[EMAIL PROTECTED] wrote:
This is the sort of question that can only (I hope) arise in a development and test environment, not in production.
I can arise whenever anyone does what you did; so it's possible in production, but we hope that packagers take steps to prevent that. (When they don't, PEs and SUPs tend to follow.)
Suppose I build a PTF and APPLY it. I have second thoughts, and rebuild it with a considerably different set of elements from the original; RECEIVE; and APPLY REDO. Is the result the same as if the earlier version had never existed and only the later were APPLYed, or is it some sort of merging of the two different versions?
SMP/E won't remove the first PTF's parts before installing the second's. Suppose PTF 1a has parts A, B, and C and PTF 1b has B, C, and D. The result after this scenario will have A, B, C, and D.
Recent unpleasant experience suggests that it merges rather than replaces: elements may appear with the RMID of the PTF, even though they are not present in the later version, only the earlier. How do I undo this mess? I suspect I might as well delete the CSI and rebuild.
No need to be that drastic. Just build a PTF with the parts you want that supercedes the first PTF number. Make sure it contains *all* the parts that were part of either version of that PTF. Following the example above, PTF 2 would have to contain A, B, C, and D; part A would be at the pre-PTF 1a level; and parts B, C, and D would be at the PTF 1b level.
(Before the second APPLY, you could have RESTOREd PTF 1a and then APPLYed PTF 1b to get almost the same result. The difference is that had you done that, the parts would have PTF 1b's RMID rather than PTF 2's RMID.)
This surely provokes a wishlist item: If I attempt APPLY REDO, and there are elements in the target zone with the RMID of the PTF, but not in the PTF itself, SMP/E should report an error and fail the APPLY. M&C should suggest "Programmer Action: Perform RESTORE and only then retry the APPLY."
I'll leave that one to you and Kurt. But I'll note that PTF packagers are a small subset of SMP/E users, so my uninformed opinion is that it's quite possible there are higher-priority things for them to do...
Related question: Is there any query that will show me all elements in a target zone having a given RMID?
I don't recall, but the SMP/E books should say. -- John Eells z/OS Technical Marketing IBM Poughkeepsie [EMAIL PROTECTED] ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
