RE: w/%editc() - Qualified Names

[jpcarr@xxxxxxxxxxxx]

>The PREFIX keyword now allows you to use real field names in your
>database, and if you like, assign a prefix to those field names at
>compile time. It's pretty cool.
>
>Bob Cozzi


I like the Prefix too.   Especially with the same file in the PGM more than
once.

However,  The Prefix, while neat for Programmers,  still leaves End Users
who use (say) Query/400 to have to qualify names,  and forces the use of
correlated names in SQL.

While you can do that in both places,  It's a tedious effort that is not
required if the DB is named that way.

I look at it from this view.

Customer# has one root definition,
All files that Contain Customer# inherit the properties of the one
definition(Class).   I like each "Instance" of Customer# to be inherently
qualified by the name of that Instance to specify which SubClass that
instance belongs to.

Root Definition

Cust

Customer Number in file XXPFZZZ
Will have the "Qualified Instance" Name of ZZZCUST

Customer Number in file XXPFAAA
Will have the "Qualified Instance" Name of AAACUST

WHERE EVER I see ZZZCUST or AAACUST I know which instance I am talking
about.
AND  it does NOT vary from program to program, display file to display
file.

Prefix names are/can be different in each program. Without an External
definition
forced upon it.

Where XX is the System name (like OE for order entry, PP for Production
Planning, etc)
PF is a part of the Object Naming meaning Physical File (DF, CL, RP, LF,
etc)

Just opinions folks.  However I never worry about field name collision. And
I have a system wide naming consistency.  Whether in RPG programs or
Query/400's or talking to end users using a graphical query tool.

The qualified Instance Name is the Qualified Instance Name.

What ever works for you.

John Carr