On Thu, Mar 25, 2010 at 7:37 AM, aki <aki275@googlemail.com> wrote:
@Frankline, I really appreciate your input. In the original thread I
had asked if anyone uses a particular plan or methodology to derive
variables and I gave an example of a design stage that I'm looking at.
The design stage can only come around if enough research has gone into
the requirements stage. Trust me, it is not that simple and a steeep
learning curve, variables is one tiny portion of the program and
placing them correctly within the code is very important. Imagine you
were a software development consultant and given a blank page to
comeup with project requirements, flow and functionality. From what
you write into the project is a blueprint for the development phase of
the project. Its like a blueprint for a house, you cannot build a
house first then make its blueprint plans later. For a house, you can
derive many variables. Some like electrical sockets, plumbing outlets,
drainage, volume and storage of water just by knowing what each floor
is going to do. The blueprint for an office block will be different
from a residential one.


Anyway, I'm still researching and hopefully be able to comeup with an
answer that makes sense on paper first. :-)

Rgds.


Aki,

Software development isn't quite like the other engineering disciplines. In software development, "the source code is the design (and not the product)". The whole process of writing your application actually involves designing it. Things change all the time as you go along.

I'd say you shouldn't worry too much about having a perfect "design" on paper before writing a single line of code. You just need to start with a general "architecture" of the system, which will also change as you flesh it out in code.

Check out these pages at the Portland Pattern Repository Wiki: <http://c2.com/cgi/wiki?WhatIsSoftwareDesign>, <http://c2.com/cgi/wiki?TheSourceCodeIsTheDesign>, and <http://c2.com/cgi/wiki?TheSourceCodeIsTheProduct>.

Joseph.