5 Dec
2002
5 Dec
'02
5:09 p.m.
1, 2, 2, 4, 3, 6, 6, 5, 6, 8, 9, 13, 10, 11, 14, 15...
I doubt the claim "a(n)/n seems to tend to 1/2" in A011754...
More likely, a(n)/n should tend to log_2(3)/2 = 0.79248+, since a bit is equally likely to be 0 or 1. (Well, the first and last bits are ones; I meant the other bits.) -- Don Reble djr@nk.ca