[Webkit-unassigned] [Bug 215986] [WPE] 2.29.90: build faiils when ENABLE_API_TESTS=ON

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Jun 16 04:43:24 PDT 2021


https://bugs.webkit.org/show_bug.cgi?id=215986

Michael Catanzaro <mcatanzaro at gnome.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mcatanzaro at gnome.org

--- Comment #6 from Michael Catanzaro <mcatanzaro at gnome.org> ---
(In reply to Adrian Perez from comment #5)
> I do not have the time to look right now into including the API tests
> in release tarballs, but I could find some time to review such a patch
> if someone else can write it.

FWIW this has been on my TODO list for a couple years now, but I don't promise to get to it anytime soon. I agree we should be running all the API tests in tarball releases. That should happen automatically when running 'make check' or 'ninja test' with no need to run further commands. I also want to run JavaScriptCore tests and WebDriver tests. Are there other tests we could/should be running?

But LayoutTests are just not going to be practical. Besides the issue of tarball size, they also take too long to execute during 'make check' so the benefit would be limited to the extremely few people who go out of their way to run the tests manually, and even they will probably be quite disappointed because the likelihood of tests passing outside the controlled upstream flatpak environment is quite low. The LayoutTests are just not intended to be run outside a controlled environment and would best remain upstream-only tests.

One more thing. We probably have many API tests that nobody bothered to add to any CMakeLists.txt, so they're just never built except by Apple. At least, that was the case last time I checked. Many of these tests would likely pass if they were built. This would be good to audit.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20210616/ed44f070/attachment.htm>


More information about the webkit-unassigned mailing list