On Sat, 16 May 2026 at 14:39, Tom Lane <[email protected]> wrote: > > Dean Rasheed <[email protected]> writes: > > Attached is a v2 patch with those changes, plus a little more tidying > > up of the regression tests. > > LGTM.
Thanks for looking. > > Arguably, this is a bug fix, but given the lack of prior complaints, I > > think we should apply this to HEAD only, like 6498287696d, meaning it > > will only appear in PG19. > > Agreed on no back-patch. It seems sensible to slip this into v19 > rather than wait for v20, since the corresponding fix in corr() > is new as of 19. But since that's a judgment call, maybe you should > get the concurrence of the RMT. Yes, that was my thinking. Does anyone from the RMT (cc'ed) have any objection to slipping this into v19? Regards, Dean
