CMIS: Boom or Bust?
Some of you may have heard about the joint announcement from EMC, IBM, and Microsoft about the creation of Content Management Interoperability Services (CMIS). The purpose of this proposed new standard? To create a vendor-agnostic way of accessing the data in content management systems from multiple vendors. In other words: Remember when SQL became a standard for accessing databases? This is the content management system equivalent.
Of course, this could be enormously attractive, making it much easier to develop applications that sit on top of those repositories. Most enterprises have multiple content management systems in place – either due to diverse needs of various businesses, size of the company, or just plain lack of content management vision. Making that content more easily accessible via protocols such as SOAP could be a major win for these enterprises.
For instance, persuasive content – used to influence customer behavior – frequently lives in multiple databases, making it more difficult for organizations to create consistent experiences across multiple channels. CMIS could enable applications to easily access previously siloed content in order enhance those experiences. CMIS could also allow for a proliferation of tools – authoring, reporting, etc. – to work with a variety of ECM systems; think of all the SQL-compliant tools out there – this would be the equivalent. And it could potentially be huge for SharePoint users, who are not always happy with the tools Microsoft provides to access SharePoint content and would like to better leverage that content.
So why am I skeptical? Well, as a former practitioner, I’ve seen too many standards fail to catch fire (most recently JSR-170, which rarely gets mentioned as a must-have by Forrester clients). Standards are kind of like political candidates: you hope they’ll live up to their initial promise and idealism, but you should prepare for the reality and inevitable letdown of their day-to-day existence.
Luckily, major players in the industry have already pledged support for CMIS: in addition to the above, the list includes Oracle, Open Text, Alfresco, and SAP. But the burden will be on those vendors to demonstrate the power of the CMIS standard by actually supporting it and providing (either themselves or in conjunction with partners) compelling applications which take advantage standard, which will further create demand for CMIS-compliant applications and repositories, which will encourage vendors to create more applications, which will encourage….well, you get the idea. Until then, CMIS remains a well-intentioned concept that has the potential to be great, but could also turn into yet another standards-based pipe dream.