Anderson on naming

[sawyl]

Having made it to chapter 6 in my slow re-read of Security Engineering, here's what Anderson has to say about naming and identity:

...a common mistake is to confuse naming with identity. Identity is when two different names (or instances of the same name) correspond to the same principal (this is known in the distributed systems literature as an indirect name or symbolic link). The classic example comes from the registration of title to real estate. It is very common that someone who wishes to sell a house uses a different name than they did at the time it was purchased: they might have changed name on marriage, or after a criminal conviction. Changes in name usage are also common. For example, the DE Bell of the Bell-LaPadula system (which I’ll discuss in the next chapter) wrote his name “D. Elliot Bell” in 1973 on that paper; but he was always known as David, which is how he now writes his name, too. A land registration system must cope with a lot of identity issues like this.

Anderson, R., (2001), Security Engineering, 1st edition, Wiley, 128

And on the problem of uniqueness:

Human names evolved when we lived in small communities. They were not designed for the Internet. There are now many more people (and systems) online than we are used to dealing with. As I remarked at the beginning of this section, I used to be the only Ross Anderson I knew of, but thanks to Internet search engines, I now know dozens of namesakes. Some of them work in fields I’ve also worked in, such as software engineering and electric power distribution; the fact that I’m www.ross-anderson.com and ross.anderson@iee.org is just luck—I got there first. (Even so, rjanderson@iee.org is somebody else.) So even the combination of a relatively rare name and a specialized profession is still ambiguous.

ibid. 130

All of which seems particularly apposite, given the recent Google+ kerfuffle...