Kaliya Hamlin, the Identity Woman, in an essay for O’Reilly, wrote

There were many who thought that an identity solution would emerge to support single sign on (SSO) shortly after the Web’s emergence in 1994. An SSO solution has proven very elusive. Solving internet identity management, creating an efficient, reliable ecosystem, is often alluded to as “The Holy Grail.” One of the reasons for this elusiveness is the fact that identity is no small matter.

Yes, identity is no small matter, but the failure to establish a widespread SSO is not just related to philosophical and technological complexities of identity.

In the early days of the web (1994-2000) online identity was mostly managed by portals offering all kind of services: search, news, email, IM, weather, online shops, etc. Portals were eager to get as much personal information about their users as they could (and as their users could bear). Today is common to register just with an email and a password, but in those dark days a subscription process often required a snail mail address, phone numbers, financial and professional details, answering polls, etc.

It was such a common practice that even small websites adopted the same strategy, gathering lots of information mostly unrelated to the tiny and probably useless service they were providing. And, of course, each big portal or small website kept his treasure of personal data for themselves supposing it was a real treasure and not something that could backfire on them in many, many ways (the technical difficulties and costs to consistently store and use those data, the privacy issues to deal with, the security issues, the trust issues, etc.).

Eventually the user got upset and some of the websites realize how big was the hassle of registering customers. So, in 2001, Microsoft came up with the brilliant solution: Passport … So, in 2005, we are still debating which is the best approach to implement an effective SSO …

It’s a very lively and interesting debate, which involves many brilliant minds: Drummond Reed for XDI/XRI, Kim Cameron for InfoCards, Dick Hardt for SXIP, Johannes Ernst for LID, Dave Smith for Passel, etc.

My feeling is that what is needed is a real user centric solution, a solution where only the user owns and knows his personal information. A solution aiming just to solve SSO and not the whole digital identity issue.