Jelte Fennema-Nio <postgres@jeltef.nl> writes:
> On Fri, 7 Mar 2025 at 11:26, Andreas Karlsson <andreas@proxel.se> wrote:
>> Out of curiosity: do you track which method works? I would expect
>> everything to be applied with either git am or patch which can be
>> applied with git apply making git apply technically unnecessary.
> I think we need all of them...
> - git apply is unable to apply the patch in 5272 on top of main
> - git am, works with that 5272, but cannot apply any patch not created
> by git format-patch (for instance 5522)
> - I've been unable to get patch(1) not to return a failure if the
> patch is already applied. So applying e.g. 5522 twice. git apply
> handles that fine.
Hm, don't you *want* a failure if the patch is already applied?
regards, tom lane
