Re: DSPF Record format ID differences

On 01-Jun-2017 14:54 -0600, Mark S Waterbury wrote:

[...]

It is possible that one version was created before some IBM PTF(s)
affected QDBLVLID, the routine that computes record format level
IDs.

The "DB" in QDBLvlId indicates the OS DataBase feature; i.e. for Data Base Files, not Device Files for which the OS component/feature is "DF". I recollect that the DB feature functions for, is invoked for, *only* the creation of database files, and that device files have their level-hash calculated within their corresponding /create/ program [or, but I think not] by a similarly separate level-identifier program.

This V6R1 PTF and its APAR are particularly troubling in the lack of
any detail:

PTF SI28827 OSP-DB-QDBLVLID change APAR SE30038

No real details are visible in either the APAR or the PTF cover
letter. Hmm....
[...]

I detest a dearth of details in APARs; I almost always gave an [attempt at a good] explanation, but seems to me, many developers prefer revealing [nearly] nothing. Perhaps so as to help avoid the fix being called a "defective PTF"; i.e. with nothing described [as to the effects], clearly nothing could be wrong [no matter the effects] :-Q

Just a guess, that APAR/PTF was per the following, which apparently should have impacted only SQL INDEX.?:
(https://archive.midrange.com/midrange-l/201006/msg00763.html)[6.1 Create Index - Format Level Identifier]