(I've added Thomas Munro to the thread.) On Fri, Jan 17, 2025 at 02:00:49PM +0900, Michael Paquier wrote: > XLogRecGetFullXid() has been introduced in 67b9b3ca3283 back in 2019, > but as far as I can see this has never been used in the code and this > is used nowhere in the core code. > > I have looked at Debian's codesearch and also looked at traces of it > on github without seeing it being used anywhere. Knowing that this > was originally intended for a hypothetical undo log patch back then, > for which no work has been done for years, is there any point in > keeping this function in core? > > This issue has been raised on a separate thread, where Noah has sent a > patch to consolidate a bit some epoch calculations for > FullTransactionIds, around here: > https://www.postgresql.org/message-id/z4i6mbulzxjk1...@paquier.xyz > > Removing it would have the benefit to do a bit less refactoring for > some of the work of the other thread, and this removes all traces of > -DFRONTEND in xlogreader.h. ;)
Seems reasonable to me. I think the counterargument is that folks developing new AMs should use this [0], but if no such users have materialized in several years, then maybe that's no longer a concern. [0] https://postgr.es/m/CA%2BhUKG%2BmLmuDjMi6o1dxkKvGRL56Y2Rz%2BiXAcrZV03G9ZuFQ8Q%40mail.gmail.com -- nathan
