Mailing-List: contact ivy-user-help@ant.apache.org; run by ezmlm
Precedence: bulk
Reply-To: ivy-user@ant.apache.org
Received-SPF: pass (nike.apache.org: local policy)
Date: Tue, 3 Jun 2008 20:32:00 +0200
From: Niklas Matthies <ml_ivy-user@nmhq.net>
To: ivy-user@ant.apache.org
Subject: Re: tag missing from ivy.xml
Message-ID: <20080603183200.GA49001@nmhq.net>
Mail-Followup-To: ivy-user@ant.apache.org
References: 
 <745B9EDF57802349B13F90E4E0B4B86C3BA402BFA4@HOUEXCH012.corp.halliburton.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <745B9EDF57802349B13F90E4E0B4B86C3BA402BFA4@HOUEXCH012.corp.halliburton.com>
User-Agent: Mutt/1.4.2.3i

On Tue 2008-06-03 at 12:10h, Shawn Castrianni wrote on ivy-user:
> The ivy.xml contains metadata about an ivy module like name,
> organization, branch, publications, and dependencies.  The branch is
> nice to record so we know what branch of source code was used to
> build a given module.  However, what if the module was built from a
> tag.  Shouldn't we have tag as an attribute of an ivy module?
> Currently I have been overloading the branch attribute to store tag
> names as well as branch names, but it seems messy.  Just curious
> what others thought.

Some thoughts:
Pro: In CVS, branches are just a special kind of tag (sharing the same
namespace too), so using the branch attribute for tags is probably
just fine.
Con: A tag is usually a symbolic name for a revision, so either the
tag should be used as the revision attribute, or else it is redundant
with the revision attribute.

-- Niklas Matthies