Date: 15 DEC 1980 1851-EST
From: DCP at MIT-MC (David C. Plummer)
Subject: Administrata and an Algorithm
To: CUBE-LOVERS at MIT-MC
CC: ELLEN at MIT-MC, DUFFEY at MIT-MC
First the Administrata. CUBE-LOVERS came close to a crisis this
weekend. There were thoughts of having it digested, but DUFFEY
came through and made some very valuable suggestions. What
happened is that the Hoey-Saxe message was long enough to confuse
the mailer, and it ended up sending it twice. It took at least an
hour and a half to send, so the mailer response was terrible.
Things have been reorganized to help make things more efficient,
and we really thank DUFFEY for doing this. In the future, please
limit your messages to one to two thousand characters. If you
must send long messages, please break it up into sections and
send a piece every hour or so. Please keep this mailing list
winning. Thanks.
Now for the algorithm. Hoey and Saxe mentioned an H pattern. By a
simple method of doing it, it can be done as follows:
FF (LLRR) BB (LLRR) RR (UUDD) LL (UUDD) DD (FFBB) UU (FFBB)
and after removing the NOPs
==> FF (LLRR) BB (LL) (UUDD) LL (UU) (FFBB) UU (FFBB)
which is another 28 mover. But I have found another way to do IT:
(UD LLRR FFBB UD) (FFBB LR FFBB L'R') ==> 24 QTW
Hoey and Saxe said that the Pons Asinorum is the only local
maximum that is PROVEN not to be a global maximum. It still is,
but if somebody can prove that Global Max must be larger than 24
(it is currently at 22), then this would be another example. The
other possibility is to find a faster algorithm to this pattern.
I have an intuitive sense that the global maximum is an even
distance away. I cannot prove it. Can anybody?