RE: SQL/DDL Column renames

Back on AS/400 if you wanted to add a column, etc you had to copy the file
to a temporary file. Create a new file and then copy the data back.

Then they modified the CHGPF command allow you to modify the DDS source,
run CHGPF and it would do it all "in place". Been this way for decades.

Therefore Mark Waterbury's suggestion bears weight.

Let's use an example.

Let's say we have this DDS.

UNIQUE
R ABR
MYKEY 11A
MYDATA 10A
MYDATE L
K MYKEY

And we compile it. We'll end up with:
Format Fields Length Identifier
ABR 3 31 2932933592714

Now let's modify the DDS to

UNIQUE
R ABR
MYKEY 11A ALIAS(I_AM_THE_KEY)
MYDATA 10A ALIAS(I_AM_DATA)
MYDATE L ALIAS(I_AM_A_DATE)
K MYKEY

But instead of "compiling" it we use:

CHGPF FILE(ROB/DEAB) SRCFILE(ROB/QDDSSRC) SRCMBR(AB)

The record format level identifier will not change.
Format Fields Length Identifier
ABR 3 31 2932933592714

DSPFFD ROB/DEAB

Field Type Length Length Position Usage Heading
MYKEY CHAR 11 11 1 Both MYKEY
Alternative name . . . . . . . . . . . . :
I_AM_THE_KEY
Coded Character Set Identifier . . . . . : 37
MYDATA CHAR 10 10 12 Both MYDATA
Alternative name . . . . . . . . . . . . :
I_AM_DATA
Coded Character Set Identifier . . . . . : 37
MYDATE DATE 10 10 22 Both MYDATE
Alternative name . . . . . . . . . . . . :
I_AM_A_DATE
Date Format . . . . . . . . . . . . . . . : *ISO
Coded Character Set Identifier . . . . . : 37



Rob Berendt