Thread: "Magic 120 Cell Solve"

------=_Part_24664_27756457.1229222304624
Content-Type: multipart/alternative;
boundary="----=_Part_24665_15792945.1229222304624"

------=_Part_24665_15792945.1229222304624
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello again,

Well, in keeping with my tradition of solving one of these puzzles during my
exam time I've solved the Magic 120 Cell. I have to say it wasn't as
difficult as its large number of permutaions would suggest, just time
consuming. I must say thank you for the piece finding feature since without
it there is no way that I could have solved this monster.

2.3 x 108126 combinations? Sounds like a lot of fun to me!
Noel

------=_Part_24665_15792945.1229222304624
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello again,

Well, in keeping with my tradition of solving one of these puzzles during my exam time I've solved the Magic 120 Cell. I have to say it wasn't as difficult as its large number of permutaions would suggest, just time consuming. I must say thank you for the piece finding feature since without it there is no way that I could have solved this monster.

2.3 x 10⁸¹²⁶  combinations? Sounds like a lot of fun to me!

------=_Part_18437_273615.1229230292483
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hi Noel,

Wow, amazing! This first solution happened much sooner than I guessed
(which was that this puzzle would resist for at least a year). Great job,
and I would love to hear about your experience and any feedback it may have
generated. 33k moves didn't outnumber your 26k move solution of the 5^5 by
too much ;) What was your overall approach, solving in order by piece type
or tackling it cell-by-cell?

I just got into Houston for a short trip, but will be back in Austin
tomorrow night. I will be sure to get the web site updated with this result
then!

All the best,
Roice
On Sat, Dec 13, 2008 at 8:38 PM, Noel Chalmers wrote:

> Hello again,
>
> Well, in keeping with my tradition of solving one of these puzzles during
> my exam time I've solved the Magic 120 Cell. I have to say it wasn't as
> difficult as its large number of permutaions would suggest, just time
> consuming. I must say thank you for the piece finding feature since without
> it there is no way that I could have solved this monster.
>
> 2.3 x 108126 combinations? Sounds like a lot of fun to me!
> Noel
>

------=_Part_18437_273615.1229230292483
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Yes, and a huge congratulation from me too! I'm also a little surprised
this happened this quickly. That said, I'm not surprised that it was by
you, Noel. You're a puzzling animal! I'm starting to wonder whether my
prediction that it would be a very long time, if ever, until we see a
*second* solution, might also be very wrong. What do you think, Noel?
Was it fun? Do you think you'd ever do it again? Maybe to take back the
shortest record once someone breaks yours? Please do tell us all about
your method, experience, and feelings about this frightening boss
monster. Most of all, aren't you glad to know that no higher dimensional
analogs exists? :) This is the end of the line for the really special
4D figures. I hope that your achievement helps you enjoy a much-deserved
relaxing holiday break.

-melinda

Roice Nelson wrote:
> Hi Noel,
>
> Wow, amazing! This first solution happened much sooner than I guessed
> (which was that this puzzle would resist for at least a year). Great
> job, and I would love to hear about your experience and any feedback
> it may have generated. 33k moves didn't outnumber your 26k move
> solution of the 5^5 by too much ;) What was your overall approach,
> solving in order by piece type or tackling it cell-by-cell?
>
> I just got into Houston for a short trip, but will be back in Austin
> tomorrow night. I will be sure to get the web site updated with this
> result then!

--00032557416a3efc6d045e30ea3c
Content-Type: text/plain; charset=ISO-8859-1; format=flowed; delsp=yes
Content-Transfer-Encoding: 7bit

Hello All,

Thank you for the congratulations! To answer your question Melinda, yes it
was a lot of fun and I might do it again sometime if I think of a faster
method! ;)

As for your question Roice, the way I solved this thing is entirly similar
to how I solve the 4D cubes. It seemed like a good starting point sice I
solve a 3D Megaminx similarly to how I solve a 3D Rubik's cube.

Now that this project is done (it only took me about a week btw :) ), I
would really like to go back to the 4D cubes but I admit, using the special
features that Roice included in the program
makes the puzzle experience much more enjoyable, not to mention faster.
Could we please have these kinds of features in the 4D cubes??? Please?? I
think everyone would love the piece finder that Roice invented and the
piece highlighting would be very useful for the larger cubes.

On another note, I know I'm not the only one that has got their hands on
the new V-cubes. I think that we know we can solve the 3D 6x6 and 7x7 it
would be cool to give the 4D versions a shot. Maybe after those I'll tackle
the 5D versions. :D

Cheers,
Noel

--00032557416a3efc6d045e30ea3c
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Hello All,

Thank you for the congratulations! To answer your que=
stion Melinda, yes it was a lot of fun and I might do it again sometime if =
I think of a faster method! ;)

As for your question Roice, the w=
ay I solved this thing is entirly similar to how I solve the 4D cubes. It s=
eemed like a good starting point sice I solve a 3D Megaminx similarly to ho=
w I solve a 3D Rubik's cube.

Now that this project is done =
(it only took me about a week btw :) ), I would really like to go back to t=
he 4D cubes but I admit, using the special features that Roice included in =
the program
makes the puzzle experience much more enjoyable, not to me=
ntion faster. Could we please have these kinds of features in the 4D cubes?=
?? Please?? I think everyone would love the piece finder that Roice invente=
d and the piece highlighting would be very useful for the larger cubes.
/>
On another note, I know I'm not the only one that has got their=
hands on the new V-cubes. I think that we know we can solve the 3D 6x6 and=
7x7 it would be cool to give the 4D versions a shot. Maybe after those I&#=
39;ll tackle the 5D versions. :D

Cheers,
Noel
--00032557416a3efc6d045e30ea3c--

One tiny correction: I think I invented the piece finding design though
it was Roice who implemented it.

I'm a little torn about supporting larger cubes in MC4D but if people
want it, I see no reason not to support it. That is assuming the
original memory explosion that happened when I initially tried that
doesn't present a problem. Perhaps more importantly, I don't want to
keep records related to larger cubes, so if people are OK with that,
I'll see what I can do the next time I work on it. That might not be for
a while as I tend to work on it between contracts/jobs, and right now
I'm happily employed working on Second Life. I would like to take a
break from that soon but things are going well there so I have no idea
when that will likely be. I definitely agree that a cubie-finder would
be extremely helpful when working on those big boys!

So you solved the 120 cell from the inside-out? Regardless, does the
program support a playback mode? In MC4D you can adjust the speed while
doing a "cheat" solve to play the solution backwards. I should really
allow for the equivalent "redo" playback to play it forward. Also, a
screensaver mode that plays forward in an endless loop would be fun too.
The loop should pause in the pristine state and do a bunch of pure
rotations before starting off again. If at all possible, I encourage you
to create a time-lapse video of your solution and post it on Youtube.
OTOH, I just did a back-of-the-envelop calculation which suggests that
even at movie speeds and one frame per twist, it could still take over
20 minutes to play! Either you'd need a higher frame rate or have to
skip several moves per frame to squeeze it into a reasonable time. Too
bad Roice's log format doesn't mark macro begins and ends, otherwise a
video or playback mode could skip over the macro details. Oh well, too
many fun things to do!

-melinda

ltd.dv8r@gmail.com wrote:
> Hello All,
>
> Thank you for the congratulations! To answer your question Melinda,
> yes it was a lot of fun and I might do it again sometime if I think of
> a faster method! ;)
>
> As for your question Roice, the way I solved this thing is entirly
> similar to how I solve the 4D cubes. It seemed like a good starting
> point sice I solve a 3D Megaminx similarly to how I solve a 3D Rubik's
> cube.
>
> Now that this project is done (it only took me about a week btw :) ),
> I would really like to go back to the 4D cubes but I admit, using the
> special features that Roice included in the program
> makes the puzzle experience much more enjoyable, not to mention
> faster. Could we please have these kinds of features in the 4D
> cubes??? Please?? I think everyone would love the piece finder that
> Roice invented and the piece highlighting would be very useful for the
> larger cubes.
>
> On another note, I know I'm not the only one that has got their hands
> on the new V-cubes. I think that we know we can solve the 3D 6x6 and
> 7x7 it would be cool to give the 4D versions a shot. Maybe after those
> I'll tackle the 5D versions. :D

With respect to playback, you can watch it in reverse the same way as in
MC4D, that is load the log file, undo a single move, then select solve from
the options menu. I need to support a "redo" playback as well. Your calc
was quite accurate, because I played it back the other night (both to see
the solution approach and to test the puzzle state engine with a full
solve), and it took 40 minutes on my computer! Nope, I didn't actually
watch every second :) This was with the rotation speed maxed out, which
turns the program into disco ball mode (only a single frame drawn per
twist). I'm betting it is rendering limited, but my computer is slow in any
case and many of you could likely get a quicker playback.
I like Melinda's idea to not render macro details as a way to compress
playback. It's not just the log files for Magic120Cell though - we'd need
to implement macros first (the fact they don't exist makes Noel's solution
all the more remarkable!).
I like the youtube video idea as well, and may do a little hacking this week
to skip frames for a compressed-solution youtube video. I'll send a link if
successful.

All the best,
Roice


On Tue, Dec 16, 2008 at 9:09 PM, Melinda Green melinda wrote:

> One tiny correction: I think I invented the piece finding design though
> it was Roice who implemented it.
>
> I'm a little torn about supporting larger cubes in MC4D but if people
> want it, I see no reason not to support it. That is assuming the
> original memory explosion that happened when I initially tried that
> doesn't present a problem. Perhaps more importantly, I don't want to
> keep records related to larger cubes, so if people are OK with that,
> I'll see what I can do the next time I work on it. That might not be for
> a while as I tend to work on it between contracts/jobs, and right now
> I'm happily employed working on Second Life. I would like to take a
> break from that soon but things are going well there so I have no idea
> when that will likely be. I definitely agree that a cubie-finder would
> be extremely helpful when working on those big boys!
>
> So you solved the 120 cell from the inside-out? Regardless, does the
> program support a playback mode? In MC4D you can adjust the speed while
> doing a "cheat" solve to play the solution backwards. I should really
> allow for the equivalent "redo" playback to play it forward. Also, a
> screensaver mode that plays forward in an endless loop would be fun too.
> The loop should pause in the pristine state and do a bunch of pure
> rotations before starting off again. If at all possible, I encourage you
> to create a time-lapse video of your solution and post it on Youtube.
> OTOH, I just did a back-of-the-envelop calculation which suggests that
> even at movie speeds and one frame per twist, it could still take over
> 20 minutes to play! Either you'd need a higher frame rate or have to
> skip several moves per frame to squeeze it into a reasonable time. Too
> bad Roice's log format doesn't mark macro begins and ends, otherwise a
> video or playback mode could skip over the macro details. Oh well, too
> many fun things to do!
>
> -melinda
>
>
> Noel wrote:
> > Hello All,
> >
> > Thank you for the congratulations! To answer your question Melinda,
> > yes it was a lot of fun and I might do it again sometime if I think of
> > a faster method! ;)
> >
> > As for your question Roice, the way I solved this thing is entirly
> > similar to how I solve the 4D cubes. It seemed like a good starting
> > point sice I solve a 3D Megaminx similarly to how I solve a 3D Rubik's
> > cube.
> >
> > Now that this project is done (it only took me about a week btw :) ),
> > I would really like to go back to the 4D cubes but I admit, using the
> > special features that Roice included in the program
> > makes the puzzle experience much more enjoyable, not to mention
> > faster. Could we please have these kinds of features in the 4D
> > cubes??? Please?? I think everyone would love the piece finder that
> > Roice invented and the piece highlighting would be very useful for the
> > larger cubes.
> >
> > On another note, I know I'm not the only one that has got their hands
> > on the new V-cubes. I think that we know we can solve the 3D 6x6 and
> > 7x7 it would be cool to give the 4D versions a shot. Maybe after those
> > I'll tackle the 5D versions. :DRecent Activity
>
> .
>
>
>

------=_Part_29883_27774929.1229615144096
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Well, this is my third attempt to get this post through.  I wrote yahoo last night and they responded that they recently upgraded their spam filters and my posts are getting marked as spam.  I guess having links in your messages increases your spam-probability.  Why they don't have members of the group white listed I can't surmise...

Roice,

Bummer about the spam filtering problem. You're right of course that
they should just whitelist all members. I definitely boot any member who
spams our group, no discussion or warning, so that would take care of that.

I didn't know that you didn't support macros. That indeed makes Noel's
accomplishment truly heroic!

I've found that MC4D speed depends more on the size of the window than
on the size of the puzzle. Playback is blindingly fast if I shrink the
window to postage stamp size.

Good luck with that video!
-melinda

Roice Nelson wrote:
> Well, this is my third attempt to get this post through. I wrote
> yahoo last night and they responded that they recently upgraded their
> spam filters and my posts are getting marked as spam. I guess having
> links in your messages increases your spam-probability. Why they
> don't have members of the group white listed I can't surmise...
>
>
> With respect to playback, you can watch it in reverse the same way as
> in MC4D, that is load the log file, undo a single move, then select
> solve from the options menu. I need to support a "redo" playback as
> well. Your calc was quite accurate, because I played it back the
> other night (both to see the solution approach and to test the puzzle
> state engine with a full solve), and it took 40 minutes on my
> computer! Nope, I didn't actually watch every second :) This was
> with the rotation speed maxed out, which turns the program into disco
> ball mode (only a single frame drawn per twist). I'm betting it is
> rendering limited, but my computer is slow in any case and many of you
> could likely get a quicker playback.
>
> I like Melinda's idea to not render macro details as a way to compress
> playback. It's not just the log files for Magic120Cell though - we'd
> need to implement macros first (the fact they don't exist makes Noel's
> solution all the more remarkable!).
>
> I like the youtube video idea as well, and may do a little hacking
> this week to skip frames for a compressed-solution youtube video.
> I'll send a link if successful.