RESOLVED FIXED 191394
Deprecate WKBundlePageSetDefersLoading
https://bugs.webkit.org/show_bug.cgi?id=191394
Summary Deprecate WKBundlePageSetDefersLoading
Alex Christensen
Reported 2018-11-07 12:53:37 PST
Deprecate WKBundlePageSetDefersLoading
Attachments
Patch (2.67 KB, patch)
2018-11-07 12:58 PST, Alex Christensen
no flags
Patch (4.42 KB, patch)
2019-02-05 15:41 PST, Alex Christensen
no flags
Archive of layout-test-results from ews106 for mac-highsierra-wk2 (2.95 MB, application/zip)
2019-02-05 16:53 PST, EWS Watchlist
no flags
Archive of layout-test-results from ews121 for ios-simulator-wk2 (2.80 MB, application/zip)
2019-02-05 17:28 PST, EWS Watchlist
no flags
Patch (6.89 KB, patch)
2019-02-07 10:52 PST, Alex Christensen
no flags
Alex Christensen
Comment 1 2018-11-07 12:58:41 PST
Dean Jackson
Comment 2 2018-11-12 10:14:22 PST
Comment on attachment 354133 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=354133&action=review r=me except for the build errors. > Source/WebKit/ChangeLog:8 > + The last use of it was removed in <rdar://45212964> Nit: .
Dean Jackson
Comment 3 2018-11-12 10:15:25 PST
Comment on attachment 354133 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=354133&action=review > Source/WebKit/WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextControllerPrivate.h:48 > +@property (nonatomic, setter=_setDefersLoading:) BOOL _defersLoading WK_API_DEPRECATED(_message, ...); WKWebProcessPlugInBrowserContextControllerPrivate.h:48:88: error: use of undeclared identifier '_message'
Alex Christensen
Comment 4 2019-02-05 15:41:17 PST
EWS Watchlist
Comment 5 2019-02-05 16:53:48 PST
Comment on attachment 361231 [details] Patch Attachment 361231 [details] did not pass mac-wk2-ews (mac-wk2): Output: https://webkit-queues.webkit.org/results/11045604 New failing tests: loader/load-defer.html
EWS Watchlist
Comment 6 2019-02-05 16:53:50 PST
Created attachment 361250 [details] Archive of layout-test-results from ews106 for mac-highsierra-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: ews106 Port: mac-highsierra-wk2 Platform: Mac OS X 10.13.6
EWS Watchlist
Comment 7 2019-02-05 17:28:01 PST
Comment on attachment 361231 [details] Patch Attachment 361231 [details] did not pass ios-sim-ews (ios-simulator-wk2): Output: https://webkit-queues.webkit.org/results/11045711 New failing tests: loader/load-defer.html
EWS Watchlist
Comment 8 2019-02-05 17:28:03 PST
Created attachment 361256 [details] Archive of layout-test-results from ews121 for ios-simulator-wk2 The attached test failures were seen while running run-webkit-tests on the ios-sim-ews. Bot: ews121 Port: ios-simulator-wk2 Platform: Mac OS X 10.13.6
Alex Christensen
Comment 9 2019-02-07 10:52:24 PST
Alex Christensen
Comment 10 2019-02-07 10:53:10 PST
Radar WebKit Bug Importer
Comment 11 2019-02-07 10:54:31 PST
Alex Christensen
Comment 12 2019-02-07 11:21:26 PST
Darin Adler
Comment 13 2019-02-09 11:47:38 PST
Comment on attachment 361416 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=361416&action=review > Tools/WebKitTestRunner/InjectedBundle/TestRunner.cpp:746 > -void TestRunner::setDefersLoading(bool shouldDeferLoading) > +void TestRunner::setDefersLoading(bool) > { > - WKBundlePageSetDefersLoading(InjectedBundle::singleton().page()->page(), shouldDeferLoading); > } This change isn’t quite right. We should be removing this function instead of hollowing it out and leaving it. A setDefersLoading function that does nothing isn’t valuable. It’s dangerous because it lets us run tests that won’t be testing what they are trying to test. We don’t need to keep the function to keep running the loader/load-defer.html test, because we are no longer running that test under modern WebKit, only legacy WebKit. But it’s not good that we keep running the two other tests that use it, loader/load-defer-resume-crash.html and loader/navigation-while-deferring-loads.html, because they aren’t testing what they claim to be testing any more. I suggest we change all of those tests into legacy-WebKit-only tests and remove this function entirely. Is that right, or am I missing something?
Alex Christensen
Comment 14 2019-02-11 11:32:03 PST
You're absolutely right, Darin. I did so in my next patch in https://bugs.webkit.org/show_bug.cgi?id=194506
Note You need to log in before you can comment on or make changes to this bug.