On Wed, Jan 17, 2024 at 04:40:10PM -0800, Junio C Hamano wrote:
> "John Cai via GitGitGadget" <gitgitgadget@xxxxxxxxx> writes:
>
> > From: John Cai <johncai86@xxxxxxxxx>
> >
> > Move t3210 to t0602, since these tests are reffiles specific in that
> > they modify loose refs manually. This is part of the effort to
> > categorize these tests together based on the ref backend they test. When
> > we upstream the reftable backend, we can add more tests to t06xx. This
> > way, all tests that test specific ref backend behavior will be grouped
> > together.
>
> So, ... is the idea to have (1) majority of ref tests, against which
> all backends ought to behave the same way, will be written in
> backend agnostic way (e.g., we have seen some patches to stop
> touching the filesystem .git/refs/ hierarchy manually), and (2) some
> backend specific tests will be grouped in a small number of test
> script files for each backend and they all will use t6xx numbrs?
>
> OK. Sounds like a good plan to me.
Yes, that's the plan. The backend specific tests will be free to also
exercise filesystem-level behaviour in order to pin down that things
work as expected. But once their behaviour is nailed down all other
generic tests should refrain from doing that to the best extent possible
and instead use Git commands to do their thing.
> > Signed-off-by: John Cai <johncai86@xxxxxxxxx>
> > ---
> > t/{t3210-pack-refs.sh => t0602-reffiles-pack-refs.sh} | 0
> > 1 file changed, 0 insertions(+), 0 deletions(-)
> > rename t/{t3210-pack-refs.sh => t0602-reffiles-pack-refs.sh} (100%)
Is there a reason why you picked t0602 instead of the not-yet-taken
t0601? If it's only because I use t0601 in my reftable integration
branch then I'd like us to pick t0601 here instead to avoid a weird gap.
I'll adapt accordingly and rename the reftable tests to have a t061x
prefix in that case so that they are nicely grouped together.
Patrick
Attachment:
signature.asc
Description: PGP signature
