I wrote:
> This leaves us with some pretty unappetizing choices about what
> to do in v18:
> 1. Try to emulate the proposed HEAD fix.
Actually, a preliminary look suggests that we can approximate
that quite well, because v18's create_unique_path() is already
responsible for preparing the list of columns to be de-duplicated
by a UniquePath. So we can stick the same logic about whether
make_pathkeys_for_sortclauses believes that each column adds something
into create_unique_path(). Unlike the case with HEAD, that doesn't
overlap with any useful work the function was already doing, but
I still think that the cost will be negligible in normal cases.
I'll try to have a patch along those lines by tomorrow.
regards, tom lane