Thread: "New multydimensional puzzle"

http://games.groups.yahoo.com/group/4D_Cubing/files/MC7D/MPUlt.zip

Finally =96 version 0.1 of "Magic Puzzle Ultimate" is ready. It has no macr=
os and
search features, and it is not tested at all. But it supports puzzles with
symmetric bodies of arbitrary dimensions, with cell-, ridge-, edge- and
vertex-centered cutting planes (for 4D) and cell- and vertex-centered cutti=
ng
planes for greater dimensions (and this limitation is caused only by twist
selection interface).
For every puzzle its description is in the text file. Now there are files f=
or
the following puzzles:
3d4.txt =96 3^4
5d.txt =96 3^5
4d_rts.txt =96 Ridge-twisting 4D cube
4d_ets.txt =96 Edge-twisting 4D cube
4d_diag.txt =96 Corner-twisting 4D cube (cutting planes go through the cent=
er)
5dd.txt =96 Shallow-cut corner-twisting 5D cube
16cell-fts.txt =96 Cell-twisting 16cell
24cell-fts.txt =96 Cell-twisting 24cell
2tetr.txt =96 Cell-twisting duotetrahedron (6D puzzle).

"Puzzle" item in the main menu asks for the description file and generates =
the
puzzle from it (or creates 3^4 if the description is not valid).
Navigation:
Left button =96 rotation of 3D projection
Right button =96 XY rotation and 4D sliding
Shift-Left button =96 4D rotation
Ctrl-Left button =96 Zoom in/out
"Switch dim" button on the control pane =96 switch direction of the "depth"
coordinate (it works for 5D and more). It doesn't change 2D projection, but
depth order of objects is changing, so rotation in 3D will be different.

Twists:
Cell-twisting in 4D is the same as in MC4D. But rotation axis is defined no=
t by
sticker under the cursor, but by the point on the cell surface (it is outli=
ned
by thin lines). If the puzzle has other kinds of twisting, click both butto=
ns
(left+right) somewhere on the face. Program finds the closest axis of 4D
rotation and outlines stickers in the outer layer of this twist direction. =
Then
you click some outlined sticker, then sticker where is should go, and so on=
.
Sometimes program may suggest starting another stickers' sequence. When the=
re is
enough information for the twist, program performs it. Left+right click in =
the
empty space repeats last twist (it's good when you rotate face to more than
1/N).
Click (left of right) in the face without twisting axis works in the same w=
ay
as left+right click.
In 5D and more only cell twists and vertex-centered twists are possible. Yo=
u
start with some corner sticker. Left click will start stickers' sequence fo=
r
cell twist, and right click defines vertex-centered twist. Again, you see
possible destinations for the selected sticker, click one of them, then nex=
t and
so on.

To change color of the face, ctrl-right-click any sticker of the "old" colo=
r and select "Edit Color" item in the menu.

There was almost no testing of the program =96 only one solve of 3^4. So al=
l bug
reports are welcome!
If you have some puzzle that you want to try, just ask about it =96 I'll tr=
y to
code it's description ))

Good luck!
Andrey

--00032555a3da2c7ddb04a1ad1d98
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

very nice Andrey :)

At first I thought to send issues I run across directly to you, but figure
it will help limit you getting double reports if I go ahead and post them
here.

Twisting on 4D puzzles:
- Outlines don't rotate with stickers
- Stickers that become mirrored during a twist don't disappear until the
end. It'd be nice if they disappeared as soon as they became mirrored.
Likewise, stickers which are mirrored at the beginning of the twist only
pop into existence at the end.

Crash in 3^5
- I'm not able to get a completely reliable reproduction, but if you do a
little right-click dragging, then left or right click a sticker, it
consistently crashes.

I like that the puzzles come from configuration files (and I've followed
your lead in this with MagicTile v2). I've only had MagicPuzzleUltimate
open for a few minutes so far, but look forward to playing with it more.

Cheers,
Roice


On Sun, Apr 24, 2011 at 10:42 AM, Andrey wrote:

> http://games.groups.yahoo.com/group/4D_Cubing/files/MC7D/MPUlt.zip
>
> Finally =96 version 0.1 of "Magic Puzzle Ultimate" is ready. It has no ma=
cros
> and
> search features, and it is not tested at all. But it supports puzzles wit=
h
> symmetric bodies of arbitrary dimensions, with cell-, ridge-, edge- and
> vertex-centered cutting planes (for 4D) and cell- and vertex-centered
> cutting
> planes for greater dimensions (and this limitation is caused only by twis=
t
> selection interface).
> For every puzzle its description is in the text file. Now there are files
> for
> the following puzzles:
> 3d4.txt =96 3^4
> 5d.txt =96 3^5
> 4d_rts.txt =96 Ridge-twisting 4D cube
> 4d_ets.txt =96 Edge-twisting 4D cube
> 4d_diag.txt =96 Corner-twisting 4D cube (cutting planes go through the
> center)
> 5dd.txt =96 Shallow-cut corner-twisting 5D cube
> 16cell-fts.txt =96 Cell-twisting 16cell
> 24cell-fts.txt =96 Cell-twisting 24cell
> 2tetr.txt =96 Cell-twisting duotetrahedron (6D puzzle).
>
> "Puzzle" item in the main menu asks for the description file and generate=
s
> the
> puzzle from it (or creates 3^4 if the description is not valid).
> Navigation:
> Left button =96 rotation of 3D projection
> Right button =96 XY rotation and 4D sliding
> Shift-Left button =96 4D rotation
> Ctrl-Left button =96 Zoom in/out
> "Switch dim" button on the control pane =96 switch direction of the "dept=
h"
> coordinate (it works for 5D and more). It doesn't change 2D projection, b=
ut
> depth order of objects is changing, so rotation in 3D will be different.
>
> Twists:
> Cell-twisting in 4D is the same as in MC4D. But rotation axis is defined
> not by
> sticker under the cursor, but by the point on the cell surface (it is
> outlined
> by thin lines). If the puzzle has other kinds of twisting, click both
> buttons
> (left+right) somewhere on the face. Program finds the closest axis of 4D
> rotation and outlines stickers in the outer layer of this twist direction=
.
> Then
> you click some outlined sticker, then sticker where is should go, and so
> on.
> Sometimes program may suggest starting another stickers' sequence. When
> there is
> enough information for the twist, program performs it. Left+right click i=
n
> the
> empty space repeats last twist (it's good when you rotate face to more th=
an
> 1/N).
> Click (left of right) in the face without twisting axis works in the same
> way
> as left+right click.
> In 5D and more only cell twists and vertex-centered twists are possible.
> You
> start with some corner sticker. Left click will start stickers' sequence
> for
> cell twist, and right click defines vertex-centered twist. Again, you see
> possible destinations for the selected sticker, click one of them, then
> next and
> so on.
>
> To change color of the face, ctrl-right-click any sticker of the "old"
> color and select "Edit Color" item in the menu.
>
> There was almost no testing of the program =96 only one solve of 3^4. So =
all
> bug
> reports are welcome!
> If you have some puzzle that you want to try, just ask about it =96 I'll =
try
> to
> code it's description ))
>
> Good luck!
> Andrey
>
>
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>
>

--00032555a3da2c7ddb04a1ad1d98
Content-Type: text/html; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

very nice Andrey :)

--0016e6dd993adbb38404a1ad5a5b
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

A little more on the "outlines don't rotate with stickers" thought... What
I am imagining is to do something like MagicTile does (with the yellow cell
outlines there), but it might be more difficult to code in this program.
The difficulty is that only portions of certain cells get moved during a
twist, so outlines need to break apart and come back together, whereas I
figure you have some code right now which draws the polytope cell boundarie=
s
without needing to worry about such complications. What you have feels lik=
e
legitimate alternate behavior, and so I'd consider this a low-priority
request for sure.

seeya,
Roice


On Sun, Apr 24, 2011 at 12:06 PM, Roice Nelson wrote:

> very nice Andrey :)
>
> At first I thought to send issues I run across directly to you, but figur=
e
> it will help limit you getting double reports if I go ahead and post them
> here.
>
> Twisting on 4D puzzles:
> - Outlines don't rotate with stickers
> - Stickers that become mirrored during a twist don't disappear until the
> end. It'd be nice if they disappeared as soon as they became mirrored.
> Likewise, stickers which are mirrored at the beginning of the twist only
> pop into existence at the end.
>
> Crash in 3^5
> - I'm not able to get a completely reliable reproduction, but if you do a
> little right-click dragging, then left or right click a sticker, it
> consistently crashes.
>
> I like that the puzzles come from configuration files (and I've followed
> your lead in this with MagicTile v2). I've only had MagicPuzzleUltimate
> open for a few minutes so far, but look forward to playing with it more.
>
> Cheers,
> Roice
>
>
> On Sun, Apr 24, 2011 at 10:42 AM, Andrey wrote=
:
>
>> http://games.groups.yahoo.com/group/4D_Cubing/files/MC7D/MPUlt.zip
>>
>> Finally =96 version 0.1 of "Magic Puzzle Ultimate" is ready. It has no
>> macros and
>> search features, and it is not tested at all. But it supports puzzles wi=
th
>> symmetric bodies of arbitrary dimensions, with cell-, ridge-, edge- and
>> vertex-centered cutting planes (for 4D) and cell- and vertex-centered
>> cutting
>> planes for greater dimensions (and this limitation is caused only by twi=
st
>> selection interface).
>> For every puzzle its description is in the text file. Now there are file=
s
>> for
>> the following puzzles:
>> 3d4.txt =96 3^4
>> 5d.txt =96 3^5
>> 4d_rts.txt =96 Ridge-twisting 4D cube
>> 4d_ets.txt =96 Edge-twisting 4D cube
>> 4d_diag.txt =96 Corner-twisting 4D cube (cutting planes go through the
>> center)
>> 5dd.txt =96 Shallow-cut corner-twisting 5D cube
>> 16cell-fts.txt =96 Cell-twisting 16cell
>> 24cell-fts.txt =96 Cell-twisting 24cell
>> 2tetr.txt =96 Cell-twisting duotetrahedron (6D puzzle).
>>
>> "Puzzle" item in the main menu asks for the description file and generat=
es
>> the
>> puzzle from it (or creates 3^4 if the description is not valid).
>> Navigation:
>> Left button =96 rotation of 3D projection
>> Right button =96 XY rotation and 4D sliding
>> Shift-Left button =96 4D rotation
>> Ctrl-Left button =96 Zoom in/out
>> "Switch dim" button on the control pane =96 switch direction of the "dep=
th"
>> coordinate (it works for 5D and more). It doesn't change 2D projection,
>> but
>> depth order of objects is changing, so rotation in 3D will be different.
>>
>> Twists:
>> Cell-twisting in 4D is the same as in MC4D. But rotation axis is defined
>> not by
>> sticker under the cursor, but by the point on the cell surface (it is
>> outlined
>> by thin lines). If the puzzle has other kinds of twisting, click both
>> buttons
>> (left+right) somewhere on the face. Program finds the closest axis of 4D
>> rotation and outlines stickers in the outer layer of this twist directio=
n.
>> Then
>> you click some outlined sticker, then sticker where is should go, and so
>> on.
>> Sometimes program may suggest starting another stickers' sequence. When
>> there is
>> enough information for the twist, program performs it. Left+right click =
in
>> the
>> empty space repeats last twist (it's good when you rotate face to more
>> than
>> 1/N).
>> Click (left of right) in the face without twisting axis works in the sam=
e
>> way
>> as left+right click.
>> In 5D and more only cell twists and vertex-centered twists are possible.
>> You
>> start with some corner sticker. Left click will start stickers' sequence
>> for
>> cell twist, and right click defines vertex-centered twist. Again, you se=
e
>> possible destinations for the selected sticker, click one of them, then
>> next and
>> so on.
>>
>> To change color of the face, ctrl-right-click any sticker of the "old"
>> color and select "Edit Color" item in the menu.
>>
>> There was almost no testing of the program =96 only one solve of 3^4. So=
all
>> bug
>> reports are welcome!
>> If you have some puzzle that you want to try, just ask about it =96 I'll=
try
>> to
>> code it's description ))
>>
>> Good luck!
>> Andrey
>>
>>
>>
>>
>> ------------------------------------
>>
>> Yahoo! Groups Links
>>
>>
>>
>>
>

--0016e6dd993adbb38404a1ad5a5b
Content-Type: text/html; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

A little more on the "outlines don't rotate with stickers" th=
ought... =A0What I am imagining is to do something like MagicTile does (wit=
h the yellow cell outlines there), but it might be more difficult to code i=
n this program. =A0The difficulty is that only portions of certain cells ge=
t moved during a twist, so outlines need to break apart and come back toget=
her, whereas I figure you have some code right now which draws the polytope=
cell boundaries without needing to worry about such complications. =A0What=
you have feels like legitimate alternate behavior, and so I'd consider=
this a low-priority request for sure.

Hi Andrey,

What a great achievement! I can't wait analyzing the 16-cell and
24-cell, although the first step is to recall the geometry of them.

I found a bug as follows:

When I hold a number key "n" and rotate something, it rotate the n-th
layer. But if n is larger than the largest available layer (for
example, try 4+"left click" in the 16 cell), a dialog box pops up
saying "unhandled exception has occurred in your application. If you
click continue, the application will ignore this error and attempt to
continue. If you click quit, the application will close immediately.
Index was outside the bounds of the array." I believe this is easily
fixable.

Nan

Nan, Roice,
Thank you for feedbacks!
I've fixed bugs with right-click in 5D puzzle (attempt to twist non-exist=
ent vertex-centered layer) and issue with buttons outside of number of laye=
rs - program is uploaded to the same place. Hiding of mirrored blocks durin=
g animation is in my todo list, and rotation of cell outline has very low p=
riority (probably zero). There are some things that I have to do: timer, tw=
ist counter, optimization and recalculation, testing of log save/load comma=
nds, library file for puzzles (instead of many separate txt files), computa=
tion of main symmetry groups for 4D, and so on...

Good luck!
Andrey



--- In 4D_Cubing@yahoogroups.com, Nan Ma wrote:
>
> Hi Andrey,
>=20
> What a great achievement! I can't wait analyzing the 16-cell and
> 24-cell, although the first step is to recall the geometry of them.
>=20
> I found a bug as follows:
>=20
> When I hold a number key "n" and rotate something, it rotate the n-th
> layer. But if n is larger than the largest available layer (for
> example, try 4+"left click" in the 16 cell), a dialog box pops up
> saying "unhandled exception has occurred in your application. If you
> click continue, the application will ignore this error and attempt to
> continue. If you click quit, the application will close immediately.
> Index was outside the bounds of the array." I believe this is easily
> fixable.
>=20
> Nan
>

--Apple-Mail-1--1042012612
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
charset=utf-8

How did you know it was my birthday today Andrey?!

(seriously it is - 21st)

I am unfortunately at my lakehouse in the middle of nowhere and do not have=
a computer with Internet access so I can't look at what sounds like an inc=
redible program quite yet, but I am dying to test it out!

Thank you so much for creating these. I suspect the 24Cell will be my all t=
ime favorite puzzle ;)

-Matt Galla

Sent from my iPhone

On Apr 24, 2011, at 2:21 PM, Andrey wrote:

> Nan, Roice,
> Thank you for feedbacks!
> I've fixed bugs with right-click in 5D puzzle (attempt to twist non-exist=
ent vertex-centered layer) and issue with buttons outside of number of laye=
rs - program is uploaded to the same place. Hiding of mirrored blocks durin=
g animation is in my todo list, and rotation of cell outline has very low p=
riority (probably zero). There are some things that I have to do: timer, tw=
ist counter, optimization and recalculation, testing of log save/load comma=
nds, library file for puzzles (instead of many separate txt files), computa=
tion of main symmetry groups for 4D, and so on...
>=20
> Good luck!
> Andrey
>=20
> --- In 4D_Cubing@yahoogroups.com, Nan Ma wrote:
> >
> > Hi Andrey,
> >=20
> > What a great achievement! I can't wait analyzing the 16-cell and
> > 24-cell, although the first step is to recall the geometry of them.
> >=20
> > I found a bug as follows:
> >=20
> > When I hold a number key "n" and rotate something, it rotate the n-th
> > layer. But if n is larger than the largest available layer (for
> > example, try 4+"left click" in the 16 cell), a dialog box pops up
> > saying "unhandled exception has occurred in your application. If you
> > click continue, the application will ignore this error and attempt to
> > continue. If you click quit, the application will close immediately.
> > Index was outside the bounds of the array." I believe this is easily
> > fixable.
> >=20
> > Nan
> >
>=20
>=20

--Apple-Mail-1--1042012612
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
charset=utf-8

Hi Andrey,

I was analyzing the pieces of the 16-cell today. I had a lot of fun twistin=
g it. But given the large number of pieces, I don't think I can solve it wi=
thout the macro.=20

I've found a glitch. Sometimes when I click a face of a certain cell, the p=
uzzle is rotating around another different cell. It happens from time to ti=
me. But I cannot find a viewpoint to reliably repeat this glitch. Fortunate=
ly one time I got two screenshots:

Before clicking the red dot, we have the solved 16-cell:

https://picasaweb.google.com/mananself/ForAndrey?authkey=3DGv1sRgCLnZ3p_kpZ=
THTw#5599719882528991298

After clicking the red dot on the dark yellow cell, that cell doesn't rotat=
e:

https://picasaweb.google.com/mananself/ForAndrey?authkey=3DGv1sRgCLnZ3p_kpZ=
THTw#5599719915923903506

The cell that is actually rotated is an invisible one. Whenever this glitch=
occurs, it always rotates around an invisible cell. I hope this clue helps=
you to fix it.

Thanks again for making such a great puzzle!

Nan



--- In 4D_Cubing@yahoogroups.com, "Andrey" wrote:
>
> Nan, Roice,
> Thank you for feedbacks!
> I've fixed bugs with right-click in 5D puzzle (attempt to twist non-exi=
stent vertex-centered layer) and issue with buttons outside of number of la=
yers - program is uploaded to the same place. Hiding of mirrored blocks dur=
ing animation is in my todo list, and rotation of cell outline has very low=
priority (probably zero). There are some things that I have to do: timer, =
twist counter, optimization and recalculation, testing of log save/load com=
mands, library file for puzzles (instead of many separate txt files), compu=
tation of main symmetry groups for 4D, and so on...
>=20
> Good luck!
> Andrey
>=20
>=20
>=20
> --- In 4D_Cubing@yahoogroups.com, Nan Ma wrote:
> >
> > Hi Andrey,
> >=20
> > What a great achievement! I can't wait analyzing the 16-cell and
> > 24-cell, although the first step is to recall the geometry of them.
> >=20
> > I found a bug as follows:
> >=20
> > When I hold a number key "n" and rotate something, it rotate the n-th
> > layer. But if n is larger than the largest available layer (for
> > example, try 4+"left click" in the 16 cell), a dialog box pops up
> > saying "unhandled exception has occurred in your application. If you
> > click continue, the application will ignore this error and attempt to
> > continue. If you click quit, the application will close immediately.
> > Index was outside the bounds of the array." I believe this is easily
> > fixable.
> >=20
> > Nan
> >
>

Hello Andrey,

I'm sorry for being so late with plaudits for your latest amazing bit of=20
programming. I've only played with it for a short time but it is clearly=20
impressive. I especially want to join the other members who are=20
especially happy to see the 16, and especially the 24 cell. The 24 cell=20
is a truly special object on par with the monstrous 120 cell in that way=20
but is much more approachable than its photogenic cousin. I hope that=20
you implement macros before anyone solves these new puzzles. Or have you=20
already solved them?

Thank you for expanding our playground yet again,
-Melinda

On 4/24/2011 8:42 AM, Andrey wrote:
> http://games.groups.yahoo.com/group/4D_Cubing/files/MC7D/MPUlt.zip
>
> Finally =96 version 0.1 of "Magic Puzzle Ultimate" is ready. It has no ma=
cros and
> search features, and it is not tested at all. But it supports puzzles wit=
h
> symmetric bodies of arbitrary dimensions, with cell-, ridge-, edge- and
> vertex-centered cutting planes (for 4D) and cell- and vertex-centered cut=
ting
> planes for greater dimensions (and this limitation is caused only by twis=
t
> selection interface).
> For every puzzle its description is in the text file. Now there are files=
for
> the following puzzles:
> 3d4.txt =96 3^4
> 5d.txt =96 3^5
> 4d_rts.txt =96 Ridge-twisting 4D cube
> 4d_ets.txt =96 Edge-twisting 4D cube
> 4d_diag.txt =96 Corner-twisting 4D cube (cutting planes go through the ce=
nter)
> 5dd.txt =96 Shallow-cut corner-twisting 5D cube
> 16cell-fts.txt =96 Cell-twisting 16cell
> 24cell-fts.txt =96 Cell-twisting 24cell
> 2tetr.txt =96 Cell-twisting duotetrahedron (6D puzzle).
>
> "Puzzle" item in the main menu asks for the description file and generate=
s the
> puzzle from it (or creates 3^4 if the description is not valid).
> Navigation:
> Left button =96 rotation of 3D projection
> Right button =96 XY rotation and 4D sliding
> Shift-Left button =96 4D rotation
> Ctrl-Left button =96 Zoom in/out
> "Switch dim" button on the control pane =96 switch direction of the "dept=
h"
> coordinate (it works for 5D and more). It doesn't change 2D projection, b=
ut
> depth order of objects is changing, so rotation in 3D will be different.
>
> Twists:
> Cell-twisting in 4D is the same as in MC4D. But rotation axis is defined =
not by
> sticker under the cursor, but by the point on the cell surface (it is out=
lined
> by thin lines). If the puzzle has other kinds of twisting, click both but=
tons
> (left+right) somewhere on the face. Program finds the closest axis of 4D
> rotation and outlines stickers in the outer layer of this twist direction=
. Then
> you click some outlined sticker, then sticker where is should go, and so =
on.
> Sometimes program may suggest starting another stickers' sequence. When t=
here is
> enough information for the twist, program performs it. Left+right click i=
n the
> empty space repeats last twist (it's good when you rotate face to more th=
an
> 1/N).
> Click (left of right) in the face without twisting axis works in the same=
way
> as left+right click.
> In 5D and more only cell twists and vertex-centered twists are possible. =
You
> start with some corner sticker. Left click will start stickers' sequence =
for
> cell twist, and right click defines vertex-centered twist. Again, you see
> possible destinations for the selected sticker, click one of them, then n=
ext and
> so on.
>
> To change color of the face, ctrl-right-click any sticker of the "old" co=
lor and select "Edit Color" item in the menu.
>
> There was almost no testing of the program =96 only one solve of 3^4. So =
all bug
> reports are welcome!
> If you have some puzzle that you want to try, just ask about it =96 I'll =
try to
> code it's description ))
>
> Good luck!
> Andrey
>
>
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>
>