home *** CD-ROM | disk | FTP | other *** search
- # From: heim@tub.UUCP (Heiner Marxen)
- # Newsgroups: sci.math
- # Subject: busy beaver Turing machine
- # Date: 24 Aug 89 13:24:10 GMT
- # Organization: Technical University of Berlin, Germany
- #
- # This is about ``busy beavers'' (is there a more appropriate newsgroup?).
- # Unfortunately I did read this newsgroup only sporadically, and don't know
- # whether this has been discussed already. My other sources of information
- # (mainly the German issue of `Scientific American') don't tell me more.
- #
- # As far as I know the 5-state busy beaver question is not yet settled.
- # With the help of a program I have found a 5-state Turing machine which
- # halts (after 11,798,826 steps) and produces 4098 ones on the tape, namely
- # A0 -> B1L A1 -> A1L // `A' is the initial state
- # B0 -> C1R B1 -> B1R // `R' is `move right'
- # C0 -> A1L C1 -> D1R // `L' is `move left'
- # D0 -> A1L D1 -> E1R
- # E0 -> H1R E1 -> C0R // `H' is the halting state
- # The best machine I knew before produces 1915 ones (published in 1985
- # by Scientific American, I believe). My questions are
- # Q1: Is there ongoing (or completed) research ? Any (theoretic) results ?
- # Q2: Are there any better 5-state machines known or published ?
- # Q3: Who else studies the busy beaver problem with general purpose computers?
- #
- # Answers to the above, hints and pointers are welcome.
- # Please answer by e-mail; if appropriate I will summarize to the net.
- # = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
- # Heiner Marxen, from europe: unido!tub!heim
- # from world: pyramid!tub!heim
- # bitnet: heim%tub.BITNET@mitvma.MIT.EDU
-
- 1. 1l2 1l1
- 2. 1r3 1r2
- 3. 1l1 1r4
- 4. 1l1 1r5
- 5. 1h0 0r3
-
-