Re: [agileDatabases] Re: Agile Methods, Universities and Misc Thoughts
- From: "Grant F."
Sent: Tuesday, February 10, 2004 7:27 AM
Subject: [agileDatabases] Re: Agile Methods, Universities and Misc Thoughts
> --- In agileDatabases@yahoogroups.com, "jm04469" <james@w...> wrote:While this is kind of off topic for this group, I'd like to comment
> > A coworker mentioned to me that agile will never gain critical mass
> > because it is not taught in universities. Students from day one are
> > taught to think in heavyweight, command and control (PROFESSOR),
> > comprehensive documentation is more important than working software
> > (research papers) and so on. Corporate America is doomed from the
> > start unless universities fundamentally change their approach.
> > Waiting to teach Agile at the Master's level is somewhat
> With all due respect to your co-worker, what a crock. VB took over
> corporate America despite active resistance from academia.
> Fortunately, businesses drive much more software innovation than
> universities do. If not, as developers & dba's, instead of being in
> the constantly changing, constantly expanding business environment
> that we are in, we'd be in a stagnant, slow to change environment.
> That sure sounds a lot more like universities to me (our latest
> intern was surprised that more than half our work was in SQL Server
> since his professors had told him that SQL Server was just a flash
> in the pan).
that there are examples on both sides, and the example of
the universities driving the almost universal adoption of a
very bad language almost by accident is quite well known.
That example is, of course, C, which was adopted because
the universities had a need for an operating system that came
with source for their OS classes, and C came along with the package.
Saying that something will or will not be adopted because
it is or is not taught in the universities is a rather hazardous
method of prediction, at least in my experiance.
- I know this is a little late in the day... But I have only just found
this forum :P
>>In light of that, what should I make of a highly normalized databaseAs pointed out by the above text, it should be quite possible to stick
>>table structure that is organized more according to what role the data
>>plays than grouped in an object-like manner?
>My experience is that a highly normalized DB schema often maps very
>to the data aspects of a well-designed object model.
to a one object/table design as long as your object model is fine
If you need to work remotely with the info in your objects (reducing
calls etc) consider mapping parts of the object graph into data
transfer objects and using a service layer.