Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!ginosko!uakari.primate.wisc.edu!ames!pacbell!indetech!david From: david@indetech.com (David Kuder) Newsgroups: comp.arch Subject: Re: LEGOs -- computationally complete? Summary: a mechanical turing machine exists Keywords: physical Turing machines Message-ID: <1989Oct9.175528.3835@indetech.com> Date: 9 Oct 89 17:55:28 GMT References: <3300071@m.cs.uiuc.edu> <4792@orca.WV.TEK.COM> Reply-To: david@indetech.com (David Kuder) Distribution: na Organization: Independence Technologies, Inc. Fremont, CA Lines: 44 In article <4792@orca.WV.TEK.COM> paulsc@radio_flyer.WV.TEK.COM (Paul Scherf) writes: >While I was in college (a few years ago), a friend of mine and I were >going to build a Turing machine. We also tried to find articles on >such things. The professors we happened to ask, were all very nice >about it, but most of them didn't understand at first. They would >either draw a Turing machine "tape" with the little squares on the >chalkboard, or explain that we only needed a short Pascal program. >Then we got a chance to explain that we wanted to build a PHYSICAL >Turing machine. No one knew of any articles, but if we succeeded in >our quest, we were invited to give a colloquium for the department. > >We got quite far into the mechanical design. When we got far enough to >start collecting parts and begin construction, we had to "take time >out" to study for semester final exams and never got back to working on >our Turing machine. (-: > >Paul Scherf, Tektronix, Box 1000, MS 61-028, Wilsonville, OR, USA 97070 >paulsc@orca.WV.Tek.com 503-685-2734 tektronix!orca!paulsc In introductory CS courses at CalTech they drag out a mechanical turing machine. The tape consists of a metal rack of three position sliders. A segment of "tape" had about 16 sliders on it. Several segments could be connected to form a longer tape. If they had more than three segments I never saw them. The bottom of the rack was geared so it could be moved past the read/write head. If I remember correctly the read was destructive: it was accomplished by shoving the slider to "off" and seeing how far it moved. A write pretty much reversed the operation of a read: shove the slider out 0, 1, or 2 notches. The rack could be moved left or right under the head. Both the head and the rack were driven by noisy relays. This thing was real loud and slow. Just right for holding the attention of freshmen :-). The finite state machine was built out of blocks. Each block represented one state. The blocks hooked together on a bus to drive the tape and read/write head. Its been a while but I think the blocks had arrows and by flopping them over you changed from left to right movement. The transition from state to state was done by plugging wires from 0, 1, or mark read holes to the input of the next block. Blocks were either color coded or had a switch to control what they wrote. It's been too long since I've seen the thing. If it still exists, perhaps someone in closer vicinity to it can describe it. -- David A. Kuder Comp.lang.perl, the time is now! 415 438-2003 david@indetech.com {uunet,sun,sharkey,pacbell}!indetech!david