898873 - Potentially clean up CUSTOM_TIMEOUTS where only one exists in the test

Status: RESOLVED WONTFIX

(Mozilla QA Graveyard :: Mozmill Tests, defect)

Description

[Jonathan French (:jfrench)]

A week or so ago I had done a little pattern search through the repo out of interest, searching for all custom timeouts (eg. TIMEOUT_DOWNLOAD) where:

1) no other custom timeout exists in that test
2) no sequentially following restart test uses that same timeout

The thinking being if there is only one timeout in the test, perhaps it's more economical to simply call those cases TIMEOUT?

Those files are:

functional/restartTests/testAddons_installFromFTP/test2.js
functional/restartTests/testAddons_installMultipleExtensions/test1.js
functional/restartTests/testAddons_installTheme/test1.js
functional/restartTests/testAddons_installUninstallHardBlocklistedExtension/test1.js
functional/restartTests/testAddons_RestartlessExtensionWorksAfterRestart/test1.js
functional/restartTests/testAddons_uninstallTheme/test1.js
functional/testSearch/testOpenSearchAutodiscovery.js
functional/testSecurity/testSubmitUnencryptedInfoWarning.js
functional/testTabbedBrowsing/testBackgroundTabScrolling.js
remote/restartTests/testDiscoveryPane_FeaturedModule/test1.js
remote/restartTests/testDiscoveryPane_FirstTimeModule/test1.js
remote/restartTests/testDiscoveryPane_installAddonWithEULA/test1.js
remote/restartTests/testDiscoveryPane_installCollectionAddon/test1.js
remote/restartTests/testDiscoveryPane_UpAndComingModule/test1.js

If there's a reason to keep them as is feel free to mark this bug resolved invalid.

Comment 1

[Dave Hunt [:davehunt] [he/him] ⌚BST]

I prefer the more descriptive name for these timeouts, but can see your point. Let's see what Henrik thinks.

Comment 2

[Henrik Skupin [:whimboo][⌚️UTC+2]]

I agree. The usage of TIMEOUT is too general and not obvious for what the constant is used. All those existing instances are on purpose. It will also not destroy the blame feature because those lines will not have to be changed, if another timeout constant gets added later.

I vote for wontfix here.