Thread: "Edging closer to a physical 4D puzzle"

--------------000007080909080805070702
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

I wonder if it would be possible to build something that would exist
superimposed upon some kind of frame. It might be a little unwieldy,
but I'm trying to picture a physical structure where you could rotate an
outer face into the center with everything else moving properly. Maybe
someone can figure out how to build something that looks like this
puzzle where it would be possible to push the "front" face into the
center and have all the correct motions happen elsewhere. The invisible
face could be an additional layer distributed around the outsides of the
puzzle. Like many of us, I've spent a little time here and there
thinking about how one could build a physical version of this puzzle,
but until now, I never thought about building into a superstructure
rather than having it be self-contained. Self-contained would be much
nicer, but I bet the problem of building it in superstructure would be
more easily solvable and may provide additional insight as to whether a
physical self-contained structure is possible.

--Jay (momentarily emerging from lurker mode)

On 04/06/2013 10:40 PM, Melinda Green wrote:
>
>
> I've always assumed that a true physical 4D puzzle would have to offer
> only a small subset of possible twists much like Don and my very first
> version of MC4D just about exactly 25 years ago, come to think of it.
> Happy 25th birthday MC4D!! I think that in that very first
> implementation you could only perform 90 twists and even then only on
> the center face plus along each of the 6 outer faces axis that
> intersects the center face. In other words, only those transforms that
> did not distort any of the pieces during the twists.
>
> For a true physical 4D cube you don't even need those outer 6 twists.
> So long as you first rotate your face-of-interest into the center, you
> can then perform all the twists you like on that face without
> distortions. I just have no idea what sort of construction would allow
> the rotations. Maybe something involving a squishy material like latex
> or something that can stretch a lot without breaking. I can almost
> imagine some latex webbing that stretches between the arms of the
> Roadblock faces. I think I'll ask Oskar to think about this problem
> too. If anyone can figure this out, he seems like the one.
>
> -Melinda
>
> On 4/6/2013 2:09 PM, Ray Zhao wrote:
>> No wonder I thought I've seen something similar before (didn't think
>> of the 2^4 when I first saw Oskar's puzzle)...There just needs to be
>> some eighth cell and the 2x2x2s still have to be able to rotate
>> individually in all 3 axes. That'll take a while =P
>> When a physical 4D puzzle is made, will it have the
>> Schlegel-diagram-like cell-centered view, since if that's the case
>> then it will be hard to turn the "inner/far" cell...
>
>


--------------000007080909080805070702
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit



http-equiv="Content-Type">


I wonder if it would be possible to build something that would exist
superimposed upon some kind of frame.  It might be a little
unwieldy, but I'm trying to picture a physical structure where you
could rotate an outer face into the center with everything else
moving properly.  Maybe someone can figure out how to build
something that looks like this puzzle where it would be possible to
push the "front" face into the center and have all the correct
motions happen elsewhere.  The invisible face could be an additional
layer distributed around the outsides of the puzzle.  Like many of
us, I've spent a little time here and there thinking about how one
could build a physical version of this puzzle, but until now, I
never thought about building into a superstructure rather than
having it be self-contained.  Self-contained would be much nicer,
but I bet the problem of building it in superstructure would be more
easily solvable and may provide additional insight as to whether a
physical self-contained structure is possible.



--Jay (momentarily emerging from lurker mode)

 

I've always assumed that a true physical 4D puzzle would
have to offer only a small subset of possible twists much
like Don and my very first version of MC4D just about
exactly 25 years ago, come to think of it. Happy 25th
birthday MC4D!! I think that in that very first
implementation you could only perform 90 twists and even
then only on the center face plus along each of the 6
outer faces axis that intersects the center face. In other
words, only those transforms that did not distort any of
the pieces during the twists.



For a true physical 4D cube you don't even need those
outer 6 twists. So long as you first rotate your
face-of-interest into the center, you can then perform all
the twists you like on that face without distortions. I
just have no idea what sort of construction would allow
the rotations. Maybe something involving a squishy
material like latex or something that can stretch a lot
without breaking. I can almost imagine some latex webbing
that stretches between the arms of the Roadblock faces. I
think I'll ask Oskar to think about this problem too. If
anyone can figure this out, he seems like the one.



-Melinda

On 4/6/2013 2:09 PM, Ray Zhao
wrote:

cite="mid:CAOFi9ucPgm5bGDVgwweazFQU-9vG3DB3Jumub86=exYhFOG6XA@mail.gmail.com"
type="cite">

No wonder I thought I've seen something
similar before (didn't think of the 2^4 when I first saw
Oskar's puzzle)...There just needs to be some eighth
cell and the 2x2x2s still have to be able to rotate
individually in all 3 axes. That'll take a while =P clear="all">

When a physical 4D puzzle is made, will it have the
Schlegel-diagram-like cell-centered view, since if
that's the case then it will be hard to turn the
"inner/far" cell...

--------------020403060003070002070500
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Hello Jay,

Great to hear from you again! For everyone else, I need to point out
that Jay has played a key role in producing MC4D. Don Hatch and I had
produced two versions by then. The first was for a specialized graphics
supercomputer that didn't do well, and the second was for Windows. That
software had a reasonably good design but was still sort of messy stuff
you get when you are rapidly prototyping. Worse, we didn't have shared
version control, or the time or desire to do the critical-but-unglorious
work needed to make it maintainable and portable. In short, Don and I
had hit a roadblock. Then Jay appeared and took the number 2 slot in the
HOF and offered to help us with the dirty work needed to turn our code
into maintainable shape in a form that we would not be completely
embarrassed for other people to see and contribute to. We will be
forever grateful for his selfless help.

Regarding your visualization of a mechanical version, your idea of the
rotations is exactly what I'm imagining, and your idea for a scaffolding
may be the key to making it work. I'm going to think about mechanisms
based on your idea. I just know that a physical 4D Rubik's cube is
possible, and you may have moved it's development a big step closer.

I hope that you post more often but am happy to know that you are
lurking and thinking about this esoteric subject that we all love.
Thanks again for your valuable contributions, Jay!
-Melinda

On 4/8/2013 6:56 AM, Jay Berkenbilt wrote:
>
>
> I wonder if it would be possible to build something that would exist
> superimposed upon some kind of frame. It might be a little unwieldy,
> but I'm trying to picture a physical structure where you could rotate
> an outer face into the center with everything else moving properly.
> Maybe someone can figure out how to build something that looks like
> this puzzle where it would be possible to push the "front" face into
> the center and have all the correct motions happen elsewhere. The
> invisible face could be an additional layer distributed around the
> outsides of the puzzle. Like many of us, I've spent a little time
> here and there thinking about how one could build a physical version
> of this puzzle, but until now, I never thought about building into a
> superstructure rather than having it be self-contained.
> Self-contained would be much nicer, but I bet the problem of building
> it in superstructure would be more easily solvable and may provide
> additional insight as to whether a physical self-contained structure
> is possible.
>
> --Jay (momentarily emerging from lurker mode)
>
> On 04/06/2013 10:40 PM, Melinda Green wrote:
>>
>> I've always assumed that a true physical 4D puzzle would have to
>> offer only a small subset of possible twists much like Don and my
>> very first version of MC4D just about exactly 25 years ago, come to
>> think of it. Happy 25th birthday MC4D!! I think that in that very
>> first implementation you could only perform 90 twists and even then
>> only on the center face plus along each of the 6 outer faces axis
>> that intersects the center face. In other words, only those
>> transforms that did not distort any of the pieces during the twists.
>>
>> For a true physical 4D cube you don't even need those outer 6 twists.
>> So long as you first rotate your face-of-interest into the center,
>> you can then perform all the twists you like on that face without
>> distortions. I just have no idea what sort of construction would
>> allow the rotations. Maybe something involving a squishy material
>> like latex or something that can stretch a lot without breaking. I
>> can almost imagine some latex webbing that stretches between the arms
>> of the Roadblock faces. I think I'll ask Oskar to think about this
>> problem too. If anyone can figure this out, he seems like the one.
>>
>> -Melinda
>>
>> On 4/6/2013 2:09 PM, Ray Zhao wrote:
>>> No wonder I thought I've seen something similar before (didn't think
>>> of the 2^4 when I first saw Oskar's puzzle)...There just needs to be
>>> some eighth cell and the 2x2x2s still have to be able to rotate
>>> individually in all 3 axes. That'll take a while =P
>>> When a physical 4D puzzle is made, will it have the
>>> Schlegel-diagram-like cell-centered view, since if that's the case
>>> then it will be hard to turn the "inner/far" cell...
>>
>
>
>
>


--------------020403060003070002070500
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit



http-equiv="Content-Type">


Hello Jay,



Great to hear from you again! For everyone else, I need to point out
that Jay has played a key role in producing MC4D. Don Hatch and I
had produced two versions by then. The first was for a specialized
graphics supercomputer that didn't do well, and the second was for
Windows. That software had a reasonably good design but was still
sort of messy stuff you get when you are rapidly prototyping. Worse,
we didn't have shared version control, or the time or desire to do
the critical-but-unglorious work needed to make it maintainable and
portable. In short, Don and I had hit a roadblock. Then Jay appeared
and took the number 2 slot in the HOF and offered to help us with
the dirty work needed to turn our code into maintainable shape in a
form that we would not be completely embarrassed for other people to
see and contribute to. We will be forever grateful for his selfless
help.



Regarding your visualization of a mechanical version, your idea of
the rotations is exactly what I'm imagining, and your idea for a
scaffolding may be the key to making it work. I'm going to think
about mechanisms based on your idea. I just know that a physical 4D
Rubik's cube is possible, and you may have moved it's development a
big step closer.



I hope that you post more often but am happy to know that you are
lurking and thinking about this esoteric subject that we all love.
Thanks again for your valuable contributions, Jay!

-Melinda

http-equiv="Content-Type">
I wonder if it would be
possible to build something that would exist superimposed upon
some kind of frame.  It might be a little unwieldy, but I'm trying
to picture a physical structure where you could rotate an outer
face into the center with everything else moving properly.  Maybe
someone can figure out how to build something that looks like this
puzzle where it would be possible to push the "front" face into
the center and have all the correct motions happen elsewhere.  The
invisible face could be an additional layer distributed around the
outsides of the puzzle.  Like many of us, I've spent a little time
here and there thinking about how one could build a physical
version of this puzzle, but until now, I never thought about
building into a superstructure rather than having it be
self-contained.  Self-contained would be much nicer, but I bet the
problem of building it in superstructure would be more easily
solvable and may provide additional insight as to whether a
physical self-contained structure is possible.



--Jay (momentarily emerging from lurker mode)

On 04/06/2013 10:40 PM, Melinda Green
wrote:

type="cite">  

I've always assumed that a true physical 4D puzzle would
have to offer only a small subset of possible twists much
like Don and my very first version of MC4D just about
exactly 25 years ago, come to think of it. Happy 25th
birthday MC4D!! I think that in that very first
implementation you could only perform 90 twists and even
then only on the center face plus along each of the 6 outer
faces axis that intersects the center face. In other words,
only those transforms that did not distort any of the pieces
during the twists.



For a true physical 4D cube you don't even need those outer
6 twists. So long as you first rotate your face-of-interest
into the center, you can then perform all the twists you
like on that face without distortions. I just have no idea
what sort of construction would allow the rotations. Maybe
something involving a squishy material like latex or
something that can stretch a lot without breaking. I can
almost imagine some latex webbing that stretches between the
arms of the Roadblock faces. I think I'll ask Oskar to think
about this problem too. If anyone can figure this out, he
seems like the one.



-Melinda

On 4/6/2013 2:09 PM, Ray Zhao
wrote:

cite="mid:CAOFi9ucPgm5bGDVgwweazFQU-9vG3DB3Jumub86=exYhFOG6XA@mail.gmail.com"
type="cite">

No wonder I thought I've seen something
similar before (didn't think of the 2^4 when I first saw
Oskar's puzzle)...There just needs to be some eighth cell
and the 2x2x2s still have to be able to rotate
individually in all 3 axes. That'll take a while =P clear="all">

When a physical 4D puzzle is made, will it have the
Schlegel-diagram-like cell-centered view, since if
that's the case then it will be hard to turn the
"inner/far" cell...