Comment 7 for bug 1253163

Mike, thanks for the comment. I only brought up the config question as possible further reason to change to STABLE, since the current immutability could potentially interfere with newer configurations being picked up.

It sounds like we agree to go ahead with the materialization being done in this branch, but with recognition that this is now a long-term change, since the PG behavior may never go back to the old, looser way.

It looks like we also agree that the functions should be lowered to STABLE. That isn't strictly part of this bug, so do we prefer to do that as part of this bug (just being opportunistic), do we open a separate bug for that, or do we roll that change into the bigger picture merge/reingest work being done?