Ed Leafe wrote:
>       I've worked for many clients who made claims like "we will never need 
> more than one phone number per customer" or "part numbers will never be more 
> than 4 places" or "640K will be enough for anyone". (OK, that last one wasn't 
> from one of my clients!).

It's become a running joke between myself and my principle contact at my main 
client. 
They keep constraining what I do based on what they think will get the job done 
faster (more cheaply). So they say, for example: "There will never ever be the 
need 
for more than one t-post" which means I don't need to set up a 1:many 
relationship or 
really do any normalization at all.

I get the job done pretty quickly and it works well. We maintain it and enhance 
it 
over the years as needs change and business rules flex with the times. This was 
1998. 
Flash forward to 2006 and customers are now wanting 2-4 t-posts plus horizontal 
t-posts (which were another "we'll never need this" option back then).

Now we need to either graft the new requirements onto the existing framework, 
or 
redesign the whole module from scratch, including converting the old data to 
the new 
model, testing, and deploying in the most non-distruptive way possible.

IOW, allowing for the 1:many design from the outset, even though the need 
wasn't 
foreseen, and even though the need for an infinite number of t-posts would 
never 
materialize, would have been the best decision for them, because making such a 
fundamental change later on takes way more time and energy than designing it in 
at 
the outset.

There are many such examples in my client relationships. For this one client, I 
make 
a point of finding the original email where she says "we'll never need this" 
and 
replying to it with a wink. But she still never learns. ;)

Paul
_______________________________________________
Post Messages to: [email protected]
Subscription Maintenance: http://leafe.com/mailman/listinfo/dabo-users
Searchable Archives: http://leafe.com/archives/search/dabo-users
This message: http://leafe.com/archives/byMID/[email protected]

Reply via email to