By irritating the majority of people who could give you meaningful
responses by requiring that they be self-proclaimed or certified
"experts" you assure that you won't get many responses. It is
generally accepted that doing a good job of describing and posting
your issue will be the best way to get qualified responses.
I've been called an expert many times over the years. Note that it's
other people's opinion, not mine. I know that I don't know all there
is to know about anything.
Software Engineering is similar to any other kind of engineering: the
desired outcome must be known before the building begins.
You should have a clear understanding of what you intend to build
before you even approach the computer. I'll list some of the steps I
use. Note that the degree of rigor applied varies with the
anticipated user base for the application.
Problem Statement: this is a statement of the problem that I am
addressing or going to solve. It can include both specific and
general issues.
Product Specification: This is a description of the solution I'm
building that will resolve each issue raised in the Problem Statement.
The two items above are always completed before I begin any "code".
If the current project is for a client, I run the above two items by
all interested parties and get them to sign-on to their accuracy and
completeness.
Functional Specification: If you are comfortable with your
capabilities in using the proposed tools (Access in this case) you
might also complete the functional spec before touching the computer.
The Functional Specification contains the descriptions of the
functions that will be used to achieve the solutions
When you are satisfied with the above three elements you have a map of
your application. It is very common that while addressing issues in
the later parts of the project you find it necessary to go back to
earlier parts to make changes. The more you use the same tools in the
same process the less that happens.
At this point I try to imagine the users going through their jobs
using the tool and try to identify and isolate the entities in the
application. When I'm satisfied that I have all that I can imagine I
design the tables, naming each table for the singular of the entity
type. The table wizard gives you ample space for field descriptions.
You'll do well to use it.
Next come the forms with their queries. Because you've done your
homework above, you'll have no doubts about where you are in the
process.
Duane Hookom made a nifty tool, Code Stuffer, available several years
ago. That tool writes a skeleton in a procedure that will put in your
name, the current date and an error handler. It works for both subs
and functions.
Using the steps above you should be able to describe what you're doing
in an Event handler sub or in a given function. Describe just what
this particular piece of code is intended to do. Don't describe the
part(s) of the application that call the procedure. If your current
procedure calls another procedure, put in a line of comment above the
call that describes what you're doing.
That's about as far as your question went ...
HTH