Saturday, February 21, 2009

 

More Thoughts on CMIS

I have learned a lot regarding CMIS from Pie and had some additional thoughts on how this specification can be improved...



I have always found it fascinating as to how vendors almost always seem to gravitate towards the lowest common denominator when a superset conversation may be the better answer. Let's agree that vendors want to do the bare minimum and their customers always want additional functionality. In this scenario, why couldn't they have noodled every single Documentum function and figured out a way to indicate whether the functionality was implemented. A CMIS implementation could throw a Not Implemented Fault.

I suspect that it may be more due to the fact that knowledgable security professionals aren't participating in the creation of CMIS and therefore you are going to get status quo solutions. The agenda doesn't seem to show any security professional participating in a security discussion.

Any predictions on whether this will actually happen? After all, they just invented it and it would take Craig Randall to acknowledge that the baby he created isn't that cute. Could you see him admitting this in public?

The WSDL used by Stellent tends to be more self-describing than the DFS stuff. It would be great if they used annotation/documentation tags along with choice to describe vs pointing users to another source for values.

My take on this is that this is determined by the consumer and therefore there should be a way to request compression within CMIS.






<< Home
| | View blog reactions


This page is powered by Blogger. Isn't yours?