OCSI Meeting Synopsis

I was in Anaheim yesterday for the Open Collaborative Services Initiative (OCSI, pronounced “oxy”) workshop, which was part of the OMG Technical Meeting. Johannes Ernst, one of the OCSI organizers, invited me to present my manifesto on collaborative tools (which will be published in Dr. Dobb’s Journal and on the Blue Oxen Associates web site).    (WI)

OCSI is an attempt to get collaborative tool vendors to make their tools more interoperable. One of its early goals is to develop a shared architectural blueprint for describing collaborative tools, perhaps initially in the form of a white paper. This has been a refrain of mine for quite some time, and so I was very glad to participate in the group’s second meeting.    (WJ)

As it turned out, there was a tremendous amount of conceptual synergy in the room. I suppose I shouldn’t have been too surprised. At the beginning of my talk, I explained that one of our beliefs (also known as the The Blue Oxen Way) is that Shared Ontology (which results in Shared Language) is a prerequisite to effective collaboration. OMG is a very strong proponent of Model Driven Architecture, which is essentially an instantiation of Shared Ontology. Not surprisingly, there was universal consensus in the room about developing a shared model of collaboration — both on the human-level (e.g. Blue Oxen‘s work with Pattern Languages) and the system-level (the topic of my manifesto).    (WK)

In his introductory remarks, Johannes made several interesting points:    (WL)

  • The word “collaboration” means many different things to different people. This simply underscores the need for a common vocabulary.    (WM)
  • Collaboration seems to be an “it” topic among CEOs and CIOs. However, as often as they mention collaboration and as important as they claim it is, the collaborative tools market has been flat the past few years. At first, this seems to be a contradiction. However, the number of corporate downloads of free IM clients over the past few years indicates that the need for collaboration is real. One of the problems is that tools are not interoperable enough.    (WN)
  • There is no horizontal industry initiative for improving interoperability of collaborative tools. However, several vertical industries have expressed interest. One of the challenges is to get the different industries to realize that they share common needs so as not to duplicate efforts.    (WO)
  • Johannes chatted with a few tool vendors about this problem. Their response: “That sounds great, but I have a product to get out.” The way to get vendors more serious about interoperability is probably bottoms-up — via the user community.    (WP)
  • In this regard, the Open Source community could play an important role. The prequisite for standards is Shared Language and free implementations. We have the latter, but we don’t have the former. If we created Shared Language and if Open Source tool-builders adopted it, we could build a compelling case for standardization. Johannes feels that it is vital to involve both the proprietary and Open Source communities in the OCSI effort.    (WQ)
  • Collaborative interfaces should be as transparent as telephone numbers. When we see a telephone number, we know what to do, regardless of the underlying service provider, protocol (POTS versus VOIP), type of telephone, etc.    (WR)
  • Cut-and-paste is a type of interoperability between collaborative tools. (A poor one, as I and others noted later in the workshop, but also a relatively effective one — a good example of loose-coupling.)    (WS)

Other talks of note:    (WT)

  • David Hartzband, VP of collaboration technology at , provided a four axes view of collaborative tools: synchronous, asynchronous, inline, and contextual. He also observed two trends in the collaborative tools space: business communications convergence (e.g. telephone integrated with email integrated with your documents, etc.) and enterprise application functional convergence.    (WU)
  • Carol Burt, CEO of 2AB, shared her vision for model-driven access management. Not only could such a model have ramifications for those developing secure applications and those selling security software, it could also potentially plug in to an OCSI model for collaborative tools.    (WV)

At the end of the workshop, Joaquin Miller (the other OCSI co-organizer) led a discussion about the next steps within the OMG umbrella. The consensus seemed to be to propose the formation of an OMG SIG, which could potentially evolve into an OMG Task Force. Not being an OMG member myself, the conversation both baffled and fascinated me at the same time. Nevertheless, the folks there seemed to know what they were talking about, which is always an excellent sign.    (WW)

The next meeting will be at the next OMG Technical Meeting in St. Louis next April. We’ll continue to collaborate via an eRoom set up by David and via the OCSI web site. Our action item for now is to share our individual high-level models of collaborative tools in order to identify commonalities and to serve as straw men for additional discussion.    (WX)