Internet Identity Workshop 2007, Day One

Quick thoughts from day one of the Internet Identity Workshop (IIW):    (M9G)

  • This is the fourth IIW. The first one was in October 2005. Amazing. It feels like we’ve been doing these for at least five years.    (M9H)
  • Over half of the participants were there for the first time.    (M9I)
  • I opened the conference with an introduction to Identity Commons. Got some good feedback, and great support from others who have been active in the rebirth of Identity Commons. My big goal is to get the community to think of Identity Commons as “we,” not “they.” We’ll see how successful we are at the end of this workshop.    (M9J)
  • We participated in a nice exercise where folks got into small groups and surfaced questions. It got people interacting, and as Phil Windley noted afterwards, people stayed in small groups chatting away well after the day had ended.    (M9K)
  • One thing that struck me about the group exercise: I heard no new questions. A common characteristic of Wicked Problems is not knowing what the questions are. A good number of us seemed to have successfully identified most of the key questions a long time ago. This is both a sign of progress and of concern. We as a community are starting to face growing pains, and community memory is becoming more and more of an issue. Doc Searls suggested that in addition to surfacing the questions, we should have asked, “Okay, who has the answers?” I think some variation of that would have made an excellent complementary exercise.    (M9M)
  • I like Pibb, JanRain‘s Web-based real-time group chat tool that uses OpenID. (Think IRC on the Web with OpenID for identities.) But I also agree with Chris Messina; Pibb needs permalinks — granular as well as thread-level.    (M9N)
  • We had a series of lightning presentations following the group exercise. They were all well done. Remarkably, they were all about basically the same thing, only told from different angles, something that Mary Hodder also observed. I think this is a good sign. It shows the ongoing convergence of our community. There was also a lot of Spotlight On Others — folks referring to each other’s work, even borrowing slides from each other — another sign of a healthy community.    (M9O)
  • There wasn’t anything new conceptually, but there were many more implementations, yet another sign of progress. Speed Geeking basically consisted of 15 different implementations of Single Sign-On, which doesn’t make good fodder for demos, but which is great for the community.    (M9P)
  • Two Speed Geeking projects stood out: Vidoop and Sxipper. Vidoop is user authentication via image recognition and categorization, which in and of itself is interesting. But what got people buzzing was its business model: sponsoring images that would be displayed to users for authentication. I don’t know if it’s viable, but it’s definitely creative. Sxipper is a Firefox plugin that handles account registration and login. What’s really interesting is what’s happening beneath the covers: It’s essentially an OpenID Identity Broker running from your browser. It looked very slick; I’m looking forward to playing with it.    (M9Q)
  • Doc Searls gave his traditional day one closing talk. I’ve heard bits and pieces of this talk many times, but I never tire of listening to him speak. He’s just a fantastic storyteller, and he’s always on point.    (M9R)
  • I carpooled with Fen Labalme, and as we were discussing our takeaways on the way back, he said, “I’m glad I didn’t sit with you at dinner.” He wasn’t joking, and I wasn’t offended! I felt the same way! One of the really special things about this community is that there are no snobs. We all like to hang out with each other, but we all also really value quality time with folks we don’t know. You could really see this at dinner. I didn’t see any cliques, and there was plenty of mixing.    (M9S)

Angry Rant on Wikis

Earlier this month, Jonas Luster invited me to speak at WikiWednesday. I didn’t have anything prepared, and I didn’t feel particularly motivated to prepare anything, so, I told Jonas that I was just going to rant. Jonas, being Jonas, loved the idea. So after IIW wrapped on May 3, I headed up to Palo Alto. I promised folks at IIW that I was going to give an angry rant on Wikis, and so several people decided to come watch, including Phil Windley, who blogged it. Feedback was great, except for a few complaints that I wasn’t all that angry. I promise to get more worked up next time, folks.    (KK2)

I’ve made all the points I made in my rant before in some form or another, often on this blog. Nevertheless, it was the first time I shared these ideas as one semi-cohesive thought, and so it’s worth rehashing the points here.    (KK3)

Overview    (KK4)

There are two things that make Wikis cool:    (KK5)

Lots of folks have latched onto the open access part, and there’s been some interesting exploration in this area. Very few folks know about or understand the Shared Language aspect. I think this is a huge loss, because it’s what makes Wikis truly transformational.    (KK8)

Open Access    (KK9)

Since I had just come from IIW, I started with digital identity. First, I said that all Wikis should support some form of distributed Single Sign-On, be it OpenID or something else. Implementing Single Sign-On does not imply loss of anonymity. Most Wikis give you the choice of logging in or not; implementing Single Sign-On would give you the additional choice of using a single identity across multiple sites.    (KKA)

Why would this be useful? Consider Wikipedia. As my friend, Scott Foehner, commented in a previous post on this topic (to be visible again when I turn comments back on), Wikipedia actually consists of a number of different Wikis, one for each language plus a number of special Wikis, such as its community site. Each of those Wikis require a separate user account. Not only is this a huge inconvenience, it effectively prevents you from having a single digital identity (along with your associated reputation) across each of these sites.    (KKB)

Simply having Single Sign-On across all of the Wikipedia Wikis would be valuable. More importantly, the identity community has converged to the point where it doesn’t make sense to roll your own protocols. There are several good existing protocols to choose from, and many of those are in the process of converging.    (KKC)

Reputation is closely associated with identity, and it’s also been one of the most popular topics in the Wiki community over the past year. However, most people have a misguided notion of what reputation is and what we should do about it. Reputation is what others think about you. Reputations exist in every system, whether or not they are explicitly represented. Reputation cannot be quantified. However, you can identify the factors that determine reputation and make those factors more explicit.    (KKD)

In Wikis, this could manifest itself in a number of ways. For example, one way to determine the quality of a page is to view the number of people who have edited it. You could make that number explicit by subtly changing the background color of that page — slightly yellowed for a page with few contributors and bright white for a page with many contributors.    (KKE)

The important point here is that you are not making a value judgement on reputation. You are not saying that a page that has many authors is better than a page that does not. All you are doing is making it easy to see that a page has many authors. Readers can determine for themselves how much weight (if any) to place on this factor for the reputation algorithm in their heads.    (KKF)

The most important button on a Wiki page is the Edit button. That button implies Permission To Participate. It should be one of the most visible buttons on any Wiki. If a Wiki looks too good, that discourages participation. Who wants to edit something that looks like a finished product? Ward Cunningham used to suggest sprinkling typos across a Wiki page to encourage others to participate.    (KKG)

At this point in the rant, I plugged both Ward and MeatballWiki. The Wikis success is no accident. A lot of the fundamental design features that make Wikis powerful were completely intentional, a testament to Ward’s brilliance. Additionally, most of what I ranted about is not new to the Wiki community. A lot of it — and more — has been discussed on the venerable MeatballWiki. If you really want to get a deeper understanding of how to improve Wikis, you should be on Meatball.    (KKH)

Shared Language    (KKI)

Last September, I wrote:    (KKJ)

What really makes the Wiki’s LinkAsYouThink feature special is that it facilitates the creation of SharedLanguage among the community that uses it. As I’ve said so often here, SharedLanguage is an absolute prerequisite for collaboration. The lack of SharedLanguage is the most common roadblock to effective collaboration, be it a small work team or a community of thousands.  T    (KKK)

It bears repeating over and over and over again. Wikis are transformational because they facilitate Shared Language. This is a feature that should be propagated far and wide, both in Wikis and other Collaborative Tools.    (KKL)

I noted two possible convergences. The first is Wikis and tagging. They both share a similar principle, namely namespace clash, and we should look at ways of combining these two concepts. For example, where’s the tag cloud view of a Wiki’s page index? Another idea: Clicking on a tag should also return Wiki pages of the same name. Technorati should be indexing Wiki pages and treating their titles as tags.    (KKM)

The second is implementing Link As You Think in all tools. Blogs that are built on top of Wikis (such as TWiki and JotSpot) have these features, but you don’t have to build a tool on top of a Wiki for this to work. This blog runs on blosxom, but it has Link As You Think. Chris Dent‘s blog runs on MovableType, and it has the same feature. It shouldn’t just apply to blogs, either. It should work in web-based forums and other Collaborative Tools.    (KKN)

Free Identity!

A suggestion for Jimmy Wales‘s list of things that need to be free: Free identity!    (JNG)

“Free” in this case has a different meaning than it does than it does with the other items on Jimbo’s list. We need to free our digital identities from the organizational silos that currently collect and control information about ourselves. I am not suggesting that all digital identities fall under an open content license; I’m saying that the individual should have the ability to decide who has access to his or her digital identity and what they’re allowed to do with it.    (JNH)

Why is this important? Privacy is the obvious and most important reason. A secondary reason is that free, or at least mobile identities are a prerequisite for Jimbo’s tenth item: Free communities! It’s not enough to be able to migrate content from one community to another if you can’t also migrate people’s identities as well.    (JNI)

How can we free identities? Technically, it’s not that’s hard, and there are already several proposed specs and implementations, all of which support some notion of Single Sign-On and profile sharing with individual control. Personally, I’m partial to the Identity Commons approach with i-names, where identifiers are globally resolvable, information is distributed, and the notion of contracts built into the data structure. In the end, it doesn’t matter. What matters is that we agree on an interoperable technical specification for identity. Fortunately, many of the folks in this space are already working on collaborating, thanks to the efforts of Owen Davis, Kim Cameron, Paul Trevithick, Doc Searls, and many others. These people have taken to calling themselves the “Identity Gang.”    (JNJ)

The social questions are the hard ones. What does it really mean to control our identities? What should the social and legal agreements between individuals and organizations look like? If I give my business card to someone, what’s the implicit contract associated with this action, and what would it mean to make that contract explicit?    (JNK)

These questions are hard, but they’re solvable. Unfortunately, we’re not devoting much energy towards these issues right now. Perhaps a more public exhortation for freeing identities will lead to an effort to address these social questions that equals the current effort to solve the technical ones.    (JNL)

WikiMania Hackfest Day 4

Bits and tids:    (JM7)

  • I didn’t plan my Hacking Days schedule very well. I missed most of the first day, when the Mediawiki developers apparently made progress on a new metadata design. Days 2 and 3, from which I based most of my criticism, focused on servers and reliability, an area to which I really couldn’t contribute, not because I’m ignorant, but because I’m powerless. This morning, they discussed Single Sign-On and usability, two areas that I do know something about. Sadly, I missed these sessions, because I was too busy spouting on and on about how we really can save the world. Owen Davis, Fen Labalme, Kaliya Hamlin, and the rest of the gang will undoubtedly kick my butt when they read this. In my defense, I managed to talk a bit about Identity Commons later in the day. I also plugged the FLOSS Usability Sprint, and met Zeno Gantner, who’s done some usability studies on Mediawiki.    (JM8)
  • I was one of the featured participants for the afternoon “Wiki developers informal discussion,” along with Ward Cunningham, Sven Dowideit, Christophe Ducamp, and Brion Vibber. Domas Mituzas, Wikimedia Foundation‘s head of operations, asked Ward, “Why Camel Case?” I won’t go into the explanation here — I have a long interview with Ward, to be published eventually, that explains this in detail — but you should know that hating Camel Case is a running joke among this community. I laughed along with everyone else, but when Sven mentioned his desire to remove Camel Case from TWiki, I felt compelled to pipe up. I gave a balanced defense, describing Camel Case’s advantages over free links, but also acknowledging the appropriateness of free links in Wikipedia. Then I got a very amusing introduction to Erik Moeller, one of Mediawiki‘s core contributors and the Wikimedia Foundation‘s chief research officer. Erik had a strongly worded response. It got a bit heated, but never overly so, and I closed by saying that we were in violent agreement. We laughed about it over dinner, but then we got serious again. We also talked about Purple Numbers. I’ve explained many times why I may seem like a poor evangelist, but I think Erik was one of the few people who appreciated my perspective. He was clearly not a big fan of Purple Numbers — as it turns out, he was somewhat familiar with my work — but after hearing my explanation, he responded, “Only intelligent people are going to understand what you just said.” Fair enough. Fortunately, regular folks don’t need to get Granular Addressability for Granular Addressability to become ubiquitous.    (JM9)
  • A group of us broke out into a small group to discuss a Wiki Interchange Format, knowing full well that this is an issue that’s been discussed many times before (Wiki:WikiInterchangeFormat, MeatBall:WikiInterchangeFormat). Nevertheless, I think our discussion was not only constructive, it has a high chance of succeeding. See my summary.    (JMA)
  • Magnus Manske, the original creator of Mediawiki, participated in our Wiki Interchange Format discussion. He also mentioned a clever idea: a “shopping cart” where people could aggregate and possibly export Wiki pages they were interested in.    (JMB)
  • Sven Dowideit demonstrated the prototype WYSIWYG editor for TWiki, based on Kupu. He also showed a WikiText editor with real-time preview, which was pretty slick. Also, Ross Mayfield showed me a prototype editor for KWiki in response to my previous post. Very good to see these things.    (JMC)
  • So many people have come to this gathering to learn from others with different experiences. Granted, all of these experiences center around Wikipedia, but I’m still envious. My neverending quest is for folks interested in collaboration to look beyond their own narrow domains for deeper insights.    (JMD)

Observations from Portals 2005

When I worked at Dr. Dobb’s Journal, I did the software development and IT conference circuit regularly. Most of those conferences were incredibly boring, but they were rarely a waste of time. What made them compelling were the attendees.    (IM0)

I’ve been spoiled in the six years since. Not only have the conferences I’ve attended been more diverse and interesting, many of them have exploited collaborative processes that emphasized participant interaction. That’s obviously an advantage if the reason you’re attending is to meet interesting folks. Additionally, most of these events were more about social good rather than corporate productivity. As a result, the energy is much more positive.    (IM1)

Attending Portals, Collaboration, and Content Management 2005 these past few days was a blast to the past for me, which was exactly why I chose to attend. I wanted to reconnect with the corporate IT community and discover what they were thinking about these days, especially regarding collaboration. I also wanted to test my ideas with this crowd, to see if I still remembered the language of this community and if my message would fly.    (IM2)

I gave the first talk in the collaboration track, and it was very well received, moreso than I expected. There was a snafu with the program, which listed my talk as, “Collaboration: What’s In It For Me,” when the actual title was, “Collaboration: What The Heck Is It?” One woman approached me afterwards and told me that she was originally planning on attending my talk, then saw what the real title was and decided to attend a different one instead. Afterwards, she ate lunch with several people who did attend my talk, and much to her chagrin, they raved about it.    (IM3)

Several people told me they enjoyed the interactivity of my presentation. That was intentional. It engaged the audience, and it gave me a chance to learn from them. My plan wasn’t to teach, it was to stretch people’s minds, to give them an opportunity to think about things in new ways.    (IM4)

Folks who know me well or read this blog regularly know how much I tout highly interactive conferences. I think there is a huge opportunity for such an event for IT workers. I heard very little that interested me in the conference tracks. The attendees were far more interesting than the speakers, and most of my learning occurred during the meals. Several people even said as much, completely unprovoked by me.    (IM5)

Some other observations:    (IM6)

  • I ran into a number of people who had been with their companies for 15 years or longer. One person suggested that the reason for this was that companies liked to put their most experienced people in charge of portals. It makes perfect sense. These folks have an innate understanding of the organizational dynamics, which portals should parallel.    (IM7)
  • Kaliya Hamlin suggested that people attending this conference would be really interested in Identity Commons. Sure enough, several people said they were looking for good Single Sign-On solutions. However, despite my active involvement and evangelism with Identity Commons, I don’t think Identity Commons provides what these people are looking for right now. The real value of Identity Commons as an identity solution is inter-organizational, whereas most IT people are dealing with intra-organizational problems.    (IM8)
  • I was blown away by the proliferation of SharePoint in organizations. During my talk, several audience members realized that they were all dealing with similar challenges with SharePoint, so they gathered afterwards to discuss. I discovered many others in similar situations. I mentioned this to some folks at the SAP Netweaver booth, and they said they were blown away by the same observation. SharePoint seems to be making real viral headway in organizations, largely from the bottom-up. Ironically, some IT people are expressing the same misgivings about SharePoint as they do about Open Source software.    (IM9)
  • I love warm, summer nights. Yes, I realize it’s still spring. An April evening in Phoenix is about equivalent to a July evening in Los Angeles.    (IMA)