Re: pgsql: Fix an ancient oversight in libpq's handling of V3-protocol COPY

Simon Riggs <simon(at)2ndquadrant(dot)com> writes:
> That's what the docs say, but Tom's patch also adds lines to handle
> NOTIFY, which is what prompted the question.

I don't believe that code can get executed given the current backend
design. I just put it in because the protocol spec says (and always
has said)

Note: At present, NotificationResponse can only be sent outside a
transaction, and thus it will not occur in the middle of a
command-response series, though it might occur just before
ReadyForQuery. It is unwise to design frontend logic that assumes that,
however. Good practice is to be able to accept NotificationResponse at
any point in the protocol.

regards, tom lane