The following was written in response to a posting at Jabber, Inc.’s Filaments Blog. The original article is available at:

http://blog.jabber.com/filaments/2007/03/05/big-content-see-viacom-v-goolge/

Dave, I agree with you wholeheartedly that providing context to users in the modern networked computing environment is essential. Although the model has been for communication to occur between individuals or groups with previously existing relationships, context cues will become increasingly important when we begin to interact with strangers we have located in search of expertise or to provide real-time information in a fast paced environment.

No offense, but I think you are overselling when you state that “instant messaging IS the desktop”. If that is true, then what the heck are we using the desktop for? There are a variety of other appliances better suited to providing IM and presence information. Mobile devices could very easily fill this role and would likely provide more accurate presence information.

True, the desktop IM client is as strong as it has ever been, but in a world where applications are being pushed into web-based delivery mechanisms I think it unreasonable to believe that IM clients won’t go the same way. This sentiment is echoed in a number of responses to the Om article you quoted above.

After all, the power of what we are discussing here is not in the messaging, but rather in the presence engine. The Jabber presence engine is quite robust, but still gains its presence information from the one (historically unreliable) source, the user. Context information provided by today’s IM applications is no more accurate or reliable than when AIM first hit the market. It is still dominated by “online” or “away” type models with very few clues from the user’s actual environment.

I am confident that this will change in time and that Jabber, Inc. will be at the forefront of this change, but in order to get there we must recognize the limitations of where we are today.

The following was posted in reply to comments on the Elastictime Blog regarding presence technology. The original article is available at:

http://elastictime.wordpress.com/2006/12/12/9/#comments

This post is right on target as regards a variety of problems facing systems that attempt to accurately represent presence to users. Given a casual glance, presence does not appear to be a complex concept. Examined in real-world circumstances, however, the standard “online” vs “away” (OvA) model is surprisingly insufficient. In order to develop an effective means of providing robust representations of presence, it is important to develop definitions we can use to describe the various components.

As you describe in your discussion of unified communications, presence systems often relate our presence relative to a device used to communicate (telephone, pc, etc.). We can infer then, that “Physical” presence, is the physical presence of a user at a known device. Physical presence often implies a locational linkage. For example, if a user is physically present at their PC and that PC is labeled as “work” or “home”, a valid assumption regarding the users location can be made. As you mention, technologies such as GPS may lead to more refined representations of physical presence, into what I call “Geographical” presence. There are many shortcomings to the current state of geopositioning technology in the office environment, but it is a proven concept for mobile computing.

Unfortunately, presence systems have not been designed to assure accurate representations of physical presence. While it is great that, according to AOL, 1/4 of users change their away message daily, the vast majority of users never change their presence status. In fact, most presence applications default to the “online” condition after a user logs in. This has the detrimental effect of changing physical presence into “Device” presence, representing that the client application is ready to communicate, regardless of actual user status.

Inappropriate handling of device presence can lead to mistrust of the system and “Are you there?” type messages regardless of the user’s representation.

Another concept you discuss is the availability of the user. I would contend that availability is a representation of a user’s physical presence combined with their willingness to be interrupted. So if we look at the communication process, the user has to be present to communicate and also willing to communicate. Interruption management is a topic for an entirely separate post, but most users recognize the importance of interruptions to their daily work tasks. Any presence system needs to allow both the interruptor and the interrupted to make value judgements about the importance of interruptions to ensure essential communications are permitted while those of less importance are not initiated.

Finally, you mention the Windows Presence Platform. Standardized protocols to provide presence are alredy in existence. One example is the Extensible Messaging and Presence Protocol (commonly known as Jabber), which has been evaluated and approved by the IETF. XMPP, although originally designed to support non-proprietary instant messaging, has grown to be a leading presence technology.

Many of the concepts here require their own greater discussions at some later time. This discussion, however, is a great place to bring these problems out of the darkness and into the light where they can begin to be solved. I look forward to participating further!

There was a time (most of us will remember) when if we wanted to speak with a friend, relative or business associate, we had to pick up the telephone and call them. If they weren’t there we were either out of luck, or fortunate enough to leave a message on a machine or with someone else. E-mail quickly changed the way we handled correspondence and was used by some for short “conversational” messages. Not until the advent of instant messaging, however, did true synchronous communication over the computer become a reality. But was messaging really the important part?

Not a chance! Sure, instant messaging has expanded our social interactions, connected people across the world, and recently changed the way business organizations interact, but what was the real paradigm shift? The answer: Presence.

At no time in the past or present, with the exception of face to face interactions, did we have the ability to determine the availability of the user with whom we desired to interact. Presence cues provided by instant messaging applications allow us to know whether a user is physically present, available to be interrupted, or too busy to expect a reply.

Sure, today’s presence models are somewhat flawed. Most of us have used an instant messaging application, great or small, where we have either logged in, set our presence to “online”, and walked away without ever changing it. In fact, most applications do this for you! Building confidence in the accuracy of such presence modeling systems is a discussion for another posting; however, it is a topic I address extensively in my research paper available here.

What have been your experiences with presence? How do you feel presence systems in existence today could be improved?