DFD as TA, was Re: Getting up to speed on UML (Was: Re: Documentation Correctness...)

[Ned Bedinger]

Sorry about busting in on Stuart's UML query, but this appearance of the 
DFD hydra reminded me of something last week that I thought was 
important, but was too busy to post last week.

Richard Lewis wrote:

>   Bottom Line: Data Flow Diagrams = Task Analysis.


Thankfully, Steve Jong took the time to put Data Flow and Task Analysis 
into parallel perspective for us last week (see 10/17/2007, from 
stevefjong at comcast.net, "Re: Documentation Correctness was Re: How many 
levels of indents and	heads are reasonable?").

I believe that, with his thinkable example, Jong completed the 
intractable "Bottom Line" thought from alias Richard Lewis.

IMHO, we can all now be free to believe that Data Flow Analysis has a 
place in Task Analysis, and even that task Analysis is not complete 
without Data Flow Analysis. But still, a major concern comes with the 
claim that the DFD is sufficient for any task analysis, and that is that 
we cannot count on our being able to infer all actor steps from the data 
flow. The inevitable conclusion is that a DFD would not suffice for all 
Task Analysis.

Many thanks to stevefjong for tackling the need for an example to help 
us think about this. And to alias richard lewis for surviving and 
sharing his/her discovery of data flow diagrams with us--I think you 
must be like this:

If you want to build a ship
don't herd people together to collect wood
and don't assign them tasks and work,
but rather teach them to long for the
endless immensity of the sea.

Antoine-Marie-Roger de Saint-Exupery.

Flames to /dev/null.

Ned Bedinger
doc at edwordsmith.com