>>If possible ... how painful would it be to do?
I'm yet to do that part, so I guess I'll find out.
> Actually it looks like you'd better, because for example aclupdate
> assumes there's only one entry for a given grantor/grantee pair.
OK, many thanks for the prompt reply :)
> BTW, are you sure Fabien did not already solve this problem in his
> pending patch?
You mean schema ownership? I thought that was just upon the first
connection to a database or something? I'm using schemas as my first
case for fixing OWNER TO commands and acls...
Chris