Talk:List of unit testing frameworks

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Computing (Rated List-class, Low-importance)
WikiProject icon This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
 List  This article has been rated as List-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.
 

xUnit[edit]

What is the difference between this article and the xUnit article? -- nolandda

Not all of these frameworks are xUnit. --Chris Pickett 05:59, 3 December 2006 (UTC)

What qualifies something as following the xUnit framework? The RUnit package for R appears to follow the architecture laid out on the xUnit page, but here it says it doesn't. I'm no expert, so I'm wondering whether this is a mistake. 128.174.127.111 (talk) 17:11, 1 March 2013 (UTC)

Layout: Tables[edit]

I converted (then IP) the bullet-list page to a big number of small tables. The point was that then you have a TOC and can click on your language. Could we discuss the pros and cons of One-big-table versus Many-small-tables? --User77764 13:04, 3 December 2006 (UTC)

Another reason for many small tables is that I want to introduce columns to the C++ table which don't make sense anywhere else. --User77764 13:11, 3 December 2006 (UTC)

And to be able to do this, I will revert to the many tables version. This is no final decision of course. --User77764 13:31, 3 December 2006 (UTC)

Yet another idea: There are many languages with only one framework listed. Maybe combine them all into a single table while those languages with multiple entries get tables of their own? --User77764 16:33, 3 December 2006 (UTC)

That will make it annoying if and when new frameworks crop up, it should be all the same style I think. So many smaller tables is fine, if somewhat more cumbersome to maintain when adding a new category to each table. I was thinking one big table for the columns that can be common between languages, and smaller tables for the columns particular to a language, but then you have to maintain the list of frameworks in two places. So many smaller tables then, but add the xUnit column to all of them. I wish table-editing was wysiwyg, would make life easier. --Chris Pickett 17:54, 3 December 2006 (UTC)
OK, so we go for one table per language... maybe add all columns that are explained at the top to all tables? --User77764 18:08, 3 December 2006 (UTC)
Sure... and it seems better if the common columns in each table go before the language-specific ones, with the exception of Remarks. --Chris Pickett 18:42, 3 December 2006 (UTC)
Ack. --User77764 20:31, 3 December 2006 (UTC)

Expansion[edit]

The list needs information. White boxes mean: We do not know. Anybody with sufficient knowledge please help out. Also, more columns should be added where appropriate. --User77764 14:58, 5 December 2006 (UTC)

In the PHP category, Code Igniter is a "Web Application Framework" so even if it contains a Unit test suite, it seems to be in a wrong place. 193.49.124.107 15:25, 25 April 2007 (UTC)

Add QtUnit? Based on CppUnit. -- pne (talk) 09:49, 5 November 2007 (UTC)

Regarding language dialects[edit]

The language list surely is too long. Recently an entry for Symbian was added. Now this is clearly NO language, is it? Afaik, they use C++ in some flavor. Anyways, they don't deserve an own entry. Where to place such candidates? --User77764 17:49, 15 December 2006 (UTC)

I wasn't sure about that one either. You could merge it with C++, but is it really a C++ framework? What other ones did you want to merge? Chris Pickett 20:19, 15 December 2006 (UTC)
Hmm, Transact-SQL and SQL? I don't know, most of the languages are unknown to me... (but I know more than one of them ;-) ) --User77764 18:28, 19 December 2006 (UTC)
I think we should just leave it. If you can find an example where compilers for either one will happily accept the other, then sure, merge to the same language. Chris Pickett 11:18, 21 December 2006 (UTC)
Well, I still think the list (of languages!) is too long. But then, the "article" lacks in more important areas, for sure. --User77764 00:28, 23 December 2006 (UTC)

Specify Visual Basic 6 (VB6)[edit]

I assume that the "Visual Basic" heading refers to Visual Basic Ver. 6, or at least I think it should. Visual Basic Ver. 6, a.k.a. "VB6", is of course quite distinct from VB .NET; the latter which I expect should be included under the ".NET programming languages" heading. So my suggestions: (1) change "Visual Basic" heading to "Visual Basic Ver. 6 (VB6)", and (2) remove csUnit entry from the VB6 group. csUnit already appears under the .NET programming languages, which it seems is correct -- from what I can tell, csUnit only supports .NET and does not support VB6. --Nmarler 16:42, 29 April 2007 (UTC)

Emacs Lisp[edit]

The link for the ELisp unit testing package times out - is this link broken? Autarch 20:27, 9 November 2007 (UTC)

The link was indeed broken, as the server said the file didn't exist today. I added a link to the appropriate page as well as adding a link to another Elisp unit testing package. Autarch 16:24, 12 November 2007 (UTC)

QTestLib[edit]

Qt (toolkit) includes QTestLib. It should be on the list, but I'm not sure about the properties. --Ropez (talk) 12:50, 21 February 2008 (UTC)

GLSL[edit]

Why is UnitTestCg listed as supporting GLSL? It does not say so on the Google Code page, or anywhere at all. I believe that the author might have mixed it up with HLSL. --193.77.149.126 (talk) 07:08, 17 December 2008 (UTC)

Fortran unit testing framework missing[edit]

What about pFunit? [1]

ANTLR[edit]

gUnit - ANTLR grammar testing [2] —Preceding unsigned comment added by Seaswiki (talkcontribs) 23:45, 24 December 2009 (UTC)

Google mocking framework is not a testing framework[edit]

What is it doing here? Also, Google Testing framework does not provide mocks - it uses google's Mocking framework to do that. Plus, it doesn't support exceptions in so far as that when one occurs when you don't specify it the entire testing framework falls over. —Preceding unsigned comment added by 82.210.249.81 (talk) 09:30, 7 May 2010 (UTC)

UnitTest++ not xUnit like?[edit]

According to the Wikipedia page, UnitTest++ has all the required features for being considered xUnit like, yet it's listed as a "no". —Preceding unsigned comment added by 82.210.249.81 (talk) 09:32, 7 May 2010 (UTC)

Wiki link validity[edit]

Many test frameworks have common names, so putting a wiki link to general page isn't helpful (eg. Fact, Specter and Pex). 99.237.58.158 (talk) 20:29, 22 May 2010 (UTC)

Distinction between unit testing and mocking[edit]

The distinction between a "unit testing framework" and a "mock framework" ought to be explained somewhere. Unit testing is often explained with nice examples in which unit testing suffice. This is fine as far as introducing unit testing goes. In my personal experience, unit testing does not make much sense without mocking support. —Preceding unsigned comment added by 142.169.202.4 (talk) 14:14, 22 July 2010 (UTC)

That seems silly. Mocks are only useful when you are trying to "unit test" a component that by all rights should be covered by an integration test. Grauenwolf (talk) 06:41, 7 May 2012 (UTC)

Time for a HEAVY trimming[edit]

This list is already in WP:LINKFARM territory, and it isn't its only problem, as it also violates WP:LSC: if a complete list would include hundreds of entries, then you should use the notability standard to provide focus to the list.

Therefore, my suggestion here is to:

  1. First, remove every entry without an article (solves LSC)
  2. Then, remove the "Homepage" column as the external links would be provided by the wikilinked articles. (solves LINKFARM)

--M4gnum0n (talk) 16:31, 24 July 2012 (UTC)

Linkfarm, yes. SpamLinks, no. In other, similar lists, links to product details pages or FAQ are supplied in their own column. That is what these should be converted to since they do not have refs. Also, other lists only have notable products, and by notable I mean, products with articles on Wikipedia. That is not a requirement on Wikipedia as M4gnum0n suggests, but it is a discussion point. I don't think this list would benefit from such pruning. --Walter Görlitz (talk) 16:35, 24 July 2012 (UTC)
I don't really think the pruning is a discussion point, as the current state of the list contradicts my direct quotation of a guideline above. --M4gnum0n (talk) 17:18, 24 July 2012 (UTC)
Thanks for sharing your opinion. I don't think that those guidelines apply here. --Walter Görlitz (talk) 18:08, 24 July 2012 (UTC)
Then let's wait for a few days and see if something moves. --M4gnum0n (talk) 09:03, 25 July 2012 (UTC)
OK, more than one week passed and no one chimed in. This led me to refer to the Dispute resolution policy, which I found somewhat confusing. If I read that correctly, it seems that the choice is between WP:3O, WP:DRN and WP:RFC. What do you think the right process should be? --M4gnum0n (talk) 11:05, 2 August 2012 (UTC)
Third opinion is not binding and usually a waste of time particularly when someone who knows little chimes in. It's better just to go for an RFC. But I would like to reiterate, this isn't a link farm since there is additional information provided for most entries and so it is not a "'mere collection of external links or Internet directories". And as for selection criteria, not all are "supported by reliable sources". --Walter Görlitz (talk) 14:01, 9 August 2012 (UTC)

RfC on trimming[edit]

Should this list only contain items that already have an article? M4gnum0n (talk) 09:49, 17 August 2012 (UTC)

  • No because there is information provided about the entries aside from the item's name which makes more than a mere collection, it adds value to the material to allow for comparison of the items. It also classifies them. While not all entries entry currently meet notability criteria for articles, entries have been verified, and it is reasonable to expect an article could be forthcoming in the future. This list is not too large to be useful to readers since it is grouped into smaller sections. --Walter Görlitz (talk) 14:06, 17 August 2012 (UTC)
  • No, if the items are important to the scope of the article, it should not matter whether they are notable enough to have their own article. To offer an analogy, a scientific paper or study may not be important enough to warrant its own article, but could be mentioned in the article concerning its field of study. Silvrous (talk) 15:28, 2 September 2012 (UTC)
    • Comment Silvrous' comments are against several guidelines such as WP:LINKFARM. Sorry, but that doesn't change the fact that the list is useful. --Walter Görlitz (talk) 23:30, 2 September 2012 (UTC)
  • No. Even if there were any merit to the criterion, it would not be relevant. WP rules are supposed to be aimed at constructiveness, not smugly rule-bound constipation. If it makes sense to refer to something in a reasonable context and good sense, as long as it is reasonably verifiable, in it goes. As for WG's comment, point(s) taken, but a more helpful lede should deal with such objections adequately, even if it does little beyond referring helpfully to the parent article. JonRichfield (talk) 03:08, 8 September 2012 (UTC)
"WP rules are supposed to be aimed at constructiveness, not smugly rule-bound constipation. " {{citation-needed}} Andy Dingley (talk) 08:19, 8 September 2012 (UTC)
My apologies; your challenge id obviously unanswerable. I shall invest in a cork immediately. JonRichfield (talk) 09:49, 8 September 2012 (UTC)
  • No That is never a good criterion. WP is not a recursive WP:RS. Andy Dingley (talk) 08:19, 8 September 2012 (UTC)
  • No - Per the comments above. United States Man (talk) 23:58, 10 September 2012 (UTC)

xUnit.net[edit]

I think that the remarks is not exact: from the license of NUNIT [license of NUNIT|http://nunit.org/index.php?p=license&r=2.6.2] i deduce that the creator is Philip A. Craig, in 2000. James W. Newkirk, the creator of xUnit.net, appears in the licence from 2002 to 2004. I think also that the [reference|http://stackoverflow.com/questions/261139/nunit-vs-mbunit-vs-mstest-vs-xunit-net] is not the most adecuate: i do not understand how it is posible deduce "xUnit.net is currently the highest rated .NET unit testing framework" from a Q/A, where, in this case, all the answers are personal opinions of the users. Maybe you can take as an indicator, but is a weak probe for an afirmation. — Preceding unsigned comment added by Narkha (talkcontribs) 11:55, 9 September 2013 (UTC)