FWiW the original scenario was non-keyed PFA with only the keyed
LFA, and an /exact/ /copy/ of this DBFile network in a separate
library. So what was proposed by the OP, save *with access paths*
and a restore verified to have restored *with that access path*
intact would have the DB files and access paths the closest possible
Evan Harris wrote:
Re: Save and restore being identical - not necessarily true.
Search on "implicitly shared access paths".
Alan Campin wrote:
The trick is that this can be modified if you have reuse deleted
records turned on. In that case, the result would be random.
As far as the order in another library would depend on how it got
copied. A save and restore should be identical.