From owner-freebsd-hackers  Tue Jun  1 16:42:58 1999
Delivered-To: freebsd-hackers@freebsd.org
Received: from alcanet.com.au (border.alcanet.com.au [203.62.196.10])
	by hub.freebsd.org (Postfix) with ESMTP id 9B41615885
	for <hackers@FreeBSD.ORG>; Tue,  1 Jun 1999 16:42:52 -0700 (PDT)
	(envelope-from jeremyp@gsmx07.alcatel.com.au)
Received: by border.alcanet.com.au id <40402>; Wed, 2 Jun 1999 09:26:40 +1000
Date: Wed, 2 Jun 1999 09:42:27 +1000
From: Peter Jeremy <jeremyp@gsmx07.alcatel.com.au>
Subject: Re: a two-level port system?
To: hackers@FreeBSD.ORG
Cc: darrylo@sr.hp.com
Message-Id: <99Jun2.092640est.40402@border.alcanet.com.au>
Sender: owner-freebsd-hackers@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

Darryl Okahata <darrylo@sr.hp.com> wrote:
>Peter Jeremy <jeremyp@gsmx07.alcatel.com.au> wrote:
>> How about storing each port as a single file in ar(5) format, which is
>> unpacked into the directory structure when make'd?  ar(5) is a text
>> format, which means it can easily be managed by CVS, which includes
>> a tool for manipulating its contents - ar(1).
>
>     This isn't all that different from the existing *.tar.gz port
>files.  If you use those, you get all the advantages of your approach,
>plus fewer disadvantages:

I think you've misunderstood me.  There are three distinct sets of
files associated with a port:
1) The port-and-FreeBSD-specific config files which are stored under
   /usr/ports/<category>/<portname>.  These files comprise a Makefile,
   various package files in pkg and various optional patches and
   scripts.
2) The original distribution file.  This may be located anywhere on
   the internet (the location and name is in the port Makefile) and
   will be stored in /usr/ports/distfiles after download.  These
   can be in any format, but .tar.gz is probably the most common.
3) Pre-built ports available as packaged from ftp.freebsd.org.  I
   think these are all .tar.gz, but some might be bzip2 format.

The discussions have all centered on 1) above.  This is the area where
there is the greatest wastage of resources (inodes and unused partial
blocks).

>  - No need to CVS commit ar files.  (BTW, CVS can also handle binary
>    files, so ASCII vs. binary is a non-issue.)
The problem is that the diffs between 2 uuencoded .tar.gz files (which
is how CVS would treat them) tend to contain the entire contents of
both files.

>> Disadvantages:
- CVS files will bloat.
- relatively difficult to examine the innards of the files for each port.

Peter


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message