(In reply to
Solution (Infinite Points) by Paul)
A valiant effort, and on first look satisfying, but as exoticorn points out, it fails at 5 points. I suspect it fails similarly at every level k where k=(2^n)+1 points.
I have not yet tested it out yet, but I'd like to suggest that point A be just ε. Point B would be Paul's Point A (ε+(1/2)). Point C would be Paul's Point B (ε+(1/4), etc. This would relieve the "strain" on the first segment at levels k=(2^n)+1 points. It may just move that strain to a different point, (It does -- at level 3 with points ε, ε+.25 and ε+.5, ε must lie between .16 and .33, a very high value for a "vanishingly small" number) and the points may need to be added in each level range 2^(n-1)<k≤2^n in an order other than simply with numerators 1,3,5... to relieve that strain.
Another possibility might be to start with A=ε, B= 1-ε, and then start adding Paul's sequence with C=ε+.5, D=ε+.25, etc.
|
Posted by TomM
on 2004-01-20 06:17:55 |