There is just no easy way to explain an Internet Identity Workshop. It is an unconference, which is important, but that’s not really sufficient to explain it. It attracts a group of people who care deeply about Internet identity technology, its design and adoption, and its social, political, and economic impact. Many of them are people who have debated, supported, and known each other for a long time, so there is a sense of camaraderie. But it’s definitely not a closed group. It’s outrageously open and inclusive of newbies. Some of the most important decisions and leaps of progress in Internet identity systems have happened at IIW, or as a result of relationships established there. In addition to all of that, there is the untalent show, which is unfortunately self explanatory.

The great usefulness and significance of IIW is hard to explain. You just need to go to the next one.

On the opening day of this IIW I gave an overview presentation of the Open Source Identity System (OSIS) working group of the Identity Commons. After that there was a “speed geeking” event in which fellow Bandit Andy Hodgkinson showed the Higgins Native Identity Agent — an open source application that is roughly functionally equivalent to Microsoft’s Cardspace. Andy is the primary author of the code and finally got a chance to escape from his office for a while and show it off in person.

On Tuesday the OSIS group held a long working session in which we brought together many projects to collaboratively test a set of capabilities and scenarios. By all accounts it was a resounding success. Code from many projects worked together in multiple combinations and end-to-end scenarios. We had far more projects involved than I expected, we checked numerous scenarios, and many bugs and inconsistencies were identified — some fixed on the spot.

Bob Blakley moderated the session and managed to get everyone moving productively even though things didn’t start off quite as planned. The wireless network had become rather unstable. Most developers had difficulty getting an address via DHCP, and even when they did, the routing configuration was off and we could not connect to external servers. At one point Pamela Dingle became a human DHCP server and manually assigned IP addresses to each participant with correct routing information.

The fun was watching the different participants work through the process. Here are some photos of Andy Hodgkinson, Chuck Mortimore, Tony Nadalin, Ian Brown, and Kim Cameron as they collaborated on information card selector implementations:

There was a team from Oracle that was attempting to get their relying party code to work with the various Identity Provider and Identity Agent projects. Things didn’t work at first, but … the expressions on their faces say it all.

Then the Oracle team asked me to take their photo with Kim. Now that’s a friendly interoperability event.

The OSIS session at IIW was intended to be an informal interaction that would help projects prepare for the more formal, and more visible, Catalyst Interop Event. Not only did it succeed in that goal, but I think most participants accelerated their understanding, improved their code, and raised their expectations for what we can show at Catalyst.

Of course there was much more done at IIW than this OSIS working group session. Phil Windley blogged a series of daily overviews.

There were a number of significant trends that seemed to me to take root at this IIW. Clearly there was a stronger interest in political and legal issues surrounding Internet identity systems. I think this is because the implementations and deployments are happening, and as people actually use this stuff they encounter subtle changes in their thinking. We start to actually think in more social terms in our internet use. It is fairly easy to recognize that there must be numerous points of contact between the identity systems and the legal system. I am sure there will be much more emphasis and discussion on these issues in the future. I know my own thoughts in these areas are still fermenting.

Another trend of this IIW was the increased emphasis on marketing. One session resulted in a new Identity Marketing working group of the Identity Commons.

The Bandit team was fortunate to be right in sync with this trend. We managed to get our Novell marketing representative (and Bandit-at-heart) Carolyn Ford to attend. Carolyn managed to soak up a huge amount of information and IIW culture in a short time.

After IIW I attended a meeting of the ITU-T Focus Group on Identity Management that took advantage of the IIW attendance to co-locate its meeting to Mountain View. While the meeting was vastly different than IIW and hosted by a vastly different organizational culture, it confirmed that the same issues around identity systems for networked devices are coming up in many areas.

I attended Kuppinger Cole and Partner‘s 1st European Identity Conference in Munich, May 7 – 10.

The trip to the conference was an adventure in itself. It started with ticket confusion and delays for international bag check-in, then long security lines out into the airport parking lot, running to another terminal with shorter lines, sprinting down the concourse only to be told that the paperwork had already been done and I could not board (even though the plane was right there and the door was not shut when I arrived). Sigh. Missed flight. Reschedule. Another missed flight in Atlanta due to lack of an available (airplane) parking place. Reroute through Paris. I’d never been to Paris before. Now I can say that I’ve run through a very nice airport in Paris. About 24 hours after I left home, I got to Munich.

At least I learned a new phrase to use when I don’t want to cooperate with someone: “I’m sorry sir, the paperwork has already been completed and there’s nothing I can do.”

I had been trying to arrive in time to hear Dr. Jeff Jaffe’s keynote. I knew part of his presentation would refer to the Bandit project and I wanted to hear both the presentation and the audience reaction. Oddly enough, even though I arrived many hours later than scheduled, the conference was running a little late and I got to hear most of the presentation.

Jeff covered some great concepts about the history of Identity Management products and the role of Open Source. His presentation went very well, and there are photos of it here and here. Since then he has written about the same concepts.

The next day I participated in a panel discussion on Trends in Open Source Identity Management. It was a very lively discussion. Tim Cole did a fine job of moderating before it got too lively. What is not obvious in the photos is that we are in front of large classroom and it’s full of people — standing room only. In the audience were Dick Hardt, Conor Cahill, Bavo De Ridder, and other experts in this area. So the session was also very interactive. At one point, David put a question to the audience, and the panel listened as members of the audience debated. In the end it was very productive. I think numerous valid points were raised, but mostly we found that, while vigorously discussed, there was not as much disagreement as expected. As I remember the main points:

• Open source development has advantages, but we don’t expect identity infrastructure to be exclusively open or exclusively closed source.
• Open standards are essential, and open source development can be very complementary with standards development.
• Numerous protocol families are gaining prominence and have valid uses, but there will not be a single dominant protocol in the near future.
• Open source identity services must plan for clear evolutionary paths from existing systems.
• We have moved out of the realm of debates about theory and possibilities and into debates about user experiences, system capabilities, and operational experiences. Deployments are happening.

Afterward the discussion continued over some fine German beverages.

The conference itself was very informative, and very well run. However, my usual programmer attire is a little underdressed for European gatherings. I’ll work on it.

Munich was beautiful. My hotel was fairly far away from the conference location, so I got to know the subway system. The subways were amazingly bright and colorful.

A few months ago the Bandit team showed an open source identity selector at Novell’s Brainshare conference.

I wrote about the demo then, and so did many others, but Gil Kirkpatrick‘s blog post about it really caught my attention. It was significant in many ways. He had not actually seen the demo, just read and heard about it from others, yet his comments were particularly insightful. He really got the important points of what was shown. As his comments about me also show, some of us have been working in the Directory Services and Identity Management space for a LONG time. And times are changing again in the identity services world. There are some new possibilities now. My perspective is that this has a lot to do with a clear movement away from an enterprise and vendor specific focus, towards more of an emphasis on integrating business with the general Internet, and therefore interoperable identity services. Gil’s comment seemed to me to reflect this perspective — he was excited about the possibilities for general use of an open source selector. He even mentioned perhaps giving news from my employer a fresh hearing (I appreciate that)!

So I contacted Gil, and we decided to renew our aquaintence at NetPro’s Directory Experts Conference.

I wanted to attend DEC for many reasons. For years I had heard that it was a great conference. My background is in directory services and I knew this was a serious conference for the same types of troubled souls as me. I also knew that Kim Cameron, Stuart Kwan, and Pamela Dingle would all be there and giving talks about information cards to the Active Directory Faithful — THAT sounded entertaining. And it was in Las Vegas which is about a 6 hours aways by car, so it sounded like a a great excuse for a road trip.

All very good reasons, but the reason I was particularly interested in attending DEC was a chance to help with an identity system used at the conference and get almost-real-world deployment experience for Bandit and Higgins components. The conference identity system in question involved information cards, embarassing photographs, and a large poultry impersonator. After all, it was in Las Vegas.

The DEC hot chicken contest, as well as it’s usage of Bandit and Higgins components, was put together by the Pamela Project. The system involved getting conference members to either avoid embarassment or win a zune by accessing a web site with an information card. I think there was some reference to a carrot and stick, but when there’s a huge chicken walking around I’m not sure whether the reference was literal or figurative. The best writeup of the overall system and it’s results is from Pamela herself, here.

The Pamela Project Cards site allows anyone to create an account and generate a managed information card so that they can gain access to the Hot Chicken site. I got to help set it up and even wrote some actual code. It was great fun, and very useful; I gathered a lot of very good input about how to improve our identity provider package in the future.

I would like to emphasize that the Pamela Project’s “Cards” Identity Provider was built from completely open source software. It included components from many projects, but the most notable to me are the Higgins STS, Bandit management, authorization and audit code, all on an Ubuntu LAMP system.

The conference itself was well run, with great sessions, great food and a positive environment for conversation and collaboration. A great time was had by all, I’m sure, and certainly pointedly educational for me. Many thanks to Gil and Pam.

Oddly enough, the very next week my friend and coleader of the Bandit team, Pat Felsted, also took a road trip to attend yet another conference in Las Vegas. His description of the experience is here.

I haven’t updated this blog for a while. It hasn’t been due to lack of interest or lack of activity. Actually, quite the opposite. A huge amount has been happening and I have a number of experiences I’d like to relay to illustrate the progress that has been made by Bandit, Higgins, the Pamela Project, and other aspects of the Open Source Identity System community…

But first….

I needed to move this domain and my blog to a new hosting site, move registrars, get a certificate, rent a static IP address, and on and on. What a pain. I wanted to make the move for various reasons; one of which was to have more control so that I can finally start using the cool new emerging identity systems. And I’m a little stubborn about how I manage my own stuff, so I have been futzing around with it between conferences and customer trips for the past few months. I appreciate the patience of my Internet tech supporters as they put up with my questions and peculiar spurts of interest. Now the site is moved, and seems to be operational. I have tried to keep it as much the same as possible, so you won’t notice many changes yet, but it represents a whole new set of possibilities for me.

For now, please notice that the login page now accepts an information card or an OpenID.

The OpenID plugin seems to be operational. There were a few hiccups in installation, but overall it went well.

The PamelaWare wordpress plugin is a development snapshot, but I am very happy with it. The installation was smooth and configuration was straightforward. The nice validation checks on the options page made the whole process much easier, and enabled me to quickly work out a few glitches in my installation.

I wish the certificate management and domain transfer had been that easy.

The co-existance of the OpenID and the PamelaWare WordPress plugins is experimental. Oddly enough, even though they implement different identity protocols, they seem to get along just fine — philosophically and practically. Please let me know (via a comment or my i-name) if you notice any bugs or if you have any suggestions.

Enabling this blog for Internet identity is just a first step. More to come.