Talk:Stress testing (software)

Field Experience

"A major software company has identified that failure to execute load and stress tests occurred in about 1 of 10 engagements."

This needs a citation (as noted by the banner at the top). However, I would also question its accuracy and relevance. Without specifying the company, or even the type of product that they offer, the statement has no context. Embedded systems and web servers may undergo more testing as they often have to work under stressful conditions (relative to the resource available), but I'd be surprised if a considerable portion of commercial applications software gets stress-tested at all. It is often considered acceptable for a program to crash occasionally as the cost of testing can appear fairly high, but a bug in the baseband software on your mobile phone can significantly harm the value of the phone, so the company is likely to invest more into testing.

In addition, there are varying levels of stress testing. Some companies may consider basic stress testing to be sufficient, whilst others may be more rigorous. I suspect that, for much of the publicly-visible software, the above statement is true, but I would also estimate that much of the remaining 90% of testing is inadequate, even though it takes place.

Of course, I have no references for the above and it is based on experience more than anything else, and for that reason I haven't added this to the main page.

Jacob (talk) 07:52, 29 August 2008 (UTC)

Agree  Done edit dif --trevj (talk) 07:36, 15 April 2011 (UTC)

Test vs. Testing

The article was called Stress testing and it was moved to stress test without any discussion. Other articles in the software testing series are title -box testing, black-box testing, gray box testing, unit testing, integration testing, system testing, acceptance testing, installation testing, compatibility testing, regression testing, acceptance testing and the partner to this is load testing. So unless someone gives me a good reason for the move, I'll be moving it back. --Walter Görlitz (talk) 19:44, 24 February 2013 (UTC)

Please see discussion Talk:Stress_test. There is more expansion to come in the stress test area (financial). If after some study you concur with the overall naming convention (Encyclopedia format) you can retract your reversion to keep everything consistent. Please also see the content additions and reforming you have rolled back in your revision. It affects several pages:

• Stress test (disambiguation)
• Stress test (broad scope article)
• Stress test (hardware)
• Stress test (software)
• Stress test (financial)

Note everything that was there prior remains, no content has been eliminated or edited, all of above is just a restructuring and expansion in some areas. Rick (talk) 19:58, 24 February 2013 (UTC)

No. We'll discuss here, thanks. This is the article you moved. This is where the discussion will happen.

Also, I redirected your new stress test article to the existing stress testing one, because it's hubris to assume that you know better than years of other editors and create a new article and then delete all of that content and point it to your newly created one. --Walter Görlitz (talk) 22:41, 24 February 2013 (UTC)

This article created 2008, not moved, not newly created. Rick (talk) 01:39, 25 February 2013 (UTC)

It should be restored to its original location in-line with the naming conventions of the rest of the project rather than those of other projects. --Walter Görlitz (talk) 03:02, 25 February 2013 (UTC)

Can you clarify what the naming convention is, and what individual articles adhere to it? Rick (talk) 07:15, 25 February 2013 (UTC)

Already done in the first paragraph. Feel free to find others linked from the software testing article. Cheers. --Walter Görlitz (talk) 07:18, 25 February 2013 (UTC)

Page titles can be changed with very minimal effort. The full history and talk page moves automatically. Yes the inbound links should be updated, however that process doesn't need to be immediate as the redirect pages handle them. It is unquestioned that Wikipedia strongly prefers nouns as article titles. Ask yourself, in these articles themselves, are you focused mostly on the actual tests? For example, describing what the tests actually are, their nature. Is each article focused on describing the various test objects, elements, attributes and describing the methods (what they do)? Are (or should) most of the cross links between these pages be to other "objects"?

Or, are you mostly describing the process of tests actually "running", for example in execution vs. design mode? Are (or should) most of the cross links between these pages be to other "actions"?

Wikipedia is not a static repository, it's constantly refactored and refined in an attempt (difficult and imperfect as it may be) to describe all human knowledge. Should most of the core, foundation, building blocks in testing be nouns or verbs? What works best "for the long haul" here? Change is often unwelcome, dogmatically resisted, thwarted, takes energy, is frequently initially misunderstood. In the context of an encyclopedia what is the very best way to go here:
Stress test vs. Stress testing
Box test vs. Box testing
Black-box test vs. Black-box testing
Gray box test vs. Gray box testing
Unit test vs. Unit testing
Integration test vs. Integration testing
System test vs. System testing
Acceptance test vs. Acceptance testing
Installation test vs. Installation testing
Compatibility test vs. Compatibility testing
Regression test vs. Regression testing
Acceptance test vs. Acceptance testing
Load test vs. Load testing
If moving in a forward direction actually doesn't take much effort, why not? Rick (talk) 17:55, 25 February 2013 (UTC)

Moved back to match the remainder of the project. Feel free to discuss moving all articles at the project. Until consensus for your excellent suggestion has been made, this article should not be the only one that does not follow the pattern. Walter Görlitz (talk) 14:52, 4 March 2013 (UTC)