Challenge: Designing for flow within one application across two user bases and two interfaces.

Wikipedia: Flow is the mental state of operation in which the person is fully immersed in what he or she is doing, characterized by a feeling of energized focus, full involvement, and success in the process of the activity. Proposed by psychologist Mihaly Csikszentmihalyi.



In the diagram above, users are either hindered by the UI, in flow, or frustrated when trying to complete a task.

My main users can be generalized in to two types. The mid level network admins with a moderate level of protocol and architecture knowledge. Best sumarised as needing a little 'hand holding' during setup and configuration. I've tailored the web based GUI to them. Then there are the 'cisco certified' power users who prefer the speed of the CLI. They spend more time planning their complex network designs, usualy in advance of working with the interface. The former tends to spend more time exploring the available options, the latter needs to simply get their tasks completed.

With user's needs roughly defined, designing the interfaces has clear constraining boundaries. The interface workflow must present features & options in a way that is neither too dumbed down nor too convoluted based on the user and their prefered interface. For the mid level admins, research is showing that they're happy with the gui. Predictably the power users spend their time in the CLI.



The issue is in the nature of the application, configuring networks. Some of the features transcend the upper & lower flow boundaries of a particular interface.

Going forward I need to understand the areas that are common. That is, the high end features in the low end interface and visa versa for the basic features in the CLI. Explore possibility of web-cli interface cross pollination, launching a CLI within a web ui, and pulling up a gui when appropriate for a CLI user which presents the feature in a more configurable light. As always, in progress.

I don't love/hate* you anymore (*delete as appropriate)

Been evolving my thoughts on the uncanny valley since the post about mimic design in March

Leisa Reichelt ponders on her observation that machine interactions seem to contain less anthropomorphism these days. She asked if it was due to the rise in social interactions diminishing the amount of human qualities we're willing to apply to the machine. Bill Higgins had a different take at an application level with the uncanny valley human/machine comparison at a desktop/web level.

I believe Mr. Higgins is right at a macro level. It ties in with what I noted in my widget design and emphasizes my notion that the uncanny valley phenomenon is not the sole domain of robotics.

Ms. Reichelts observations are very insightful but I feel she arrives at the wrong conclusion. I believe the median state of interfaces has progressed to the point where they are not so quirky anymore. They’re more functional, they’re starting to do what they promise.
When we’re past the lull in the valley, when we’re running intelligent agents on our machines. Then I think we’ll see a full blown revival of anthropomorphism.

Learn from history or repeat its mistakes

The philosopher John Loche's (1632-1704) take on the events that led to the recent Digg community revolt

He believed that we create a social contract when we form a community/society, whereby we cede certain natural rights to an authority in return for security and other gaurantees. Locke's argument is that any fair social contract must have certain qualities: It must respect its citizens rigths to life, liberty, and property. If these rights are violated, we are entiled to rebel against the authority

Lifted from The Intellectual Devotional

The singularity will be savant... In progress

There follows a snippet from an email thread with another body, just capturing it here for now and will tie it in to a more formal essay on the subject soon.

"My bet is we will create the singularity. I do not think it will be apocalyptic, nor an immediate massive watershed, more of a major milestone in computer development. I believe the singularity will be much like a savant. An autistic personality, tremendous archival capabilities and able to view/process the collective knowledge on the net in new ways. The first queen bee of the hive mind. A step towards a true human colony transcending race, gender & national differences. Helping us solve some of the issues we face such as climate change, energy efficiency - possibly.

The second thing to come from the singularity, as I see it, is its ability to copy itself. A fundamental of computer science. These copies will solve the biggest issue in computer interaction - personal AIs (or agents). The agents, nothing as sinister as the matrix, will be our personal assistants. Remembering preferences so we dont have to check the same godam box to not remember a password for the nytimes, recognising faces in photographs and tagging according when it uploads to flickr. That sort of stuff.

I think the negative views of HAL 9000/Terminator/Electric sheep are based on the premise that computers will have some form of personality. I'm my dealings with computers and people in the autistic spectrum, I just dont think the singuraity, the net-savant, will have one. Therefor it wont go psychotic, it might not care that it gets switched off, as it might know it will inevitably be switched back on in the normal pace of progress."