Higgins and Vendor Sports

A few days ago IBM, Novell, and Harvard's Berkman Center announced the Higgins Project.  Higgins is a "user-centric identity" system that will be open source and managed by the Eclipse Foundation.    When we held the Internet Identity Workshop last October, Paul Trevithick spoke about Higgins (listen to the podcast).  As Paul said then:

The goal of the Higgins Trust Framework is to address four challenges: the lack of common interfaces to identity/networking systems, the need for interoperability, the need to manage multiple contexts, and the need to respond to regulatory, public or customer pressure to implement solutions based on trusted infrastructure that offers security and privacy.

If you read the press responses to the Higgins announcement, you're likely to get lost in all the vendor sports talk: "IBM takes on Microsoft with Open Source Answer to InfoCard" and all that.  The reality is that Higgins is complementary to InfoCard.  Microsoft, no matter how big, can't be everywhere.  Higgins covers some of the ground that Microsoft can't. 

I had a discussion with Dale Olds from Novell yesterday about Higgins.  Dale is a Novell Distinguished Engineer and their point man on Higgins.  Here are some of my questions and his responses (edited for brevity):

Windley: Tell me why Higgins matters.

Olds:  Higgins is open source.  That helps build trust in consumer's use of system because they can inspect what's going on, if they want.  Higgins is really a very thin layer than application can write to that supports plug ins to identity systems underneath, so it links identity systems together.   Whereas Liberty [Alliance] is about backend connections between providers, Higgins is focused on the user aggregation point instead.  

Windley: The press has played this up as "open source vs Microsoft."  Do you see it that way?

Olds: Not at all.  Higgins and InfoCard get along and share similar goals.  There's a huge amount of territory here.  The idea of metasystems has a lot of room.  There's territory there that Microsoft isn't covering and that leaves room for other parties to make contributions.  The design work that Microsoft has done in the identity metasystem is very good.  The way that data is transfered is very important and we expect to work with them in those areas and embrace areas where they can't go. 

Windley:  There's a danger that we get user choice in the way that Doc Searls calls "your choice of silo" rather than a universal system.  Do you see that happening.

Olds:   Think about TCP/IP, email, and IM (instant messaging).  TCP/IP and email won because of interoperability, not best feature set.  IM is a counter example.  Those silos are a detriment.  The reason for identity to exist on the Internet is to have relationships and communication with other people.  Thus, interoperability is the goal. 

Windley: Why are Novell and IBM working together on this?

Olds: IBM and Novell are working together because they had seen silos in the identity space (directories) and were interested in creating interoperability.  IBM pointed Novell toward Higgins.  This fits well with Novell because just as Novell has had an initiative of taking the enterprise network and moving beyond the bounds of the firewall, the same principles apply to consumer identity on the Internet.  We can take our experience on identity and roles and identity management and refactor them for user-centric identity. 

Windley: When Microsoft announced InfoCard, I expected the Novell would create a project to build and InfoCard compatible client for Linux.  Is that idea off the table?

Olds: Novell might still implement a Linux client for InfoCard.  This isn't an either/or thing.  Higgins needs a plugin supports InfoCard.  Higgins provides an interface that apps can write to to deal with identity systems.  InfoCard, SXIP, and others will all  have Higgins plugins.