Jacob_5_I

I am pretty familiar with task flows.  They work well to convey the logical connections between actions and states in a design.  However, what I was not aware of before reading this book is what Saffer calls Use Cases (Saffer 107). I thought it was pretty interesting, and it actually parallels something else I am familiar with called notecarding from a book called Big Java (3rd Edtion) by Cay Horstmann.

The idea of notecarding is to identify the verbs and nouns inside the system you need to design.  Once you do this, you can group nouns inside of other nouns (where it makes sense to do so) and then associate all actions with the appropriately corresponding nouns.  (You can also put actions inside of other actions: i.e. the action of running may require the action of moveLegs.)

The Use Cases idea seems a little more focussed than note carding, or rather more structured.  Where the notecarding is very general, the notion of Use Cases seems to mimic the task flow - it sort of plugs the designer into any given point in the design and says "what happens here?"

The other nice thing is that a use case seems to be generalizeable.  In other words, I can talk about a very specific use case or a very broad use case (i.e. the entire use of the system from beginning to end),  albeit I think in the broader use of the notion of use case one might run into some problems with Poka-Yoke, but I think that's the fun of design.  E.g. You might say that "At the beginning, the user sees the Web site and can access all of his tax information.  At the end, he has filed his taxes."  This is an informative use case to me; it describes in general what the design should accomplish.  But for more complex Websites with many possible beginnings and ends, it gets a little fuzzy.

So, I think it just depends on your style and how you want to approach designing the "hows" of the design.