From: David Bremner <david@tethera.net>
To: Jani Nikula <jani@nikula.org>, Mark Walters <markwalters1009@gmail.com>,
	notmuch@notmuchmail.org
Subject: Re: [PATCH] configure: only install bash completion if supported
In-Reply-To: <87vbwxw6u2.fsf@nikula.org>
References: <1391359677-12497-1-git-send-email-jani@nikula.org>
	<877g9dqup5.fsf@qmul.ac.uk> <878utt1b9s.fsf@maritornes.cs.unb.ca>
	<87vbwxw6u2.fsf@nikula.org>
User-Agent: Notmuch/0.17+56~gb9c3d8e (http://notmuchmail.org) Emacs/24.3.1
	(x86_64-pc-linux-gnu)
Date: Sun, 02 Feb 2014 19:18:31 -0400
Message-ID: <87lhxt14u0.fsf@maritornes.cs.unb.ca>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: list

Jani Nikula <jani@nikula.org> writes:

> On Sun, 02 Feb 2014, David Bremner <david@tethera.net> wrote:
>> Mark Walters <markwalters1009@gmail.com> writes:
>>
>>> This LGTM (untested)
>>
>> I did test it, at least completely removing bash completion works as
>> expected.  
>>
>> Unfortunately --with-bash-completion does not override this test because
>> the order things are processed. Do you think this is a bug?  I wondered
>> if users that "know what they are doing" (TM) might want to force
>> installation even though the pkg-config test fails.
>
> I thought it was a feature, not a bug, but I'm fine either way.

I can live with the current patch. As you pointed out on IRC, this is
the usual way missing dependencies work.  And I think we should avoid
extra complications in the configure script when we can.