Date: Fri, 03 Nov 2006 21:47:10 -0800 From: Doug Barton <dougb@FreeBSD.org> To: Matthew Seaman <m.seaman@infracaninophile.co.uk> Cc: Charles Sprickman <spork@bway.net>, freebsd-ports@freebsd.org Subject: Re: [nycbug-talk] creating "local" ports (fwd) Message-ID: <454C295E.8060002@FreeBSD.org> In-Reply-To: <454BB10E.4020900@infracaninophile.co.uk> References: <Pine.OSX.4.61.0611031519500.4567@dyn-160-39-250-49.dyn.columbia.edu> <454BB10E.4020900@infracaninophile.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
Matthew Seaman wrote:
> That's a generic problem with ports -- instead of registering a
> dependency on the package that provided the file that satisfied the
> test specified by the FOO_DEPENDS variable in the port Makefile, the
> dependency is registered on the default package to install to resolve
> the dependency as shown in the second half of the FOO_DEPENDS line.
That's both a positive and a negative. The positive is that the ports
infrastructure works transparently when a dependency is satisfied by
another port at build/run time. But you're right, it doesn't have a
mechanism for translating that knowledge into something to record in
the +CONTENTS file.
> No idea if either portmaster or portmanager have any similar
> alternate dependency functionality though.
What portmaster does is to read CONFLICTS for each port that is
depended on, and then look to see if you have one of those ports
installed. If it does, it treats that port as the one to satisfy the
dependency for purposes of looking for updates, building your port,
etc. It does not twiddle with the DEPORIGIN in the +CONTENTS file
however.
hth,
Doug
--
This .signature sanitized for your protection
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?454C295E.8060002>