Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 24 Jun 2011 00:50:48 +0000
From:      Alexey Dokuchaev <danfe@FreeBSD.org>
To:        Florian Smeets <flo@FreeBSD.org>
Cc:        cvs-ports@FreeBSD.org, cvs-all@FreeBSD.org, ports-committers@FreeBSD.org
Subject:   Re: cvs commit: ports/www/squid Makefile ports/www/squid30 Makefile ports/www/squid31 Makefile
Message-ID:  <20110624005048.GA12150@FreeBSD.org>
In-Reply-To: <201106231511.p5NFBIaG008249@repoman.freebsd.org>
References:  <201106231511.p5NFBIaG008249@repoman.freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Jun 23, 2011 at 03:11:18PM +0000, Florian Smeets wrote:
> flo         2011-06-23 15:11:18 UTC
> 
>   Modified files:
>     www/squid            Makefile 
>     www/squid30          Makefile 
>     www/squid31          Makefile 
>   Log:
>   change all squid ports from CONFLICTS to CONFLICTS_INSTALL
>   
>   PR:             ports/158194

This makes me wonder about the usefulness of the idea of segregate
CONFLICTS: 1) Why do we need CONFLICTS_(BUILD|INSTALL), and 2) Why do we
check for CONFLICTS before the build, not before installation?  Maybe I am
missing some weird scenario where some port indeed cannot be built in the
presence of another (we have a few cases like that, but they are usually
solved by trivial patching (which is correct: marking ports as conflicting
in this case would simply mask out the problem instead of solving it
properly)?  I also do not understand why we preventively forbid something
which would not fail: every time I try to build something conflicting, I
get this message about "They install files into the same place".  Well, I
am not installing anything yet, am I?  :-)

./danfe



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20110624005048.GA12150>