Re: [PATCH v2 14/14] test: pull-options: revert unnecessary changes
- From
Felipe Contreras <felipe.contreras@gmail.com>
- Date
- Dec 5, 2020, 01:28 UTC
- Message-ID
- <CAMP44s1-Q=3AN8o03gncQZtExewYqwvKos1o1k1Ap8CcGKn23A@mail.gmail.com>
- In-Reply-To
- <CABPp-BEfxvkuVsDR9duwg1QCMxV9O4RqGF_rm2NuiStczggQoA@mail.gmail.com>
On Fri, Dec 4, 2020 at 5:49 PM Elijah Newren <newren@gmail.com> wrote:
Show 20 quoted lines
>
> On Thu, Dec 3, 2020 at 10:16 PM Felipe Contreras
> <felipe.contreras@gmail.com> wrote:
> >
> > Commit d18c950a69 changed these tests, but it's unclear why. Probably
> > because earlier versions of the patch series died instead of printing a
> > warning.
>
> Another case where log --pretty=reference notation would be more
> helpful in referring to commits.
>
> Looking up that commit, I see that commit made a number of other test
> changes which seem to be better motivated by the commit message and
> code changes. I wonder if we could make that clearer in the commit
> message somehow. Perhaps
>
> Commit d18c950a69 ("pull: warn if the user didn't say whether to
> rebase or to merge", 2020-03-09) changed a number of tests in t5521
> and added some new tests in t7601, but it is not clear why the changes
> in t5521 were made...Fine by me.
I had a mind to do some mail archeology and find the likely reason myself, but we all have limited time on this Earth.
-- Felipe Contreras