Re: Data Flow Diagrams

["Anthony Markatos" <tonymar -at- hotmail -dot- com>]

Chris Kowalchuk said:

Let me first say that I too, as a tech writer who sometimes helps
develop software specs, typically resort to data flow diagrams. Whenever I 
do, the programmers are amazed.....

Tony Markatos responds:

Wow! - there are at least two others on this planet who create DFDs (you and 
Win).  What a feeling! - I am not alone!   :-)

Chris Kowalchuk said:

[Computer programmers do not create DFDs because] ........Computer 
programming has a long history of being approached from a "just make it 
work" attitude....

[But (software) engineers do create DFDs because] ..... an engineer who 
could not do a proper spec or map out a process would probably not make it 
out of engineer school.

Tony Markatos:

First let me emphasize that I am talking about REAL Data Flow Diagrams, not 
"poor-man" imitations - often called things like Process Flow Diagrams, 
Process Diagrams, Flow Diagrams, etc.

Now to comment on your comment: The older I get, the truer rings the old 
saying "Believe none of what you hear and only half of what you see."   And, 
while your comment logically makes sense, it contradicts what I have seen.

I have, as a business analyst type, lead (larger-scale) requirements 
specification projects which included "engineers" having Phds in IT from MIT 
and Stanford (and a Harvard MBA type, but he don't count).  These guys were 
sharp, like real sharp - except when the need was to create Data Flow 
Diagrams.  Then, mysterious things would always happen, and I would, 
inevitably, be the only one doing them.  They were "hungry" for my output, 
but I was the only DFDer.

At one(a large Southern California aerospace) company, I was a member of an 
elite DFD "gruppie" department.  (I was drafted into it - I don't "gruppie" 
anything.)  These people ate, slept, and breathed DFDs.  They constantly 
talked about the benefit of them.  They developed a large in-house library 
on them.  They taught evening courses in them.  One guy - a real 
"borne-again Christian" type - would often loudly quote famous passages from 
DeMarco and Yourdon (famous DFD authors).  But, funny thing, when it came 
time to actually do DFDs (like for work); I was, once again, alone.

So maybe you have worked with engineers who create DFDs.  If so, then I have 
only worked poorer-quality companies (Note: I have worked coast-to-coast, 
large-medium-small, and for several international consulting companys.)

Don't get me wrong Chris, I am definetly pro-DFD (and other SSAD 
techniques).  Properly used they are, as it sounds like you have discovered, 
very powerful.

I think that a major reason for there gross underutilization has been the 
way they were taught.  Example: Yourdon used to say that DFDs without an 
accompanying data dictionary were merely "rough-sketches".  This scares 
people.   For in the final analysis, Structured Systems Analysis is merely 
rigorously enforced honesty.  And the human capacity for honesty is limited 
- increasing I think, but still limited.

Tony Markatos
(tonymar -at- hotmail -dot- com)




______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com