Based on the discussion so far, it seems to me that the sane course of action is to continue to register the grantor, because the standard mandates that it should be there; but ignore the parts where we revoke selectively, because that's a stupid thing to do. So we do deviate, if slightly.
So we will have pg_dumpall do nothing special if the grantor has gone away since granting the privilege. That is, exactly the patch that was submitted, no new code needs to be written. (Maybe a mention in the "compatibility" section of REVOKE is warranted, though I'm not sure). Does anyone object to this course of action? -- Alvaro Herrera http://www.CommandPrompt.com/ PostgreSQL Replication, Consulting, Custom Development, 24x7 support ---------------------------(end of broadcast)--------------------------- TIP 2: Don't 'kill -9' the postmaster
