Re: [PATCH v3 4/5] Makefile: the "cscope" target always creates a "cscope.out"

From: Junio C Hamano <gitster@pobox.com>
To: Taylor Blau <me@ttaylorr.com>
Cc: "Ævar Arnfjörð Bjarmason" <avarab@gmail.com>,
	git@vger.kernel.org, "Denton Liu" <liu.denton@gmail.com>,
	"Felipe Contreras" <felipe.contreras@gmail.com>,
	"Kristof Provost" <Kristof@provost-engineering.be>,
	"Jeff King" <peff@peff.net>,
	"Ramsay Jones" <ramsay@ramsayjones.plus.com>
Subject: Re: [PATCH v3 4/5] Makefile: the "cscope" target always creates a "cscope.out"
Date: Thu, 22 Jul 2021 11:50:49 -0700	[thread overview]
Message-ID: <xmqqfsw69o8m.fsf@gitster.g> (raw)
In-Reply-To: <YPmxsCatUIByQnig@nand.local> (Taylor Blau's message of "Thu, 22 Jul 2021 13:58:08 -0400")

Taylor Blau <me@ttaylorr.com> writes:

>> >  cscope.out: $(FOUND_SOURCE_FILES)
>> > -	$(QUIET_GEN)$(RM) cscope* && \
>> > -	echo $(FOUND_SOURCE_FILES) | xargs cscope -b
>> > +	$(QUIET_GEN)$(RM) cscope.out && \
>> > +	echo $(FOUND_SOURCE_FILES) | xargs cscope -f$@ -b
>
> But that alone is a good change in my mind at least. Then it's clear
> that that target is responsible for generating cscope.out and nothing
> else.

Probably.  The preparatory $(RM) is close enough to the invocation
of cscope that anybody adding other options like '-q' should be
careful enough to notice that the line needs to be touched, too, so
I can be talked into thinking that $(RM) change here is a good one.
I do not know about "-f$@", which is "Meh" to me.  Is there a good
reason to suspect that they might want to change the default output
filename?

> So I'd be in favor of rewording the patch message and only retaining
> this hunk (and dropping the other two).

Yup.  I do not mind dropping one half of this hunk, too, but again,
I do not mind keeping -f$@, either [*1*].

[Footnote]

*1* if the patch to add cscope support anew were done with -f$@, I
   wouldn't have insisted removing it, out of the principle "any Meh
   change is not worth applying once the code was written and
   working".