OpenID is an easy to setup identification system for both users and developers that is beginning to gain industry momentum. OpenID uses URL's as unique identifiers which is a flexible and memorable way of tackling what your identity is. As it is relatively simple in principle OpenID is not a solution for all identity tasks but in a loosely joined, Web-centric context it is well suited.
All you need to do to use it is sign up for a free OpenID account, or if you are really adventurous implement your own system. When you login to an OpenID-enabled website/service it takes your submitted url and password and passes it on to the OpenID service for authentication. The first time you successfully authenticate you must tell the OpenID service exactly what (if any) information the target website/service should be able to see.
Firewalling what information a third party can see about you is vitally important as it allows you (the owner of your information) to tailor what others can learn from your OpenID account. The upshot of this process is that sites you trust implicitly can learn about your physical address and personal details, whilst a site you hardly know can be left completely in the dark about who you are. This is a far more useful approach than the all or nothing (or simply nothing) strategy employed by older identity systems like LDAP. Because your identity is centered around your OpenID account maintaining your personal information becomes a lot simplier, there is only one password to remember and one set of contact details to manage. Just to help out Drupal has a handy module which allows your website/blog URL to be your OpenID handle which redirects to your chosen OpenID provider (for example myopenid.com).
Overall OpenID is a very promising technology that if it gains critical mass may significantly improve the identity space on the Web. It certainly is not going to dominate the entire space but it gives small-scale developers a great starting point for implementing an easy to use universal identity system.