Hey, Michel, and thanks for the reply.

>Well, there is always a reason why a request to enhance existing infrastructures
>would apply. And, when this happens, this is because the design has to be adjusted,
>a new framework is used, there are new employees doing additional tasks, it was time
>to expand the application, data normalization was applicable, etc.

Exactly what we're going through here, Michel.

>And, whenever this is happening, we always find ourselves in the situation saying "Oh,
>I would have done that differently than what they did!". Well, it's just the way
>it is.

Too true.

>In many of them, some of those designs have been built at first with a minimum set of
>requirements. So, it was doing the job right. But, the company has expanded, the
>technology has moved on and the requirements kept coming in. So, after a while, the
>corporate people sat down and decided it was time to revamp the entire design.

This is what's going on here, too. Personally, I like the VFP/VB front-end
and MySQL on the back-end, with MySQL killing two birds with one stone: It's
a RDBMS, and there's tons of support for scripting languages, such as PHP
and the like.


>For most of my consulting, this is what I do. I don't work on existing code.
>I basically revamp the entire infrastructure by upscaling the application to actual
>technologies, implementating new development standards, doing data normalization and
>menthoring the team so they would be able to move on as soon as I finish my job.

Pretty much what I was hired for, Michel, although we have to fix, maintain,
and add code to an existing production client-server app.


>But, in many situations, while analyzing existing code, yes, it is somewhat difficult
>to understand what is going on there.

In days gone by, I worked for a company called "Interactive Digital Devices."
Basically, we designed mil-spec (practically) joysticks, rudder pedals, etc.
for the video game industry. Lockheed Martin, General Dynamics, Boeing, G.E.,
Bell, and NASA wound up being our best customers :^). One of my duties while
working there was to disassemble 3D flight-sim apps that used only digital
joystick routines (for the Amiga), and then I would patch them to use our
analog products, with all of the patch code being applied to the app in memory.
I had to disassemble hundreds upon hundreds of thousands of lines of other
peoples code, and then figure it all out in reading assembly generated with
C compilers. Now _THAT_ was a pain! :^). So, reading other peoples VFP code
is a cake walk; however, it is still very time consuming, especially when you
think you can rewrite the code faster.



>I found however that there is much more comments than it used to be. I also menthor
>that as well. The more you have in the code, the better it is.

Especially in assembly code! You _have_ to comment the ever-living heck out
of it, or you come back to it in a month or two and say "Huh?" :^). And I'm
talking about your own code, too. :^)

Nice talking with you,

Randall