SPUG: Incorrect subtraction result

Fred Morris m3047 at inwa.net
Tue Aug 26 21:50:48 PDT 2008


I had vague amusing notions that I would put up a web site which would let 
people put in their 10 digit telephone numbers, and give them two integers 
which when treated as dividend and divisor would yield them a repeating 
decimal fraction, repeating their telephone number. That turned out to be 
pretty interesting. I learned a few things; decided I'd defer the exercise.

The problem is that low-order primes don't necessarily occur in the sequence 
1/9,1/99,1/999.... or something like that. Anyway it worked with varying 
degrees of success for some telephone numbers and not at all (within 
reasonable integer precision) for others. It's late, I have a job; crypto 
geeks are welcome to ping me back on the weekend.

On Tuesday 26 August 2008 07:51, Aaron West wrote:
> I'd like to add that if it seems strange that some fractions (such as 1/5 or
> 1/10) cannot be represented accurately in binary, consider the fact that 1/3
> and 1/7 (for example) cannot be accurately represented in base 10 (decimal).
> The reason is the same.



More information about the spug-list mailing list