Re: [PATCH v1 1/4] t: introduce WITH_BREAKING_CHANGES prerequisite

From: Junio C Hamano <gitster@pobox.com>
Date: Mar 11, 2025, 17:07 UTC
Message-ID: <xmqq4izz4hte.fsf@gitster.g>
In-Reply-To: <Z9AzKoWLeOJ63j_E@pks.im>

Patrick Steinhardt <ps@pks.im> writes:

Show 20 quoted lines

> On Mon, Mar 10, 2025 at 04:16:49PM -0700, Junio C Hamano wrote:
>> diff --git a/t/test-lib.sh b/t/test-lib.sh
>> index 9001ed3a64..12fe82f660 100644
>> --- a/t/test-lib.sh
>> +++ b/t/test-lib.sh
>> @@ -1862,6 +1862,11 @@ test_lazy_prereq CURL '
>>  	curl --version
>>  '
>>  
>> +test_lazy_prereq WITH_BREAKING_CHANGES '
>> +	test -n "$WITH_BREAKING_CHANGES"
>> +'
>> +
>> +# DEPRECATED; DO NOT USE THIS IN NEW TESTS
>>  test_lazy_prereq WITHOUT_BREAKING_CHANGES '
>>  	test -z "$WITH_BREAKING_CHANGES"
>>  '
>
> Do we maybe want to state that this can be removed once the next release
> cycle is over?

Perhaps. As 'seen' is pretty-much closed at this point and there is nothing in flight that uses WITHOUT_ variant in there, v2 of this series can just do without it, which may be simpler.

> I find it to be a bit more actionable when stating hard
> dates after which something can be dropped

True, that is a good strategy for a transition that takes longer time.