Message #1937

From: schuma <mananself@gmail.com>
Subject: 4D puzzles in Gelatinbrain’s applet
Date: Sun, 27 Nov 2011 02:06:04 -0000

Hi all,

As some of you may know that Gelatinbrain(GB)’s applet contains hundreds of 3D puzzles. Now two 4D puzzles have been added. To see them, please go to

http://users.skynet.be/gelatinbrain/Applets/Magic%20Polyhedra/dodeca_f0.htm

then wait for it to load (it may take two or three minutes). Then switch puzzles by clicking the "File" button on the top left corner, then following New > tessaract to find 8.1.1 or 8.1.2.

8.1.1 is 2x2x2x2. 8.1.2 is the deep-cut (half-cut) RT hypercube, which can also be found in MPUlt 1.09 as 4D_half_RT.

There have been discussions between the GB solvers and GB here:
http://twistypuzzles.com/forum/viewtopic.php?p=271270#p271270

Neither of them has been solved in that applet yet, according to the ranking page. I’m not comfortable solving the 2^4 without using macros -_-;

There are some unique features in GB that’s not there in MC4D or MPUlt.

(1) It has two views. So the hidden cell is directly visible.
(2) The global orientation of the puzzle is automatically snapped to some fixed angles. The rigid viewpoint has pros and cons. For example, if the macro function becomes available (GB is working on it), one can apply macros without specifying the reference point. This is less flexible but pretty convenient.
(3) When a puzzle is solved, the move counts and time are automatically submitted to a scoreboard.

There are also differences between GB’s puzzles and the corresponding existing puzzles here. As Brandon pointed out, the moves in GB are limited to 90-degree turns only.
(1) In MC4D one may click on edges or corners to do 120 deg or 180 deg turns. But in GB 8.1.1 these moves not allowed.
(2) In MPUlt > 4D_half_RT, when turning around the face between cells A and B, one can use a single move to swap the whole cells A and B, affecting some other cells. This kind of moves are not allowed in GB 8.1.2. The 90 deg turns only rotate cells A and B without swapping them.

In spite of the limitation, GB 8.1.1 reaches all the states of the "fully functional" 2^4; and GB 8.1.2 reaches all the states of 4D_half_RT in MPUlt. The latter statement is not straightforward but is true, because using only 90 deg turns one can do arbitrary 3-cycles.

In conclusion, we have a new software implementing 4D puzzles. The author, Gelatinbrain, is also a talented and diligent programmer. I would expect more 4D puzzles coming in GB’s applet.

Nan