Workflow and Lightroom

Workflow is a concept the computer world and academia have been theorizing about and experimenting with for years. What's fascinating about computer-mediated workflow is that workflow systems can be really effective or a disaster.

In more restricted environments where the workflow is shallow - meaning it doesn't deal with complex decision trees - the tasks typically are well defined. An example is a materials return authorization process in which the sequence of steps is fairly straightforward (receive request; check product return terms; if OK for the product to be returned, issue a return-merchandise application number; if not OK, issue a request rejection).

A successfully implemented, shallow computerized workflow is likely to have a big payoff. When it comes to more complex processes, however - insurance claim handling, technical support or interpersonal communication, for example - it is much harder to create an effective workflow because the logic is more complex and exceptions - points at which the workflow logic can't handle the events that occur - become a big problem.

There are two choices for handling exceptions: Ignore the exception (typically the job continues to be handled, ignoring the exception, or is aborted) or allow a human (or other smarter process) to deal with it.

As workflows become more complex, the consequences of ignoring exceptions or letting them be handled too generically increase rapidly. A good example occurred in a product called The Coordinator, published by Action Technologies in the 1980s and adopted for corporate use at the time by Novell. The Coordinator was supposed to streamline messaging in an organization by classifying all messages in terms of interactions between people. This meant that you could make a Request ("Please send the sales figures"), have a Conversation for Possibility ("Which sales figures?") or send a Promise ("I will send the sales figures tomorrow"). This might sound promising, but there was a big problem that involves a fourth type of interaction: a Note.

The problem came from users' lack of discipline in using the system, which wasn't unreasonable as most users couldn't bear the complexity.

To be disciplined in using The Coordinator, users had to consider what kind of communication they intended, select the correct form, and fill in the details - for example, when their Promise was due. Until they provided the details, they couldn't send the message.

As you might imagine, after being trained, the average user worked with this complex methodology for a few weeks (or more usually, a few days), then resorted to sending Notes that required the least amount of effort. Of course, this invalidated the entire workflow. Not surprisingly, today The Coordinator is an almost forgotten historical footnote. For a good description of workflow and references on the topic, we recommend Wikipedia's discussion.

What got us musing about workflow is a beta product we have been testing that we think has incredible potential: Adobe's Lightroom.

Lightroom addresses the problem of managing the workflow associated with digital photography by providing a system that handles the importing, organizing, categorizing, tagging, comparing, selecting, "developing" and showcasing of a large number of digital images - essentially it's a specialized content-management system.

Available for Mac OS X (both PowerPC- and Intel-based) - Windows support is planned - the full release is scheduled tentatively for year-end. That said, the product's FAQ makes it clear Adobe is making no promises.

Lightroom on the Mac is remarkable. The attributes that really struck us are how intuitive the user interface is and how much it stays out of the way. This is an excellent example of how user interfaces should be built.

Next, we'll take a deeper look at Lightroom. 'Til then, lighten our darkroom at gearhead@gibbs.com or on Gibbsblog.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Take IDG’s 2020 IT Salary Survey: You’ll provide important data and have a chance to win $500.