Re: Conceptual, Logical, and Physical views of data
Date: Tue, 30 Aug 2005 22:44:41 -0400
Message-ID: <Kumdnd_JZeMFhIjeRVn-rQ_at_rcn.net>
Kenneth Downs wrote:
> BobTheDataBaseBoy <"xxx at rcn dot com"> wrote:
>
>
>>David Cressey wrote: >>
>
>
> Is there anything they know that is worthwhile in your world?
>
>
sure. they should know the business. they shouldn't be allowed to specify a data model. anymore than a plumber should do neurosurgery. they should know what data they need at each discreet point in the process about which they are Expert. they should know what those points are. they should be able to describe how to get from one to another; or not. that is, the spreadsheet/word processing model vs. the defined process model. both work. not interchangably, however.
of the younger crop of Gurus, Spolsky has written about the conflict between ease of use and ease of learning (which is how he describes it. it's the same thing). this conflict has implications for both code and data structures. unless one has built to the defined process model, it's very existence in the here-and-now might not be evident.
years ago, i worked on some BOM systems. such software is by its nature pretty general; both as code and data structure. the user/expert decides how his/er business works by defining the materials and operations and such. this is SME territory. deciding how the data structure looks isn't.
summary: so long as the conceptual model is restricted to defining the data needed, OK.
BTDB Received on Wed Aug 31 2005 - 04:44:41 CEST