On 2019-Apr-04, Andres Freund wrote: > I'm totally not OK with this from a layering > POV. CheckAndSetAllVisibleBulkInsertState is entirely heap specific > (without being named such), whereas all the heap specific bits are > getting moved below tableam.
This is a fair complaint, but on the other hand the COPY changes for table AM are still being developed, so there's no ground on which to rebase this patch. Do you have a timeline on getting the COPY one committed? I think it's fair to ask the RMT for an exceptional extension of a couple of working days for this patch. -- Álvaro Herrera https://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
