Talk:Black-box testing

Test design techniques

I removed from the Functional testing. Functional testing isn't a design technique. Black box tests can be functional or non-functional, but it isn't a design technique. I also removed Cross-browser testing. That isn't a technique either. You might use a decision table to design a cross-browser test, but the test itself isn't a design technique. Anyone disagree? DRogers 16:05, 5 March 2007 (UTC)

what is eqivalence testing.

The Equivalence partitioning section needs expanding. At the moment it doesn't really explain what Equivalence partitioning actually is. --StoatBringer 10:26, 29 June 2006 (UTC)

I added an own article for equivalence partitioning and another one for boundary value analysis and reduced the sections in this article to an outline with links to the other two articles. Ebde 09:39, 2 July 2006 (UTC)

I suggest that the link to "Testing Principles" is removed. That site has incorrect statements like "Tests are always specified and recorded." Isn't it better to refer to the Syllabus at http://www.istqb.org/ ? Epim 10:46, 3 October 2006 (UTC)

I agree. The current link makes absolute, black and white statements. Those statements may be true for the the organization that wrote that, but they aren't for me. Maybe it's a subtle language issue, since it's a German site. Anyway, I also agree that we should replace it with the ISTQB link. DRogers 13:30, 3 October 2006 (UTC)

Merge with Acceptance test

Regarding the merge tag that has been added: It would have been useful if some discussion had been added here to state why the merge is proposed. Acceptance testing is usually used to denote testing specifically to determine whether software is acceptable to customers, generally by using the software in as close to a real-world environment as possible. Black box testing is a technique for testing software, generally as part of testing by in-house staff - the Acceptance test article as it stands could be misleading in this respect. The two are completely different, so the merge makes no sense. --Michig 10:58, 1 December 2006 (UTC)

I agree fully. Acceptance test and black-box testing should remain two different articles. Epim 14:03, 1 December 2006 (UTC)

I'm sorry. I added the merge tag because I got confused. I was originally looking for functional testing. Functional testing redirects to black box testing. Acceptance test says that acceptance tests may be referred to as functional tests. It also says they are performed through black box testing. Hence I assumed acceptance test implies black box test. I read the black box testing article again and it says it is applicable to all levels of development; my assumption was wrong. So I guess functional testing should redirect to acceptance test, and acceptance test should explain that it is a kind of black box test. Does this make sense? --Chris Pickett 17:17, 1 December 2006 (UTC)

No, of course it doesn't make sense. How about a separate functional testing page, and making black-box testing point to functional testing and acceptance testing? I'll delete the merge tag. --Chris Pickett 17:22, 1 December 2006 (UTC)

Okay. I redirected functional testing to acceptance test, and tried to clarify some text. Let me know if there are still problems. --Chris Pickett 19:40, 1 December 2006 (UTC)

doc with examples to design techniques

Hi, i added a doc which has examples to some black box design techniques. Feel free to add your comments. --Erkan Yilmaz 13:44, 8 December 2006 (UTC)

Functional testing

I don't know anything about this topic, but I noticed that there was a link in the article to functional testing, which is, strangely enough, a redirect right back to black box testing! I removed the self-link, but after looking at some of the comments here and at talk:acceptance testing, maybe it should point somewhere else? Or have it's own article?--Aervanath 19:37, 29 January 2007 (UTC)

Lead paragraph

I would like to make this topic more accessible to, for instance, my boss (a PhD in computer science! ahem.) Before I change the article, I'd like some feedback please:

Black box testing is a methodology for understanding the behavior or implementation of a system whose internals are unknown or (purposefully) ignored. The methodology is to examine the system as nothing but a set of inputs and outputs. Examining the systems involves selecting sets of inputs and recording the output for each input set. Gray box testing is very similar, but some knowledge of the internals of the system are used to better select the input sets. White box testing is performed when the testers are aware of the system's implementation. By contrast, in white box testing, testers select test cases based on their knowledge of the system's implementation.

Though the term itself refers to hardware, this method of test design is applicable to all levels of software testing: unit, integration, functional testing, system and acceptance. The higher the level, and hence the bigger and more complex the box, the greater the requirement for black box testing. While this method can uncover unimplemented parts of the design, one cannot be sure that all functionality of the system has been tested.

--Otheus 18:49, 5 March 2007 (UTC)

I think this is fine, except for "White box testing is performed when the testers are aware of the system's implementation" as it could give the impression that if the internals of the system are known, then that makes it White box testing, when in fact its only white box testing when the internals are purposely examined. This may seem a subtle distinction but I have come across people who think that testing carried out by the developers of a piece of software *is* white-box testing, even if they are only performing black-box tests (i.e. they are aware of the implementation but ignore it). Perhaps this sentence would be better as "White box testing is performed when the testers are aware of the system's implementation and take it into account when designing test cases"--Michig 19:36, 5 March 2007 (UTC)

Now? --Otheus 19:49, 5 March 2007 (UTC)

Looks fine to me.--Michig 19:56, 5 March 2007 (UTC)

quote "one cannot be sure that all functionality of the system has been tested."

is this proven to be true? where can i find more info, please?

Where is this quote from? DRogers 12:52, 11 May 2007 (UTC)

it can take indefinite amount of time to find all functionality

Spelling

I see 3 different ways this type of testing is spelled just on this page. There's the official title "black-box", the image uses "blackbox", and the text refers to "black box". I don't know which is right, but there should only be one usage. White box testing has a similar problem. -Dmulter (talk) 22:45, 15 June 2009 (UTC)

Test Levels

Functional testing is not a Test Level... It's Test Type. Functional and non-functional tests can exists on any level (unit, integration, system, acceptance). -Nschoot. I removed functional as a test level. Reference: http://istqb.org/download/attachments/5439596/ISTQB+Glossary+of+Testing+Terms+2+1.pdf —Preceding undated comment added 09:25, 11 November 2011 (UTC).