Talk:Test harness

This page pretty flagrantly plagiarizes http://www.devdaily.com/java/junit/node9.shtml —The preceding unsigned comment was added by Docjowles (talk • contribs).
 * You're right. I've removed the copyvio, effectively reverting to the June 2006 version with a few typographical improvements. —Caesura(t) 20:20, 19 February 2007 (UTC)

metaphor
Can someone explain the harness metaphor/etymology? I don't get it. --Abdull (talk) 12:08, 2 April 2008 (UTC)

One of the definitions of harness from WordNet (r) 2.0: "stable gear consisting of an arrangement of leather straps fitted to a draft animal so that it can be attached to and pull a cart." A test harness is analogous to this definition. A test harness is software which is used to control and direct another piece of software. In this case the software being tested is being controlled. -- Simon --217.149.208.50 (talk) 08:07, 19 May 2008 (UTC)

Please do not merge "test harness" with "test automation framework". I'm working for 5 years now a test automation framework for functional black box testing. That's something different from "test harness" which is more something for a developer white box test. Sorry if this contribution is not on an appropriate way. This is the first time I use Wikipedia. Erik tV —Preceding unsigned comment added by 167.202.201.4 (talk) 13:53, 1 March 2010 (UTC)

Test automation framework as I understand it is a set or collection of tools that help with test structure organization and implementation, but a test harness is an implementation, or collection of implementations of tests than make can make use of a test automation framework. In my mind, this is similar to the concept of a test plan compared to a test case where a test plan is a higher level abstract used to describe generally how tests will be conducted, and the test case are steps to perform an individual test. In my opinion these are two different things. 12.108.25.2 (talk) 21:59, 31 March 2010 (UTC) Barry

You guys both bring up very good points. A TAF is a set of components, they can be tools, code, and applications or all combined. So for example in HP=Mercury's 4th generation tool BPT the framework consists of Quality Center (TM), QuickTest Professional (R&P) and BPT (4th generation). These tools by themselves provide a 4th generation test automation framework. However this framework can be extended through use of resource libraries within QC/BPT. On the other side TestFrame, the original 4th generation TAF can use any test automationtool, uses an execution engine known as a controller (TestFrame Engine) is written in C++ and implemented as a Windows .dll. The test cases are written in Microsoft Excel (application) and the wrappers (scripts) are written in the language of the test tool.

So I think in terms of a test harness, it is an implementation that uses a TAF, and as Barry is saying a collection of tests. So essentially the TAF can provide the physical structure for the implementation, the test harness provides the method by which the tests are implemented. So I think they should be kept separate, since there is enough of a difference. ProfessionalTST (talk) 16:26, 16 July 2010 (UTC)

In "Release It!" by Michael Nygaard the term "test harness" is used to describe a remote system emulator that produces specific error conditions. In this article the term "test harness design pattern" is coined to describe ways of setting up a test fixture to test a component. A native speaker should do some research about this article and its usage of the term. —Preceding unsigned comment added by 212.227.35.68 (talk) 09:27, 10 September 2010 (UTC)

Test Harness v. Test Framework
I think most comments here are in agreement with the idea that a Harness and a framework are related but distinct in nature. I would like to drive the update for this point. The article needs to be modified to describe a test harness (opposed to a testing framework). I am not suggesting that this article describe the difference but that references to developing the functional automation be removed and simply focus on the scheduling, execution, results and reporting which describe a test harness.

I am suggesting this change be submitted and review no later than February 1, 2014. All changes be agreed upon before making the actual changes.

-=Dirk=-