**stevev@tekchips.UUCP (Steve Vegdahl)** (08/24/84)

In response to a recent article part of whose definition of fib was: if n < 2 then fib := 1 This should have been (assuming n positive): if n < 2 then fib := n The sequence is 0, 1, 1, 2, 3, 5, 8 ... not 1, 1, 2, 3, 5, 8 ... Reference: Knuth, volume 1, page 78.

**ljdickey@watmath.UUCP (Lee Dickey)** (08/27/84)

Steve Vegdahl says: > In response to a recent article part of whose definition of fib was: > if n < 2 then fib := 1 > This should have been (assuming n positive): > if n < 2 then fib := n > The sequence is > 0, 1, 1, 2, 3, 5, 8 ... > not > 1, 1, 2, 3, 5, 8 ... I tend to agree, but the properties of this sequence that make it interesting to me do not depend on how many initial terms one omits.

**jpl@allegra.UUCP (John P. Linderman)** (08/27/84)

>Steve Vegdahl says: >> In response to a recent article part of whose definition of fib was: >> if n < 2 then fib := 1 >> This should have been (assuming n positive): >> if n < 2 then fib := n >> The sequence is >> 0, 1, 1, 2, 3, 5, 8 ... >> not >> 1, 1, 2, 3, 5, 8 ... > >I tend to agree, but the properties of this sequence that make it >interesting to me do not depend on how many initial terms one omits. The question is one of indexing, not of omitting. One of my favorite identities involves Fibonacci numbers and greatest common divisors, namely GCD(fib(i), fib(j)) = fib(GCD(i,j)) but it relies on the "proper" indexing of the Fibonacci's. John P. Linderman Small fib Department allegra!jpl