>> For what it is worth, I work this way:  I keep everything for a project in a
>> folder.  When I start to rework or improve, I make a copy of the folder and
>> set it aside.

For what it's worth, I work this way:

I first create a project folder (i.e., myProject 1.0). When I have the
program doing something of value, I then duplicate the entire folder and
only rename the project folder (i.e., myProject 1.1). I also add "read me
files along the way to describe specific differences/featrues/problems.

For specific functions/features/aspects, I create stand-alone programs that
I add as needed. By making them stand-alone, I build a library.

I never alter the internal code of any project to reflect any version
number, other than for client beta versions.

At the end of any project, I have literally dozens of "past" project files.
I keep what I think are important and trash the rest. And, as a result of
my methodology, I have numerous stand-alone functions that can be used for
future projects. It's a constantly growing process.

tedd

PS: Your mileage may vary.


___________________________________________________________________
<mailto:tedd@...>	               http://sperling.com/