Andrew Dunstan <andrew(at)dunslane(dot)net> writes:> I take this as confirmation that calling pg_usleep is the Right Thing (tm).> Here's the patch.
Applied.
regards, tom lane