Dynamic Identity Numeric Representation (DINR)
I have just realised what seems a good way of giving each person a dynamic number that seems optimal for representing someone identity with a number, to be more precise with a vector of 10 ordered pairs.
As I explained in my article: "Earth Cubic Spacetimestamp (ECS)",
https://bitcointalk.org/index.php?topic=141141.0 you can use 22 digits to specify any geolocation on Earth, using cubes with 1 meter edge. 21 million cubed gives us: 9261 000 000 000 000 000 000 cubes.
Now if you register the cube you are on every second, making a dynamic list of them, updated whenever possible, you can use the analysis of the space geolocations, given by the cube numbers made of 22 digits,
and get DINR: [(1,%),(2,%),(3,%),(4,%),(5,%),(6,%),(7,%),(8,%),(9,%),(0,%)]
That is, the vector will always be, for the maximum number ECS cubes numbers you can consider reliable to have in your register to generate the updated DINR, the most precise and exact percentage possible for the frequency of each of the algarisms in that list of ECS cubes numbers.
I'm not sure how many decimals for the percentages would be a reasonable to have, but I think a good few of them would do.
It should fit an ID card with Liquid Crystal Displays, IMO. And ensure the ID's would be unique, except very rare exceptions like two people living in the same body, or something.
The good thing about it is that you don't need to divulgate the list of ECS to others and still this simple representation is the most accurate uniqueness of them, in a particularly compact format.
Also it paves the way to a new paradigm of very anonymous and dynamic way of representing our ID in time, as you can always had a timestamp to it and add an extra percentage parameter to tell how much percent of your life does it pertain with.