Posted 8:33 AM (2) comments

Whose data is it?

The Burton Group's Bob Blakley has a great post ("Antisocial Networking") today about the Facebook-Scoble story. The essence (or, at least one essence) of Bob's note is that relationships are a different order of data from attributes. As he says:

"Even the fact of your relationship with Scoble is not Scoble’s property, it is common property, like the kids in a joint custody arrangement. Both you and Scoble are obligated by the laws of relation here and here to treat the fact that you have a relationship, and also the details of the relationship, according to certain understandings and social conventions. If you don’t believe this, meditate on whether you think it would be OK for adultfriendfinder.com, match.com, and linkedin to share friend lists. The information Scoble tried to take out of Facebook is NOT Scoble’s property; it is relationship information. Scoble is not free to do whatever he pleases with relationship information; if he violates social understandings and conventions by disclosing the existence of or certain information about his relationship with you in the wrong context, he may embarrass or endanger you, and he will certainly endanger the relationship."

And that's what it's all about.

Of course, not all relationships are reciprocal. I have a relationship with Edith Piaf - I'm a great admirer of her singing. The relationship isn't reciprocated, of course, and not only because she's been dead for many years. But I also have a relationship with the very lively Tom Hanks, of whom I'm a fan. I don't think Tom is one of my regular readers, though, so I doubt the "fan" relationship is reciprocated.

Human relationships may need to be classified similarly to mathematical transitivity. There are:

• reciprocal relationships (e.g., a is friends with b and b is friends with a);
• non-reciprocal relationships (e.g., a is a fan of b but b is not a fan of a);
• relatively reciprocal relationships (e.g., a is father to b, b is daughter to a); and
• asymmetric relationships (e.g., a loves b, b can't stand a).

Some of these relationships will need joint permission for publication, some won't. Some will allow unidirectional publication, some will require it. It's not going to be easy, it's not going to happen soon, but a relationship calculus is going to be necessary for this to work at all.

Labels: attributes, ownership, relationship, social networks

Posted 10:38 AM (2) comments

Uniqueness

The Burton Group's Mike Neuenschwander has posted a mild rant touched off by trying to bag memorable, unique - but recognizable as his - OpenID's on various sites. He tells a good story before getting to his point, which he summarizes as: "There are no identifiers, only attributes."

Mike falls into the trap of the absolutist generalization (there's a lot of that going around, isn't there?). But he concentrates on your given name as an "identity":

"Names are slippery. Most people have many more than one legal name, none of which are unique. They also have several dozen nicknames. There’s no practical way to get any of these every-day-use names onto a global namespace. And what’s a name after all but a synthetic attribute—a foreign key that we hope the receiving party stores somewhere so we can remember them later? "

An "identifier" does need to be unique within a particular namespace. In a family, this is done thru a combination of given names and nicknames. e.g., while my son was at home growing up, I was known as "Dave" and he was called "David". Outside of our immediate family, of course, ambiguity quickly cropped up. There's my nephew (his cousin) David Kearns , for example.

But we are all familiar with unique identifiers within the digital world. Your email address - every single one of them - is a unique identifier within the entire world of the internet. There's also your ip address, but in a DHCP world, that can change without warning. It would still be unique, but tying it solely to you would be a more difficult task.

OpenIDs are unique. A little judicious shopping for an OpenID provider (OP) should get you one who has available an identifier that associates well with you, and that you wouldn't be ashamed to share with others.

Labels: attributes, openid, uniqueness