?
geoo: hatch order is probably the only one that makes my solution work
geoo: execution is not that bad the way it is I think, once you figure out the moves

-- Simon

[Proxima]

New version of Eye of the Needle: Fixed Nepster's backroute.

[Nepster]

Finally solved the original version of Trading and Cooperation, which turned out to be very similar to my solution with the wrong hatch order. I am not totally happy with the current version, because execution is pretty tough, especially two points:

Spoiler

1) Placement of the blocker: One brick less and the builder does not close the gap to the upper platform and one brick more and the jumper may not turn. Why not make the middle gap slightly larger and let the platformer become a blocker after placing the last brick? This would be much more natural.
2) After having placed the blocker correctly, the jumper placement is extremely tough to get right: She has to start at the very first moment it becomes possible without hitting the ceiling, otherwise the jumper will not make it to the blocker or jumps over the blocker. After turning around she will barely land at the outmots pixel of the lower platform. With all the required (almost) frame-precise assignments, it is far from obvious that this jumper trick works at all! In fact, this was the reason why I couldn't solve the level for a very long time: I thought this trick was impossible and looked for alternative approaches.
I am not sure what the best method is to simplify execution here. But I found that letting a lix jump, who moves in the same direction as the platformer (either while still platforming or jumping over the blocker), is a lot easier to execute.

[Proxima]

Simon asked me for an up-to-date zip of the lemforum pack, which I've attached. There are two levels for which I don't have up-to-date versions:

Won't Get Fooled Again: I believe lowering the lix count on this level has been suggested, but I don't know the details.
We're in This One Together: I believe redesigning this level has been suggested, but as I haven't solved it, I haven't been following closely.

I've also attached my replay for the current Slipping Again. Let me know if there are any other levels that need new replays.

[Simon]

Merged.

-- Simon