Talk:Smoke testing (software)

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Proposed merge with Build verification test[edit]

Seems to a kind of smoke testing (or maybe a Microsoftism for the same concept). QVVERTYVS (hm?) 01:32, 20 December 2013 (UTC)[reply]

They're similar concepts but have different goals. If that can be captured and explained, I'm OK with merging. The fact that it is used interchangeably in some corners is a good reason to merge. Walter Görlitz (talk) 01:43, 20 December 2013 (UTC)[reply]

See also previous discussion, which discussed merging into another article. QVVERTYVS (hm?) 02:44, 20 December 2013 (UTC)[reply]

This article lacks one essential feature, the origin of the term. "Smoke testing" sounds idiomatic. --Philip lamb (talk) 02:59, 15 June 2018 (UTC)[reply]

@Philip lamb: I'm not sure what this has to do with a discussion that several years old about merging with another article, but at the top of the article is a "hatnote". It states "For other uses, see Smoke testing." On that page are other terms and Smoke testing (mechanical) is where the origin of term is discussed. Walter Görlitz (talk) 03:11, 15 June 2018 (UTC)[reply]

Proposed merge with Sanity testing/software development[edit]

Sanity testing#Software development suggests that "sanity testing", in software development, is the same as smoke testing, and it has a reference to corroborate this. QVVERTYVS (hm?) 01:35, 20 December 2013 (UTC)[reply]

First there should not be a split. All that needs to happen at the Smoke Testing article is summarize the various sections.
Second, a sanity test is not a smoke test in software. It's very different. Walter Görlitz (talk) 01:45, 20 December 2013 (UTC)[reply]
Then please tell me the difference and provide a reference. QVVERTYVS (hm?) 02:09, 20 December 2013 (UTC)[reply]

--Qside (talk) 14:56, 5 June 2014 (UTC)This article is talking about Intake Testing, which is distinct from smoketesting. Article should be renamed. Reference is...[reply]

Standard glossary of terms used in Software Testing V2.2 (dd. October 19th, 2012) Produced by the ‘Glossary Working Party’ International Software Testing Qualifications Board

http://science.df.lu.lv/kaab13/istqb_glossary_of_testing_terms_2.2.pdf

There can be a difference Smoke testing is Build Verification Test checks whether the build launches successfully after installation and Sanity testing is shallow testing done after regression is over with an intent to verify that end user requirements are met on not.(3rd paragraph of <https://en.wikipedia.org/wiki/Sanity_testing)[1]


I don't like idea of merging those 2 things. Programmer can insert a sanity check in code in order to rule out if for example configuration of the program is sane. Lets say a piece of code that checks if configuration entry describing the TCP port number is in fact a number and is below 65535. In this meaning a sanity check is an inherent part of the program. Smoke tests on the other hand are always something that comes from the outside. Let it be a user testing if some basic functionalities are working or an automated test, that is not an inherent part of the running program.

Emergie (talk) 17:44, 23 August 2016 (UTC)[reply]

References

redirect until this is resolved[edit]

The article should never have been created and I am being bold in redirecting until a decision has been made. The confusion has endured long enough. Walter Görlitz (talk) 05:21, 20 December 2013 (UTC)[reply]

Etymology[edit]

So why is it called a "smoke" test?

77.201.128.233 (talk) 14:49, 28 February 2019 (UTC)[reply]

This is discussed at the other articles. Its earliest use was in pipe-fitting. They would inject smoke into the system to see if any leaked-out. It continued in electronics where you applied power to the breadboard and if anything smoked, you had a problem. Walter Görlitz (talk) 15:39, 28 February 2019 (UTC)[reply]
I have copied the Etymology from Smoke testing (electrical), where we have the etymology cited (even if it isn't correct!). The text in that article may now be redundant (see its talk page).
A very plausible alternative view is that the term derives from typography (see Punchcutting) where a newly carved punch was tested with soot (instead of ink) on paper to check it was OK.
--EdJogg (talk) 11:02, 16 July 2021 (UTC)[reply]
I too learned the term from Kaner, but have subsequently learned that smoke test had an earlier definition with the same result. Walter Görlitz (talk) 17:57, 19 July 2021 (UTC)[reply]

Smoke Test: Heavy load[edit]

For me a smoke test, tests if the system works under heavy load.

I am not the only one who uses the term like this. Example:

https://smartbear.com/blog/test-and-monitor/7-types-of-web-performance-tests-and-how-they-fit/

Quote: 7. Smoke Test: A smoke test is a test run under very low load that merely shows that the application works as expected. The term originated in the electronics industry and refers to the application of power to an electronic component. If smoke is generated, the test fails and no further testing is necessary until the simplest test passes successfully. For example, there may be correlation issues with your scenario or script – if you can run a single user test successfully, the scenario is sound. It is a best practice to initate one of these “verification” runs before running larger tests to ensure that the test is valid. — Preceding unsigned comment added by Guettli (talkcontribs) 10:34, 17 February 2020 (UTC)[reply]

The fact that the idea that a smoke test started in the electronics industry and not plumbing shows this is not a reliable source. Secondly, they don't actually suggest measuring the load on the system. That they state the system itself is not under an unusually high load does not equate to verifying that it's not under an unusually high load. Since we do not state that the system should be under an under unusually high load, we leave it to the reader to understand that the system should be under a normal load.
I'll go further to say suggest the following. Assume you're testing a COTS app, such as a word processor. I would expect one of the smoke tests to be that it read a file from disk into memory. If that file is somewhat large, or the volume is fragmented, the disk could reach maximum throughput, thereby putting one part of the system under high load for a moment. Yet, that would be a valid smoke test. Walter Görlitz (talk) 19:00, 17 February 2020 (UTC)[reply]