Royal Pain

Henry Spencer henry-lqW1N6Cllo0sV2N9l4h3zg at public.gmane.org
Sat Jun 19 05:47:20 UTC 2004


On Fri, 18 Jun 2004 cbbrowne-HInyCGIudOg at public.gmane.org wrote:
> And that "well defined specification" part is where things generally go
> SO wrong that the process is made clearly futile to everyone involved.
> They may come up with "well defined specifications" for aerospace system
> components, but those get written by engineers for engineers.

Moreover, truly well-defined specifications are major projects in their
own right, requiring massive efforts.  (They are often done poorly, on the
usual "well, you know what we mean" principle; I'm talking about what it
takes to do them well.)  So much so that it's rarely deemed worth the
trouble. 

> In contrast, business systems tend to have fuzzy descriptions, and are
> anything BUT well-specified.

Worse, the specifications will inevitably change when the prototypes start
running.  Partly because the specs weren't that precise in the first
place; partly because people often can't clearly visualize what they want,
so the specs weren't *right* either; and partly because what they want
*changes* once they get to play with an actual system and get some idea of
its potentials and limitations. 

                                                          Henry Spencer
                                                       henry-lqW1N6Cllo0sV2N9l4h3zg at public.gmane.org

--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml





More information about the Legacy mailing list