Mailing-List: contact ant-dev-help@jakarta.apache.org; run by ezmlm
Message-ID: <38FE2495.C25224C@eng.sun.com>
Date: Wed, 19 Apr 2000 14:26:45 -0700
From: James Duncan Davidson <james.davidson@eng.sun.com>
MIME-Version: 1.0
To: ant-dev@jakarta.apache.org
Subject: Re: Ant Principles
References: <852568C6.00716A2F.00@d54mta04.raleigh.ibm.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit


> 1) Can we settle on a design pattern where we can easily distinguish
> between these three cases?

Taken as an action item to investigate this in depth today. Will report
back.

> 2) Is the current implementation sufficiently broken to justify this
> change? 

The current implementation isn't clear enough. One of the things that I
want to see is that it's so clear and documented well enough so that
when people look at the code and think of additions, that they are in
line with the design intent. I don't want to be in a position to have to
explain nuances in long detail to convey intent. If it can't be done in
a para....

> Last I heard, one of the primary reasons given for the change was:
> 
>     But if we go with the slightly less flexible model, then all
>    methods are set*** methods.. Cleaner and simpler.

Yes.. Like I said, I'll take a deep look at the approaches and report
back on pros and cons of the alternatives.

.duncan