On Tue, Jan 16, 2018 at 8:56 AM, Robert Haas <robertmhaas@gmail.com> wrote:
> I've seen that before as well.
>
> I have also noticed people complaining about patches that apply "with
> offsets", which also seems like needless nitpicking. If the offsets
> are large and the patch has been sitting around for a long time,
> there's a small chance it could be applying to the wrong place, but
> that is extremely rare. Most patches have small offsets, just a few
> lines, and there is no problem.
+1
The parallel CREATE INDEX patch is something that I've worked on
(fairly inconsistently) for 2 years now. I remember two occasions in
which somebody else changed a function signature for functions that my
code called, and without that causing even a compiler warning after
rebasing on top of these changes (e.g., changing an int argument to a
bool argument). On both occasions, this led to a real bug in a version
of the patch that was posted to the list.
Mechanical detection of problems is great, but there is no substitute
for vigilance. I think that people that complain about stuff like
patches applying with offsets have a false sense of security about
detecting problems mechanically. Rebasing a patch without conflicts
(including seeing a warning about offsets) does not mean that your
patch didn't become broken in some subtle, harmful way. Mechanical
detection is only useful to the extent that it guides and augments
human oversight.
--
Peter Geoghegan
