icculus.org
[svn] - [hg] - [git] - [bugzilla] - [ssl] - [mail]

News archive: (submit news)

• Notable .plan update from chunky (posted 2002-05-08 07:00:04 by fingermaster):
chunky updated his .plan file at 2002-05-07 06:41:37 with the following info:

```[note: old plan entries are all avilable at
http://icculus.org/~chunky/oldplan]

[2002-05-07]

http://www.mefferts.com/images/puzzles/new/megaminx-l.jpg

Notes on this cube:
1) It's not a "cube", I just think of it as one.
2) It's not as complicated as it looks. It's actually exactly the same
as a normal Rubik's Cube, except for one [just one!] difference:
Instead of having 4 sides on each face, this has 5. As a natural
progression to end with a platonic solid, you complete the exercise with
a dodecahedron.
As a result, almost all of the normal cube-like moves actually work
here, also. You just have to imagine that the 3 faces you mess with
are perpendicular instead of vaguely perpendicular, and everything just
slots into place.

Hmmm. Anyone who can solve a vanilla Rubik's Cube [3x3x3], you can
[probably] solve this already, depending on how you solve a normal one. If
you do it a slice at a time, this is almost completely the same. You can
solve the whole cube except the bottom slice using just your intuition
and the move that takes a piece from the bottom to the side, without
munging the rest of the center slice, or the stuff above it.

Except one thing that was bugging the living f**k out of me.

What that picture doesn't show you is that the colors on opposite sides
are exactly the same. Hence, the edge-middle pieces have /two/ of each
color pair, and the corner pieces have two of each color group [eg, RGB]
but are handed, hence, unique. For example, going clockwise around the
piece, there's one Corner that goes Green, Red, Blue, and there's another
that goes Green, Blue, Red.

Since the centers never move:
1) The centers are effectively different colors WRT each other
2) The corners are all unique
3) There are two of each side piece.

Which means that
4) Sometimes, you'll get to what looks like an insoluble cube. It is. You
need to start again.

Which was my mental block. But it's also a pretty daft design, vaguely;
you /can't/ garauntee that you're solving the thing correctly; as opposed
to other normal cubes, where you're garaunteed to have the correct piece
if you just look at the colors. Or it doesn't matter if you have the
wrong pieces, so long as they're the right color [like the 5x5x5 cube,
which has groups of 4 indistinguishable pieces, but they can be in any
order when you place them].

Anyways. I'm now off to try to make pretty patterns on it; solving
Rubik's cubes is generally "simple", wheras making pretty patterns all
over them is
1) harder
2) more interesting
3) more time-consuming
```

--fingermaster.

[ email newsmaster | get news rdf ]