Pavel Stehule <pavel.stehule@gmail.com> writes: > When function is overwritten, then regproc result contains schema, although > it is on search_path
There's no "fresh regression" here, it's done that more or less since we invented schemas. See regprocout:
* Would this proc be found (uniquely!) by regprocin? If not, * qualify it.
git blame dates that comment to commit 52200bef of 2002-04-25.
Admittedly, qualifying the name might not be sufficient to disambiguate, but regprocout doesn't have any other tool in its toolbox, so it uses the hammer it's got. If you're overloading functions, you really need to use regprocedure not regproc.
It is false alarm. I am sorry. I shot by self. Thank you for explanation
