Sunday, September 24, 2006
Is the Liberty Alliance still relevant?
When the Liberty Alliance was first formed, it was a forum in which end-user enterprises could with one cohesive voice define not only their basic requirements but also propose solutions in which vendors should implement. Today, there are more software vendors than large enterprises and this trend doesn't seem to be changing.
Likewise, the early influence that Liberty had with SAML was good but now they have moved on to less important issues. Wouldn't it be wonderful if the Liberty Alliance pushed the folks at Oasis to incorporate the notion of impersonation into SAML 3.0?
The biggest proponent of Liberty Alliance in the blogosphere seems to be Pat Patterson of Sun. I have yet to see him in his blog talk about how SAML can be used in other contexts. For example, the folks over at BEA have incorporated SAML support into their Weblogic Server. Should all J2EE containers embrace this approach?
What would happen if a Sun employee decided to submit an enhancement request to the Java Community Process asking for these folks to consider figuring out a way that JDBC could also support SAML? What about if databases from Oracle, Microsoft and IBM could also support SAML? Do you think it may mean that the notion of identity propagation may become a reality?
If identity propogation became real, then all enterprises may be able to tighten up on their Sarbane Oxley deficiencies by having one identity transcend all tiers. SAML can be a savior in the SoX world too...
The one thing that I have struggled to figure out in the Liberty Alliance is not in the realm of case studies or any of the other public information on the site but what the members of large non-software enterprises are telling as their internal story. Of course, I understand at a high-level they may babble about ROI, increasing security, etc along with the usual CIO magazine cliche phrases but wouldn't it be interesting if one of its members shared their internal story with the rest of the world?