Re: Create Table Like - Format Level Check

[CRPence <CRPbottle@xxxxxxxxx>]

  Is the "original TABLE" a file that was just-created, or was that 
file existing on disk having been created sometime in the past?  If the 
latter, does the problem go away when using a newly-created version of 
that file; i.e. the re-created "original TABLE" gets a different record 
format level identifier?  If so, this could be a side-effect of a 
correction for the VARCHAR issue discussed last year, but for which no 
APAR was opened, and the PMR was simply closed with IBM saying they were 
not going to correct the issue.  That was an issue with a difference 
between how DDS CRTPF using VARLEN and DDL CREATE TABLE using VARCHAR 
generated some /bits/ of the record format that are part of the hash. 
The described effects seem to be closely related to that same issue:
http://archive.midrange.com/rpg400-l/201202/msg00051.html

  FWiW: the long column names [or ALIAS names] are not part of the 
hash, nor is a primary key.

Regards, Chuck

On 23 May 2013 13:16, Anderson, Kurt wrote:
> I have a file created in SQL using Create Table.
> I have another file that is defined like that table, yet the format
> level IDs are different. I have walked through the DSPFFD output and
> the files look identical.
>
> The second file I have created as:
> CREATE TABLE CdrMiscP2 LIKE CdrMiscP
>    INCLUDING IDENTITY COLUMN ATTRIBUTES
>    INCLUDING COLUMN DEFAULTS
>    INCLUDING IMPLICITLY HIDDEN COLUMN ATTRIBUTES
>    INCLUDING ROW CHANGE TIMESTAMP COLUMN ATTRIBUTES
>    RCDFMT CdrMiscF;
>
> Format Level ID: 3F61F0D0B2DB3
>
> I have also ran the command without the INCLUDING (I had only added
> that in later after reading online to try including those) and got
> the same Level ID.
>
> The original create table:
> <<SNIP>> MiscValue  VarChar (100)  ALLOCATE(30) <<SNIP>>
>
> Format Level ID: 3FE1F0D0B35B3
>
> If I do a CRTDUPOBJ the Format Level ID comes out the same. I
> suppose I can create the file using CRTDUPOBJ, but I prefer to have
> source for the files we use. (I guess it can be source saying, "Hey,
> create it with CRTDUPOBJ.")
>
> I originally ran into this issue with a much larger file (that did
> not have a Primary Key). I was testing on a smaller file to see if I
> still ran into the issue and I did.
>
> I'm on 7.1.
>
> Should I open a ticket with IBM?  Am I missing something?