Hi FRED.

26-Apr-03 19:02:00, FRED KANTOR wrote to HERMAN TRIVILINO


 FK> In a message dated 04-25-03, Herman Trivilino said to Fred Kantor:

 FK>> PMFBI... I'm interested in the foundations of physics;
 FK>> especially, re amount and representation of information in
 FK>> physical systems.

 HT>> Would that be information theory?

 FK> No. The area of work is called 'information mechanics'.

 FK> IMHO, what is called 'information theory' would seem to be based
 FK> on a set of assumptions which does not seem able to support
 FK> formal, quantitative treatment of amount and representation of
 FK> information, i.e., 'information bookkeeping'.  Let's go into that
 FK> a little, here:

 FK> Amount of information equals log base 2 of the number of
 FK> possibilities from which one is designated.

 yes, if you're measuring it in bits (which is the usual unit of information
 theory)

 FK> That amount of
 FK> information does not depend on which one of the possibilities was
 FK> designated.

 FK> Built into the definition of 'the real numbers' and the definition
 FK> of 'the integers', is the idea that you can always add another;

 yes.

 FK> That assumption that you can always add another point, means that
 FK> one element out of such a set is assumed to be one out of
 FK> arbitrarily many possibilities.


yes.

 FK> For this reason, the amount of
 FK> information assumed to be represented by the designation of any
 FK> one specific such value appears as log_2(arbitrarily many), which,
 FK> in their mathematical picture, can be made bigger than any
 FK> particular value.

yes, but innformation theory is not usually applied to elements of infinite
sets.

 FK> 'Information theory' was based on such mathematics.

yeah, it's main computer related uses are in error correction,
and data compression,



 -=> Bye <=-

---