RE: Thompson's Lamp

From: Mark Williamson <>
Date: Mon, 20 Oct 2003 18:40:54 +0100

Just been musing about this on the train home.
There are two parts of the system an ideal lamp and a
clock generator.
The clock generator is what is interesting and it is implementing
a pulse train representing zeno's paradox.
The clock generator will fail in the time quanta just prior to two
Either by exploding/divide by zero or whatever depending on how its
implemented. The lamp state then is the same as the output of the
failed clock generator ie. Not defined.
Your assumption "Yet the clock runs" is a red herring - the clock
only runs until a point after which it is undefined. If you wish
to alter the rules (ie. it does continue running - you are defining
new behaviour not seen in our world and therefore you must
specify it just as you have with the behaviour of the lamp).

Can you tell that I'm a software engineer :-)



ps. what is this list I didn't know I was subsribed to it and have
no memory of doing so (although I must have done)


From: Norman Samish []
Sent: 20 October 2003 08:07
Subject: Thompson's Lamp

I've been looking for an idiot savant to answer this question: Perhaps
you've heard of Thompson's Lamp. This is an ideal lamp, capable of
infinite switching speed and using electricity that travels at infinite
speed. At time zero it is on. After one minute it is turned off.
After 1/2 minute it is turned back on. After 1/4 minute it is turned
off. And so on, with each interval one-half the preceding interval.
Question: What is the status of the lamp at two minutes, on or off? (I
know the answer can't be calculated by conventional arithmetic. Yet the
clock runs, so there must be an answer. Is there any way of
calculating the answer?)

        ----- Original Message -----
        From: incarn81 <>
        Sent: Saturday, October 18, 2003 11:36 PM
        Subject: Joining

        I'm mainly an idoit, sometimes a savant. I get most of the
references that I've read so far, but don't really have a deep technical
background in any one area.
        Can't wait to catch up on the archives!
Received on Mon Oct 20 2003 - 14:07:12 PDT

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:08 PST