Many of you were instrumental in OpenID, and I was really excited about it at first. I wanted it for one reason: To provide a single unified log-on for weblog comments. To do that, it had to provide two things:

1. Provide an identifier in the form of a URL that I could use.
2. Provide basic persona information, a name to show, maybe some contact info.

As the protocol evolved, the conversation ended up in all sorts of places off in the weeds, and I thought "well, those are two simple primary functions of this system, I can just step out and let everyone else fight out their needs".

When the dust settled, though, interoperability was (and continues to be) bad enough that it only did half of the first. And, yeah, there's been a little bit of progress on the second, but all of those other features, especially the ones that created the "Nascar Badge" situation (ie: multiple users can provide a single starting-point URL).

This is why I'm speaking up now, and why I'm trying to steer the discussion to solving real problems.

Text version of the slide

Solve Real Problems

• OpenID had two uses for me:

• Provide a single-sign-on identifier
• Provide persona information

• It might have solved the first one if we'd had test suites.