Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 11 Jun 2010 15:57:40 +0100
From:      RW <rwmaillists@googlemail.com>
To:        ports@freebsd.org
Subject:   Re: Data files and ports
Message-ID:  <20100611155740.1779827f@gumby.homeunix.com>
In-Reply-To: <1276264730.2503.20.camel@hp-laptop>
References:  <1276264730.2503.20.camel@hp-laptop>

next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 11 Jun 2010 10:58:50 -0300
Jesse Smith <jessefrgsmith@yahoo.ca> wrote:

> I'm trying to port a program which is distributed in two separate
> packages from the upstream project. One package contains the
> executable program and the other contains data files. The Data
> package rarely changes. The idea being packaging them together would
> use up a lot of extra bandwidth.

> Which brings me to the question: Since the executable relies on the
> data files being in place before it's run, how should I handle that
> in the port? Should I just get the executable to install and let the
> user manually get the data files? Should I create a second port for
> the data package? Or should I find some way of making the
> executable's makefile download and unpack the data package?


I think it depends on how the data is versioned. If it has a version
number in the filename that's the same as the executable, then a single
port is best. If they have separate version numbers, then go with two
ports.


If the data is not versioned at all, it might be best to create a
data port that uses a snapshot that you host separately.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100611155740.1779827f>