Up until the spring of 2008, if you asked me what "friendly" numbers were, I would refer you to my listing for amicable or aliquot numbers. Two numbers were amicable (friendly) if the sum of the factors of each equaled the other. 220 and 284 are the oldest pair known, and they date back to Pythagoras. The proper divisors of 220, 1, 2, 4, 5, 10, 11, 20, 22, 44, 55, and 110, add up to 284 and the same is true the other way around. The numbers were inscribed on "magic charms" in the middle ages which were sold to insure the fidelity of ones lover. Other stories suggest that the gift of 220 Goats from Jacob to Esau in the Biblical story was an expression of love made significant by the use of one of the pair. Until very recently, at least to the best of my knowledge, this was what people meant when they referred to "friendly" numbers. This is how Simon Singh describes them in his book Fermat's Enigma; Western mathematics knew only the one pair until 1636 when Fermat discovered a second pair; 17,296 and 18,416 . This was also how Hoffman described them in his book about Paul Erdos, The Man Who Loved Only Numbers, and how Alfed Posamentier had used the term in Math Charmers: Tantalizing Tidbits of the Mind. These were world class mathematician/authors. I assumed I was in line with the current usage.
Then recently, I came across a referrence on mathworld that described them as the ratio of the sum of the divisors divided by the number itself. For example, 8 can be divided by 1, 2, 4, and itself, 8. The sum of its divisors is 15, so its ratio, is 15/8; this is sometimes called the "abundance" of the number. Do not confuse this with the much older term "abundant" for a number for which the sum of the proper divisors (factors, or divisors not including the number itself) is greater than the number itself.. For example the proper divisors of 8 are 1, 2, and 4, which total 7, so 8 is NOT abundant, but deficient.
The classification of numbers as being deficient (the sum of the proper divisors is LESS than the number), abundant (the sum is greater than the number) and perfect (the sum is equal to the number, as in 6 and 28) goes back at least to Nicomachus (about 100 CE) who separated the even numbers into abundant (it was the mistaken belief for a long time that all odd numbers were deficient), perfect, and deficient.
When the ratio of the sum of the numbers divided by the number equals two, the number is perfect, so all perfect numbers are friendly under this new usage. Other numbers that are mutually friendly besides 6, 28 and the rest of the perfect numbers include 30 and 140, with a ratio of 12/5, as well as 80 and 200, whose ratio should be 2.325, if I calculated right. There are many numbers, such as all the primes, that are known to be solitary, that is, they have no friends. All the primes have a ratios of (p+1)/p, so their ratio would get smaller towards a limit of 1 as the size of the prime grew larger. There are other numbers, some relativly small like 10 and 14, that we do not know if they are friendly or solitary. I wondered as I computed these if there is a number or a pair with the greatest ratio? (write if you know, please)
I am not sure, and am presently searching to find the first use of the newer use of friendly numbers, but it seems to exist at least since the 1970's (surprised me!) from a citation on Sloans integer sequence site for ...Anderson, C. W. and Hickerson, D. Problem 6020. "Friendly Integers." Amer. Math. Monthly 84, 65-66, 1977. Ok, if you know about this stuff, drop me a line and set me straight.
A few weeks after I wrote the above request, Walter Nissen sent me a nice note confirming that this newer use of "Friendly" does date from about 1977 (he gave me 1975). He also gave me a link to his site where he explains much more, (and much more clearly) about these numbers, and has some links for further exploration.