| « home | Browsing all comments (spoilers: on | off) (info: on | off) | next 25 |
 |
<N7DOT 3> At first I didn't know what level this was based off of due to the name "Mars Drift", but when I saw the group of four bots at the top, I quickly realized the level was another modification for "Seven bots for Seven buttons" I'm pretty sure this will get good ratings. |
 |
<Dave> (SPOILER) Added speedrun: 569 moves (old: 570). Now that's a bit more like a cook... and all to shave 1 move. |
 |
<radiant> (SPOILER) That move sequence isn't a solution to the Pons Asinorum analog position, and it's not actually a solution here either (since the exit doesn't open up). In addition to checking for the right colors, the panels also encode edge parity, even if not in a way that's particularly visible. <N7DOT 3> still really good level, though <N7DOT 3> darn, i just realized that should have been a spoiler! at least the level is easy enough without help... so WARNING! DO NOT READ COMMENT DIRECTLY BELOW IF YOU DON'T LIKE SPOILERS! sorry, realised it immediately after posting <N7DOT 3> very nice implementation. however, there are two non-equivalent positions that satisfy the starting position, meaning that from the position given, the sequence L U R D L U results in a solution, while on the actual puzzle, depending on which starting position is chosen, it may or may not result in a solution, and so although the machinism may be flawless, it isn't a full implementation. |
 |
<N7DOT 3> Sorry, I just had alot of trouble making this level, and I'm also sorry about forgetting to credit the inventor of the puzzle this is based on. <Dave> radiant: binary adder was an impressive demonstration of how much boolean algebra can be compressed into such a small area. I agree that the 3x3x1 is implementable, but for this to be a worthwhile excercise we need to get into 3 dimensions and find a solution to the 2x2x2 problem. My question is: how do you implement six colours whilst maintaining any reference to the tactile feel or appearance of the cube? <radiant> Observe that in a Rubik's Floppy using U, D, L, and R as generators, each move consists simply of swapping two pairs of blocks, and the central block on each side is invariant under all moves (except that on the actual device it has a parity signal, not that this has any effect on otherwise-valid solutions). I see a line of attack to implement this without allowing invalid moves (pushing a side off to the right or down and then not putting it back, or else manipulating the blocks so that they go back in non-reversed order), and if I can do it in reasonably compact space then I might just upload it. It's well-established by now that Escape is Turing-complete and so even the full 3x3x3 cube would be possible to implement in theory, except that the limits on board dimensions probably cap at too low a value to allow all 6 generators to be expressed. Demonstrations like this can still get points in style; they just need to be sufficiently recognizable (the 1x3x3 "puzzle" is far more obscure than the traditional 3x3x3, and even the 4x4x4 and 5x5x5 variants) and work properly. See something like #2894 (8-Bit Binary Adder), which has the additional point that there's at least an (admittedly trivial) goal which has to be accomplished before the exit opens, as opposed to just being able to head off to the exit at any time, even mid-move. In fact, if not for me telling you now, you wouldn't know that the level was originally supposed to have a "Clear" mechanism so you could reset everything and try multiple additions without having to go back to move 0; this was eventually scrapped after I had to settle for a carry mechanism that couldn't account for the garbage it would leave. <Dave> A Rubik's Cube implementation relies upon simultaing a rotational movement with only linear motion available. This is where your level fails as soon as you put the side back in place and try to make a second turn. If I had thought of a way of doing this I would have done it already, which is not to say it's impossible, I just can't see a suitable method. I will keep pondering it though. PS. You already know my thoughts on the ratings situation, it is a problem. <N7DOT 3> oops... i meant ALL my implementations, not NONE of my implementations... <N7DOT 3> why did I waste so much of my time on this level? none of my implementations are either already done or flawed to the point that they no longer are worth the time, it seems. I have two other implementations that will be neat, and one will be kind of difficult, but they are difficult enough to make that i don't see myself uploading them. if this level has poor style, i'd really like to see somebody actually TRY to make a rubik's cube implementation WITHOUT using this level as a guideline; A.K.A. from SCRATCH. sorry if this seems a harsh conclusion... it's just there are some levels i really want to make, but the work is too much to be justified, due to getting any rating beyond 5 for style becoming very hard to get, unless the difficulty is above 6 and the level still has style above 4 otherwise. I Absolutely hate to say it, but it really does seem like Escape is dying. |
|
|
<Dave> (SPOILER) Hi kyran, I forgot to mention something: This level is messy... but I for some reason really I like it. |
 |
<Gil Anthony> Most levels I get,this one not so much. I just hate it when I got to revisit. good level. |
|
|
<Dave> (SPOILER) The little 7x5 room with the switches and lights near the end. <kyran> (SPOILER) Which room? <Dave> (SPOILER) Still skipped a room. We'll call it a shortcut rather than a cook :) |
 |
<Gil Anthony> I took the LONG escape exit 2404 moves before I saw the 15 move short cut.Luck me. |
 |
<N7DOT 3> (SPOILER) Added solution "New 'speedrun'": Well, I've set a newer legal speedrun, but as the original is unbeatable, I have to upload this as a non-speedrun. |
|
|
<N7DOT 3> Kyran: Normally you have to delete your levels, by hovering the mouse over the level and pressing the delete key. (only the author of the level can do this) <N7DOT 3> (SPOILER) Cook: same thing... <kyran> I fixed it, delete this one. <Zog> (SPOILER) Cook: Did the author really intend a fifteen-move finish? |
 |
<Karlodinium> (SPOILER) Cook: Skipped the bottom right section. |
 |
<N7DOT 3> It also doesn't discriminate against using the mouse to press the buttons, or playing more than once (by pressing enter at death, or closing and opening level). Tyler, Please play some of the higher rated levels like OneLeft (full implementation) or any of the Memory Hole series, and see for your self what players want. If your even reading this, that is a good thing, because the people who play this game enough to keep it alive don't like these levels. I know that when I started playing, quite a few years before I registered, I didn't know what good levels were. However, I quote another player here ( wowei or kirima/mm), that in many games, the fun is in getting surprised or in some hidden information, and that some players like this game simply because it is not one of those games. If your levels are going to rely on hidden information for gameplay, try to be EXTREMLY innovative and creative, as these techniques get old immediately after it's debut. For example, see noname's "Guesswork" or Joshua Bone's "Combination Lock", which were quite clever and only practical to solve in the level editor. But these levels also need to be made in moderation. try to be clever, not difficult. trying to be hard will bring nothing. I am glad that this only has 4 choices, but this level build has been used in about 125 levels, and that's just a guess. This isn't what we want to see. So please try to make other types of levels, as guessing levels rarely are good. |
| « home | Browsing all comments (spoilers: on | off) (info: on | off) | next 25 |