For all the time we spend detailing use cases for ever imaginable "happy path", when was the last time we stopped to create a use case that accounts for the "death" of a user? Are we good/humble enough developers to handle the potential that our users might want to, well... leave?

"User death" was a topic that I kept running into at CSCW this year. Not in any papers or presentations, instead the topic was relegated to quiet conversations where people dared challenge the impenetrable user/technology dyad. During one of the first nights at CSCW, I spent a good deal of time speaking with Mike Massimi, a PhD candidate at the University of Toronto. He was kind enough to share some recent theoretical work he submitted to SIGCHI about what he calls "thanatosensitive design." Quoting one his professors, “It's an odd feeling seeing a recent e-mail in your inbox from someone who is no longer here to receive the reply.” Massimi suggests that we need to reconsider user-centered design to account for our inevitable deaths.

Later that week, while talking to Janet Vertesi, we shared our mutual fascination with what we might call "digital identity death." She shared her interest the way that blogs change into memorials once their authors die, while I talked about the growing number of social networking profiles that are renamed to "In Loving Memory". Each of these binds users and friends into a new social network of sorts. Users come and converse around a shared interest: their deceased loved one.

Today, however, I am thinking about something much more fundamental: Why can't I delete my account? Or perhaps, what does it mean when I do? Since Van Gelder's 1991 account of online identity fraud and Dibbell's descent into the MOOs and MUDs, we have had to deal with the blurring of what we traditionally thought of (as Donath says) "one body, one identity". When online, what counts as a "body" or "identity" emerges out of the coconstruction, negotiation, and even contestation of users and technologies. While users may prove their existence with each Cartesian account (i.e., "I login, therefor I am"), the terms of their existence is often preregulated by the technology. Moreover, these jealous applications may go to extremes to prevent you from leaving. Technology does a great job of enabling our own sense of immortality.

To be clear, most of these online "identities" are rather facile. The username and password seems the most common and simple way of reconstituting the body online. MySpace doesn't care about your government issued ID, it is your username and password that proxies your body and allows access. It is not surprising, then, that the mass exodus from MySpace earlier this year during Simon Owens' "International Delete Your MySpace Account Day" equated to a form of cult-like suicide.

From the developer's perspective, it is hard to understand a user's desire to leave our software behind. A quick survey of some of the applications around my association reveals that in many cases, this isn't even an option. While drinking my morning joe, I casually accosted colleagues, forcing them to face their mortality, and asked: "Why can't you delete your user account?"

The most common response was a blank stare, followed by a "Why would you want to?" The answers varied, but were all rather weak: data integrity issues, research reasons, and my favorite "it's not really a business priority." It is certainly not an area that developers spend a great deal of effort on. Owens' blog post encouraging people to leave MySpace resulted in a microflurry of users trying to figure out how to do it in the first place. One of our architects admitted that he sometimes wonders what will happen to his Facebook account when he dies. "I guess my wife will clear it out?"

Clearly, the relationships we have with our online identities are not clear. What responsibility t\do we have to the identity artifacts we create, and what responsibilities do software developers have to us as users? My biggest take home is that no one seems to want to talk about this, let alone create clear exit strategies from the user/technology relationship. Perhaps preparing for the death of a user is similar to preparing a living will. Maybe we are procrastinating a task to a future date that may not even exist in the first place.