Message #1889

From: schuma <>
Subject: Re: yet more new puzzles and a prize
Date: Sat, 05 Nov 2011 01:00:22 -0000

Hi guys,

I finally got a computer that can run MagicTile v2 preview. Knowing that the challenge is still open, I looked into the three {3,7} puzzles. I wonder if anyone is attempting these puzzles. After spending several minutes playing, I have the following thoughts.

VT is obviously the easiest. The main difficulty is that, there are too many colors. 56 colors, right? Sometimes I’m embarrassed because I can’t find the piece I need. Sometimes I thought I found a desired piece. But only when I moved it to the destination, I found the colors were actually different. Because of this difficulty, the puzzle is not as easy as it should be. But still, it’s not hard. It just takes a little bit more time. By the way, under the default color setting, there are two blue triangle (color 3 and color 33). They are identical colors and one should change one of them.

ET is something serious. I quickly found all the 3-cycle algorithms. They are pretty short and easy to find. As far as I currently know, the difficulties may be:
(1) Too many colors. Hard to find a piece.
(2) Too many pieces, nearly 1000 pieces. Tedious.
(3) Hidden global orientation. This is tricky. In the beginning one should not assume a triangle to be the white triangle and start solving. He/she needs to study the scrambled state, and deduce which triangle is the "white" triangle, which is the blue one etc. Then he/she can start solving based on the deduced "orientation". If you don’t do it in the beginning and make an arbitrary assumption, it’s probably wrong and you are going into a dead end. I met this issue on edge turning dodecahedra, e.g., Gelatinbrain 1.4.10. Brandon must know what I’m talking about.

FT is even harder, because the 3-cycle algorithms seem to be longer than ET. So macros would be very helpful here. At this point I can’t tell if I’m patient enough to solve it or not. I guess one need approximately or more than 20 hours to solve it. I won’t attempt it before solving VT and ET.


— In, Roice Nelson <roice3@…> wrote:
> Hi all,
> MagicTile v2 has been coming along well lately, so I wanted to share a
> preview (not that we need any more puzzles at the moment!). It is a big
> user experience improvement over v1, mostly due to my favorite new
> capability - *smooth hyperbolic panning*! Click and drag to your heart’s
> content.
> Not a huge number of puzzles in the menus yet, only the ones I’ve been
> distracted towards or used in development, but they are unique. Also, they
> are configured in xml and easy to set up, so if anyone would like to see
> something in particular, just let me know. I’m finding the most difficult
> (but perhaps most interesting) task is working out the coloring patterns -
> that is, how to identify cells with each other. So if you request a puzzle,
> I may turn around and ask for help like I did for the {8,3} tiling.
> Here’s what is available so far:
> - {3,7} face/edge/vertex turning
> - {8,3} 24 Color, 10 Color and alternate 12 Color (The last two surprised
> me and the 10 Color is especially pretty.)
> - {4,6} and {6,4} "Petals"
> - {5,5} - These are seriously strange… non-orientable, 6-colors.
> Should be awesome, so check them out!
> - {4,4} Variants
> - Klein Bottles - There’s a Nan/Brandon/Andrey difficulty-level
> puzzle that is both face+edge turning :)
> - {6,3} - A 7-color and a vertex turning 16-color.
> - Also nice is that you can no longer pan to the edge of the euclidean
> tilings - they will just go on forever.
> I like that Melinda gave out a prize in the speed cubing competition, so I’d
> like to offer to send a MC4D crystal
> cube<>to
> *the first person to solve all three of the {3,7} puzzles*. I know there
> are 115 stickers that move through a twist in the face turning puzzle, some
> of them tiny. I know there are a ton of colors. I know there are no macros
> or piece finding yet. I only offered the prize, and didn’t say it would be
> easy! I hope the challenge will hold the group at bay for a while, maybe
> until new features are implemented, maybe forever. Yeah right! :D
> Here are the files (I haven’t made an installer). It requires the .NET
> Framework 4<>,
> but I’d only worry about that if it doesn’t work after simply unzipping and
> attempting the run the exe.
> Cheers,
> Roice