Message #2047

From: Roice Nelson <>
Subject: Re: [MC4D] Re: 4D-interactive puzzles in MagicTile
Date: Tue, 06 Mar 2012 11:51:51 -0600

Thanks for your thoughts and positive feedback Andrey :)

That’s a really cool observation about edge orientation behavior!
Also, about the duoprisms being planar tori, I wanted to mention that the
skew polyhedra are only a different view of the traditional puzzles in
MagicTile. For the duoprisms, you can switch to the {4,4} planar tori view
by setting the option "Show as Skew" to false. Changing this setting for
the bitruncated simplex will show its associated hyperbolic tiling too.

About the twisting direction relationship to the side you click on, I bet
it can be changed without too much effort like you say, so I will plan to
do this relatively soon.

aside: I did mentally debate whether the current behavior was ok or not,
and the same thing happens on all of the IRP puzzles as well. Besides
easier coding, one advantage is that it gives a visual clue of the two
different halves the skew polyhedra divide space into. A similar sort of
thing came up before with non-orientable tilings of the plane - mirrored
faces were twisting in an opposite sense. That had advantage (visual clues
to the topology), but the same disadvantage of making solving difficult. We
settled on making all faces always twist CCW when left-clicked, and it
sounds like I should take the same path here. I would like to make solving
as pleasant as possible for the skews.


On Tue, Mar 6, 2012 at 10:03 AM, Andrey <> wrote:

> Roice,
> New puzzles are beautiful! It’s very nice to see these sceletons of our
> familiar 4D bodies and wander around them :)
> I solved only one puzzle so far - bitruncated simplex (my favorite
> polytope :) ) in F:0:0:1 slicing. It was not very easy, and one of main
> problems is that twising rotation depends on side of face that you click:
> if you see one side, face twists clockwise, but if you go to the other side
> and make the same click, it twists counterclockwise. I think that it’s not
> difficult to fix.
> This puzzle has a kind of "global non-orientability": if you have wrong
> oriented edge, you can’t reorient it by moving around the vertex, you have
> to make a loop around a tube. I guess that same effects will be in 5x5 and
> 7x7 duoprisms (that are actually just another implementations of planar
> tori :) )
> Thank you for these puzzles!
> Andrey