Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 15 Apr 2004 18:50:50 +0200
From:      Jens Schweikhardt <schweikh@schweikhardt.net>
To:        Alexander Leidinger <Alexander@Leidinger.net>
Cc:        freebsd-ports@freebsd.org
Subject:    Re: ports and DESTDIR (was: Second "RFC" on pkg-data idea for ports)
Message-ID:  <20040415165050.GA992@schweikhardt.net>
In-Reply-To: <20040414210630.4353e27c@Magellan.Leidinger.net>
References:  <p0602040cbca10a7dbe52@[128.113.24.47]> <200404131516.i3DFGMJA078941@green.homeunix.org> <20040413181311.GA5858@schweikhardt.net> <20040414210630.4353e27c@Magellan.Leidinger.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Apr 14, 2004 at 09:06:30PM +0200, Alexander Leidinger wrote:
# On Tue, 13 Apr 2004 20:13:11 +0200
# Jens Schweikhardt <schweikh@schweikhardt.net> wrote:
# 
# > For starters: one real BIG problem (at least for me) the ports system
# > still has is a lack of DESTDIR support. We can
# > 
# >   make installkernel DESTDIR=/foo
# >   make installworld  DESTDIR=/foo
# > 
# > but try this with ports and let all hope go. There is even a PR in the
# > queue that solves some issues, but it has been left to rot for years,
# > http://www.freebsd.org/cgi/query-pr.cgi?pr=ports/28155
# 
# I fail to see how you want to solve the DESTDIR= issue for every
# installation procedure we have in the ports tree. You have to convince
# every autoconf, automake, custom made, what ever "makefile" to respect
# DESTDIR in his "make install" target. For 10.000 ports this is a lot of
# work. Any idea how to solve this problem?

I realize that this is next to impossible (but I like to say one's
personality grows with the herculean task you succeed at :-)

I can think of at least two possible ways to avoid messing with all
third-party Makefiles.

* do a regular install with DESTIR=/, see what has been installed
  (maybe use PLIST), use that information to bring the files to DESTDIR.
  Yup, that's gonna be tricky for stuff in /var/db/pkg, but it does not
  look flat out impossible.

* chroot magic

Regards,

	Jens
-- 
Jens Schweikhardt http://www.schweikhardt.net/
SIGSIG -- signature too long (core dumped)



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040415165050.GA992>