Add didFinishProgress BundleUIClient callback for WebKitTestRunner.
Created attachment 142092 [details] Patch
Comment on attachment 142092 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=142092&action=review > Source/WebKit2/WebProcess/qt/QtBuiltinBundlePage.cpp:77 > + 0 // didFinishProgress indentation is inconsistent with above lines
Created attachment 142237 [details] Patch with indentation fix
Comment on attachment 142237 [details] Patch with indentation fix Clearing flags on attachment: 142237 Committed r117276: <http://trac.webkit.org/changeset/117276>
All reviewed patches have been landed. Closing bug.
Is it safe to add to this client interface without bumping the version number?
(In reply to comment #6) > Is it safe to add to this client interface without bumping the version number? No, it was not safe to add to this client interface without bumping the version number. This broke WebKit nightlies, for example.
(In reply to comment #7) > (In reply to comment #6) > > Is it safe to add to this client interface without bumping the version number? > > No, it was not safe to add to this client interface without bumping the version number. This broke WebKit nightlies, for example. Apologies - I thought this was on a version 0 client. It was on a version 1 client. We're good!
(In reply to comment #8) > (In reply to comment #7) > > (In reply to comment #6) > > > Is it safe to add to this client interface without bumping the version number? > > > > No, it was not safe to add to this client interface without bumping the version number. This broke WebKit nightlies, for example. > > Apologies - I thought this was on a version 0 client. It was on a version 1 client. We're good! Thanks!
I have a deeper question about this, though. This change unskipped a qt-wk2 test only... Was this truly QT-WK2 specific? On the surface it seems like if this callback was truly important for this test, then no WK2 port could have passed without this callback. What am I missing?
(In reply to comment #8) > (In reply to comment #7) > > (In reply to comment #6) > > > Is it safe to add to this client interface without bumping the version number? > > > > No, it was not safe to add to this client interface without bumping the version number. This broke WebKit nightlies, for example. > > Apologies - I thought this was on a version 0 client. It was on a version 1 client. We're good! FWIW, we got confused about which client was changing because the bug and Change/Commit Log said BundleUIClient, but it was actually the BundleLoaderClient :)
(In reply to comment #11) > (In reply to comment #8) > > (In reply to comment #7) > > > (In reply to comment #6) > > > > Is it safe to add to this client interface without bumping the version number? > > > > > > No, it was not safe to add to this client interface without bumping the version number. This broke WebKit nightlies, for example. > > > > Apologies - I thought this was on a version 0 client. It was on a version 1 client. We're good! > > FWIW, we got confused about which client was changing because the bug and Change/Commit Log said BundleUIClient, but it was actually the BundleLoaderClient :) Sorry about that. I realized it also now :(
(In reply to comment #10) > I have a deeper question about this, though. > > This change unskipped a qt-wk2 test only... Was this truly QT-WK2 specific? > > On the surface it seems like if this callback was truly important for this test, then no WK2 port could have passed without this callback. > > What am I missing? I meant to check in other wk2 skipped lists too. But forgot about it before submitting my patch. I see it in mac-wk2 and gitk-wk2 lists as well.
Bumped the client anyway in https://bugs.webkit.org/show_bug.cgi?id=86638