From: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Date: Sat, 25 Sep 2010 17:05:41 GMT
Subject: Re: [RFC PATCHv2 16/16] Add missing &&'s throughout the testsuite
Message-ID: <AANLkTi=0ByRX2E_MDE7eeb4LR=bKXeah1xD1MRpW1+Jk@mail.gmail.com>
In-Reply-To: <AANLkTincF9qu62w1+WLcidxFAYb8iqf_Jw+ETwuy=rg9@mail.gmail.com>

On Sat, Sep 25, 2010 at 16:50, Elijah Newren <newren@gmail.com> wrote:
> On Fri, Sep 24, 2010 at 5:18 PM, Ævar Arnfjörð Bjarmason
> <avarab@gmail.com> wrote:
>> On Fri, Sep 24, 2010 at 22:22, Elijah Newren <newren@gmail.com> wrote:
>>>        (
>>> -               unset GIT_CONFIG GIT_WORK_TREE GIT_CONFIG
>>> +               unset GIT_CONFIG GIT_WORK_TREE GIT_CONFIG &&
>>
>> These are all bad, see c289c315c2bfcd5fde99f9e4dbb109f89e29dfce for
>> why. If it's possible maybe we should have a new:
>>
>>    test_returns_whatever unset FOO &&
>>
>> To make this more self-documenting
>
> Isn't that what test_might_fail does (even if documented as being only
> for git commands)?

Yeah, maybe that should be used. But it might also be better to have a
different one for OS commands so they can be easily grep-ed for. I.e.:

    easily_greppable_string unset FOO &&

Instead of:

    unset FOO;