ucim wrote:It represents the inverse of the degree of privacy, using any reasonable metric you wish to establish. All metrics are problematic to some degree, but that's not the detail that matters.Derek wrote:What does that number represent? The probability of being seen? The probability of being seen in an ideal world? The consequences of being seen? The portion that can be seen?
One such metric could be "number of people who could easily see it" divided by "number of people you interact with in your life".
Delegating it to "degree of privacy" doesn't answer the question. How do you measure "degree of privacy"? The only reasonable metric I can think of is the same way computers deal with privacy: A set of boolean flags indicating who has permission to see or know something and who does not. But even this is weak, because in this case permission is transitive (if someone knows your private information, they can tell others), so it only takes one person with a big mouth to ruin your privacy. And as I've said in other threads, wants something is public it is always public, you can never make it private again.
Your proposed metric does not seem very good. If someone does something so that only one person can see, but that one person could be anyone in the world, is it really private? The implication is that any non-zero amount of publicity (ie, at least one person can see) may as well be entirely public, because any particularly person could see it and there is no telling what they'll choose to do with that information. This metric is guaranteed to create an endless stream of disasters where someone said or did something "privately", but the wrong 1 person saw it and exposed them. Any model that encourages people to act privately while not guaranteeing that privacy is an untenable model, imo.