I agree with Roice that Don lives on a higher mental plane. It's been a
multi-decade collaboration that's worked really well since it required
both of our strengths. Even still we couldn't have brought it even half
this far by ourselves. We were able to produce version 1 in 1988 in C
using the Dore 3D graphics library that we were helping to build. That
version was little more than a demo that has since been lost in time.
Don then reimplemented a better engine and I built version 2 around that
using C++ and Direct3D. If version 1 was a demo, version 2 was a product
prototype. It was good enough to share the binaries, but the code was a
real mess and neither of us felt good enough about it to publish it as
open source, nor did we have the energy to clean it up. So there it sat
until an angel named Jay Berkenbilt came to our aid by offering to do
all the dirty work of getting it under source control and massaging it
into publishable shape as well as implementing the Linux version. From
that cleaned-up source I was able to translate it into Java, and Don
implemented a simple rendering system to project the 4D views all the
way down to a sorted set of 2D polygons that Java could draw. Thus
version 3 was born. Most recently, Don created this amazing new polytope
engine which required yet another rewrite to support. This is where
Roice really stepped up and got Don's prototyped version installed in
the Google Code repository where he then took over management of Don's
engine while providing invaluable help and encouragement while I rebuilt
the UI around it. Now with a passionate group of users (That's You!), we
are now getting invaluable help testing, discussing, and filling out the
project wiki, and are very close to launching version 4 to the world. It
really does take a village, and I'm just as amazed and grateful and as
the rest of you.
My congratulations to Anthony as well. His first solution to the much
coveted pentagonal duoprism is very impressive. It's doubly cool that
his short solution is exactly 2000 twists. I'm looking forward to seeing
a bunch of new firsts from you guys!
To Chris and Roice's points about individual records in the wiki in
addition to the firsts and shortests: Roice is right that it is already
moving in that direction. Klaus had the brilliant idea of linking from
the puzzle names to a page for that puzzle. Currently there is just the
one for the 3-simplex, and the links from both tables point to the same
page on the history of its shortest solutions. My suggestion would be to
make them point to different pages (or different anchors on the same
page), with the link from the "firsts" table linking to a hall-of-fame
page containing a table for all first solutions by all people of each
puzzle size. I especially like Roice's suggestion of including screen
shots of each puzzle. Perhaps we could even put little ones in each of
the first column of puzzle names. Everyone loves pictures, and these
would make it clear just how impressive each puzzle is.
Finally, I'd like to include a personal note to all the readers of this
list who never post. I know that there are a lot of you because you've
download the new versions and because you haven't unsubscribed to escape
the sudden flood of traffic on the list. I'm betting that many or most
of you feel nervous about speaking out in a forum with so many
brainiacs, but I assure you that you are fully qualified to chime in,
and that your feedback is very much appreciated. For example, where do
you get confused by the program? What do you think would be helpful *to
you*? Even if you don't want to speak out here, we would love to get
your participation in the issue tracker
Thanks to every one of you. You guys are awesome!
-Melinda
Roice Nelson wrote:
> Hi Chris,
>
> You're welcome, and thanks for all your appreciation :D This also
> gives me the opportunity to publicly thank Melinda and Don myself.
> When Melinda shared this with me in mid-June, I was blown away. Don
> did an amazing job with the bones of the program (I think he
> experiences permutation puzzles on a higher plane of existence, or
> maybe a better way to put it is "on a hyperplane of existence"), and
> Melinda has been amazing in making those bones a polished product.
> I'm really thankful to have been included, and feel privileged to get
> to play a part with it.
>
> Congrats to Anthony too. I was using the the "{5}x{4} 3" as a test
> run of the program, so he trumped me as well :) To answer your
> question on the wiki, the sky is the limit. One reason we wanted to
> do the wiki was to give power to everybody in helping to direct things
> like this. I think a list of "full solutions in order of achievement"
> type table would be nice to have for each puzzle, and I encourage you
> to start one for this puzzle as soon as you complete your solution.
> The wiki could evolve a number of ways as the number of solutions
> grow, but in my mind I was imagining a main puzzle page with a basic
> list of all the puzzles and the first/shortest, which linked to a
> puzzle specific page for each. The latter would have info about each
> puzzle (maybe a picture, the piece/permutation counts, etc), and the
> full list of everyone who has solved it, as well as the history of
> shortests. I've been pleased to see that things already look to be
> heading this direction, and I guess we'll see how it turns out!
>
> Thanks also for your thoughts on solutions. I am finding the
> pentagonal duoprism quite challenging, with (dare I say) "parity
> problems" wreaking havoc for me. I hope to complete it at some point
> as well.
>
> All the best,
> Roice
>
>
> On Wed, Oct 28, 2009 at 6:44 AM, Chris Locke
>
>
> First of all, I want to thank Melinda and Roice for writing and
> sharing this great piece of software with the world! You've done
> a great job and we are all very thankful for your hard work.
>
> Second, I want to congratulate those who have solved some of the
> new puzzles so far. In particular, Anthony for solving the 3
> layer deep pentagonal duoprism. I actually started working on the
> same puzzle too a few days ago and managed to solve about half the
> puzzle before you uploaded your solution to the wiki. Congrats!
> Upon looking at your solution, I see you were able to solve the
> puzzle with much more finess than I have been working with. I
> tend to rely on macros quite a bit to work my way out from 2 color
> pieces, up to the 4 color pieces. As such, I often work with
> reset puzzles to find algorithms to permute/orient just the pieces
> I want, and then use these algorithms to build up other
> algorithms. So really, my way of solving is mostly just being
> careful and patient. It also helps that after a while, you find
> out what kind of moves tend to give potential good algorithms.
> Commutators are one useful building block. Also, you can
> sometimes find new algorithms by taking other sequences (like a
> simple aba'b') and applying it multiple times. Doing this you
> will sometimes find it fixes much of what it does, leaving just
> what you need moved around (like it's almost nilpotent). Of
> course, conjugation can then vastly increase the usefulness of
> each algorithm. .... but that's a bit more of a digression than
> I intended :D
>
> Before I forget, this actually brings me to a question I have.
> Will there also be a record page for everyone who's solved the
> puzzles, not just the first and shortest solutions? If not, then
> I will still finish my puzzle, but I won't be in as much of a
> hurry, so I can do other things too in my evenings :D. However if
> we will have such a list, I might try to finish it earlier rather
> than later if I find the time.
>
> Chris
>
>