Posted by Reilly Atkinson on August 17, 2000 at 18:39:41:

In Reply to: Re: what's the way to represent tacik knowledge of an enterprise? posted by John Tieso on August 10, 2000 at 21:46:18:

John -- Small world indeed. I worked with Ross and Softech in the mid 70s primarily in marketing and selling SADT, their Structured Analysis and Design Technique,to the Feds. It was a tough sell, and we never quite got the cigar -- the possible clients ranged from standard civil servants to a PhD physicist, the Chief Scientist of EPA ( also a friend )-stodgy to sophisticated. People are generally suspicious of new ways to think, unless there is overwhelming evidence that the new ways are highly sucessful, and worthy of new jargon and buzz words.

My take then on SADT was that it was good for designing relatively simple data processing systems, dealing with basic transactions, some interrupt-driven systems, monitoring a persons bank accounts including batch updates and so forth.

But I found it virtually useless in my work at the time, which was highly focussed on complex simulation models -- of population and household flows from neighborhood to neighborhood, of local housing markets, or product and financial flows in a logistics system with a probabilistic driver.
With these types of problems, SADT was too confining, as would be any varient of SADT and probably any technique that atempted to provide "automaticthink". All these things are very much "in the box". But these problems eventually yielded in an unstructured way to my physics bag of tricks.

It was basic to my training as a physicist that my job was to find the right tools to solve the problem, not to fit the problem into a particular tool. We were trained to be "out of the box", to be inventive, and develop our own tools when we saw fit.

Quite honestly, I think people in training for most anything need to be taught less structured stuff, and more about actually solving problems, particularly story book algebra problems. That way the student can build up their own bag of tricks, rather than someone else's. After a good bit of this, it's OK to learn structured stuff, and the student can then fairly judge the stuff's utility. I guarantee that a couple of years of this type of hands-on training, and the students will be way out of the box -- and, possibly, hard to manage.

Regards,
Reilly

---

Follow Ups:

• Re: what's the way to represent tacik knowledge of an enterprise? John Tieso 21:54:54 08/23/00 (4)
  • Re: what's the way to represent tacik knowledge of an enterprise? Reilly Atkinson 22:52:27 09/02/00 (0)
  • Re: what's the way to represent tacik knowledge of an enterprise? Robert Benjamin 03:49:37 08/24/00 (2)
    • Re: what's the way to represent tacik knowledge of an enterprise? John Tieso 14:43:43 08/26/00 (1)
      • Re: what's the way to represent tacik knowledge of an enterprise? Robert Benjamin 02:19:21 08/28/00 (0)

---

Click Here to Post Follow Up in New Forums

    Knowledge Management Think Tank (New)

  

 Subject: 

Message: