On Thu, Jun 04, 2020 at 10:59:05AM -0700, Andres Freund wrote:
> What's the point of that change? I think the differentiation between
> heapam_handler.c and heapam.c could be clearer, but if anything, I'd
> argue that heap_get_latest_tid is sufficiently low-level that it'd
> belong in heapam.c.
Well, heap_get_latest_tid() is only called in heapam_handler.c if
anything, as it is not used elsewhere and not publish it. And IMO we
should try to encourage using table_get_latest_tid() instead if some
plugins need that. Anyway, if you are opposed to this change, I won't
push hard for it either.
--
Michael