Mailing-List: contact dev-help@ant.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Ant Developers List" <dev@ant.apache.org>
To: dev@ant.apache.org
Subject: Re: Testing of Ant
References: <200307211315.43256.magiel@software-improvers.com>
From: Stefan Bodewig <bodewig@apache.org>
Date: 21 Jul 2003 13:46:29 +0200
In-Reply-To: <200307211315.43256.magiel@software-improvers.com>
Message-ID: <m3k7acw19m.fsf@bodewig.bost.de>
Lines: 36
User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Portable Code)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii

On Mon, 21 Jul 2003, Magiel Bruntink <magiel@software-improvers.com>
wrote:

> To validate theoretical results, I have used the Ant sources as the
> subject of a case study.

Then you already have our main point of testing, our unit tests (in
the src/testcases subdirectory).

> 1. Do you use some kind of testing criterion?

As much as possible 8-)

There are no hard rules.  When a bug is discovered we prefer to have a
unit test that verifies the bug.  This one will become part of our
testsuite and ensure that we won't create the same bug again.  But
writing tests is no requirement.

> 2. Is the level of compliance to the testing criterion subject to
> measurement?

As there are no rules, there is nothing to validate.

> Do you use a tool to automatically calculate your code coverage?

Not methodically as part if Ant itself.

Conor MacNeill has done so[1][2] over time using Clover, Stephane
Bailliez used to do so with JProbe coverage IIRC (sorry, no URL).

Stefan

Footnotes: 
[1]  http://codefeed.com/blog/archives/000004.html
[2]  http://codefeed.com/blog/archives/000020.html


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@ant.apache.org
For additional commands, e-mail: dev-help@ant.apache.org