From: Junio C Hamano <gitster@pobox.com>
Date: Tue, 26 Jan 2016 20:24:32 GMT
Subject: Re: [PATCH v2 19/19] mingw: do not bother to test funny file names
Message-ID: <xmqq1t94ccn3.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <CAPig+cQXVwfyaEzf9fR2tFNon-Jq93iPCMwW9gsB8BnUVv5p3A@mail.gmail.com>

Eric Sunshine <sunshine@sunshineco.com> writes:

>> +if ! test_have_prereq MINGW && touch -- 'tab   embedded' 'newline
>> [...]
>> +test_have_prereq !MINGW &&
>
> Where negation is concerned, is there a non-obvious reason that this
> patch sometimes says:
>
>     ! test_have_prereq MINGW
>
> and sometimes:
>
>     test_have_prereq !MINGW
>
> ? Is one form preferred over the other?

The latter may be preferrable as it can be extended more easily,
e.g.

	test_have_prereq !MINGW,SANITY

without having to cascade "&& test_have_prereq ANOTHER_PREREQ" in a
verbose way.

It is a minor thing that is not worth bothering to "fix" from one to
the other style once it is committed to my tree, but it may be worth
sticking to one style in new code.