|
|
 |
|
| Gullet |
Sat 21 Dec 2013 22:00 |
 |
<N7DOT> 'Gullet' uploaded by N7DOT:
Hmm... there is an abundance of pieces but no obvious way to get them out... |
|
| |
| Get Movin' |
Sat 21 Dec 2013 16:11 |
 |
<Jim> (SPOILER) Devin, isn't this the same thing as Soko_Very_Hard? |
|
| |
| ScrambleMods 1 |
Sat 21 Dec 2013 01:29 |
 |
<Dave> (SPOILER) Added speedrun: 23 moves (old: 27). |
|
| |
| Squish |
Sat 21 Dec 2013 01:24 |
 |
<Dave> (SPOILER) Added speedrun: 466 moves (old: 470).
Messy first solve. I like it, it's deceptive: it looks hard but it actually isn't. :=) |
|
| |
| ScrambleMods 1 |
Fri 20 Dec 2013 22:46 |
|
|
<N7DOT> 'ScrambleMods 1' uploaded by N7DOT:
Ok, so -this- is a modification/organic level series where instead of a nice, tidy starting point with no contents, the starting point is a random soup. Stuff wil likely have to be removed, but that's supposed to happen here. This idea was based off how my first Improv level got modded a whole bunch.
Also, maybe we could avoid making this get super difficult fast, just so that everybody can get ample time to contribute? The newer attempts have been pushed to maximum difficulty at much faster rates which means that very quickly it's just going back and forth between a couple authors because they are the only ones who can solve it. |
|
| |
| Squish |
Fri 20 Dec 2013 22:25 |
|
|
<N7DOT> 'Squish' uploaded by N7DOT:
Probably super easy - maybe make a mod of it? (i.e. timer addition might be ok as it is so short, addition of or changing hazards, etc.) |
|
| |
| tesseract |
Fri 20 Dec 2013 22:06 |
 |
<N7DOT> It would be nice if you added some way to see what the puzzle is implementin, and what each move did on the tesseract. (a comment describing in fuller detail would be nice too) |
|
| |
| Get Movin' |
Fri 20 Dec 2013 20:48 |
|
|
<devin> 'Get Movin'' uploaded by devin:
Another Sokoban level remade. |
|
| |
| Crunches |
Fri 20 Dec 2013 04:50 |
 |
<mm> (SPOILER) Added speedrun: 397 moves (old: 434). |
|
| |
| tesseract |
Fri 20 Dec 2013 02:30 |
|
|
<Dave> (SPOILER) Interesting problem. My first reaction after reading you comment is to tackle it from a completely different angle. Whether my thinking is correct remains to be seen. Hopefully have time to do a bit of programming over the weekend. :) |
|
| |
| Crunches |
Fri 20 Dec 2013 01:10 |
|
|
<benjamin> (SPOILER) Added speedrun: 434 moves (old: 479). |
|
| |
| tesseract |
Fri 20 Dec 2013 00:49 |
|
|
<benjamin> 'tesseract' uploaded by benjamin:
All of the wire tiles that must be in place to finish the level represent vertices on a 4d hypercube, as do the buttons that you can press. When pressing a button, it activates vertices connected to it in 4d space. I generated this level using a program, and the maximum number of moves needed for any configuration that I have found searching through literally thousands of them is 5. I am attempting to figure out why that is the case. There are 32768 possible configurations, but only a few of them are solvable. When running my program on completely random levels instead of scrambled levels I never found a solvable one (giving up after 10 moves and running for 30 minutes). |
|
| |
| Crunches IV |
Thu 19 Dec 2013 22:15 |
 |
<Dave> (SPOILER) Added speedrun: 408 moves (old: 412). |
|
| |
| inversion 2.1 |
Thu 19 Dec 2013 11:30 |
 |
<Jim> (SPOILER) Added speedrun: 76 moves (old: 96). |
|
| |
| Double Sokoban |
Thu 19 Dec 2013 11:29 |
 |
<Jim> (SPOILER) Added speedrun: 147 moves (old: 285). |
|
| |
| inversion 2.1 |
Thu 19 Dec 2013 05:24 |
|
|
<Dave> 'inversion 2.1' uploaded by Dave:
Not much of a mod but it does eliminate the extraneous inversion. |
|
| |
| inversion 2 |
Wed 18 Dec 2013 21:51 |
 |
<N7DOT> (SPOILER) I get the feeling this is cookable, yet it manages to evade my attempts. If this is possible it would have to be really hard. |
|
| |
| Double Sokoban |
Wed 18 Dec 2013 21:15 |
|
|
<N7DOT> 'Double Sokoban' uploaded by N7DOT:
If anybody wants to try building a better sokoban pair using this concept, go ahead. The first puzzle is really easy, and the second puzzle is medium-easy, but the point is that the blocks and walls swap after the first half, so the walls become blocks and the blocks become walls. |
|
| |
| inversion 2 |
Wed 18 Dec 2013 20:56 |
|
|
<benjamin> 'inversion 2' uploaded by benjamin:
This has a similar idea as inversion but uses transponders and has no robots. |
|
| |
| Crunches IV |
Tue 17 Dec 2013 21:37 |
|
|
<mark> (SPOILER) Added speedrun: 412 moves (old: 440).
I must have went through everything that was not possible before finding the possible.
Great mod Jim once again... |
|
| |
| Sokobite |
Tue 17 Dec 2013 21:35 |
 |
<N7DOT> (SPOILER) ... I'm also 99% sure that 71 moves is optimal, but here the proof is non-rigorous and may have a flaw. |
|
| |
| inversion |
Tue 17 Dec 2013 21:33 |
 |
<N7DOT> (SPOILER) I'm fairly sure my solution is provably optimal through a parity argument (the number of steps required to get to exit if the gate cell is initially left untouched must be even, and the distance from the goal the generation before it is on must be greater than 1. so for even parity this is the fastedt possible. for odd parity, you need to go up and around the bot, which takes 4 moves, which is at least 3 moves slower as the parity change saves one move (and yes, I have found the 53 move solution.)
<N7DOT> (SPOILER) Added speedrun: 50 moves (old: 59).
multitasking
<N7DOT> (SPOILER) Added speedrun: 59 moves (old: 60).
One move shaved...
<benjamin> 'inversion' uploaded by benjamin. |
|
| |
|
|
|
|