summaryrefslogtreecommitdiff
path: root/Source/WebKit/blackberry/ChangeLog
diff options
context:
space:
mode:
Diffstat (limited to 'Source/WebKit/blackberry/ChangeLog')
-rw-r--r--Source/WebKit/blackberry/ChangeLog207
1 files changed, 207 insertions, 0 deletions
diff --git a/Source/WebKit/blackberry/ChangeLog b/Source/WebKit/blackberry/ChangeLog
index e4cb6eb99..62914fd11 100644
--- a/Source/WebKit/blackberry/ChangeLog
+++ b/Source/WebKit/blackberry/ChangeLog
@@ -1,3 +1,210 @@
+2012-10-22 Jocelyn Turcotte <jocelyn.turcotte@digia.com>
+
+ [Qt] Fix "ASSERTION FAILED: !document->inPageCache()" when loading a page
+ https://bugs.webkit.org/show_bug.cgi?id=98514
+
+ Reviewed by Kenneth Rohde Christiansen.
+
+ * WebCoreSupport/FrameLoaderClientBlackBerry.cpp:
+ (WebCore::FrameLoaderClientBlackBerry::transitionToCommittedForNewPage):
+
+2012-10-20 Andrew Lo <anlo@rim.com>
+
+ [BlackBerry] m_mainFrame should be null checked before use in WebPagePrivate destructor
+ https://bugs.webkit.org/show_bug.cgi?id=99916
+
+ Reviewed by George Staikos.
+ Internally reviewed by Arvid Nilsson.
+
+ Internal PR228029
+ Add null checks for m_mainFrame for things that happen during
+ WebPage destruction.
+
+ * Api/WebPage.cpp:
+ (BlackBerry::WebKit::WebPagePrivate::setVisible):
+ (BlackBerry::WebKit::WebPagePrivate::setCompositorDrawsRootLayer):
+
+2012-10-19 Mike Fenton <mifenton@rim.com>
+
+ [BlackBerry] Convert input style to int64_t
+ https://bugs.webkit.org/show_bug.cgi?id=99848
+
+ Reviewed by Yong Li.
+
+ PR 227381.
+
+ Switch input style mask to int64_t to match the the updated definition.
+
+ Reviewed Internally by Nima Ghanavatian.
+
+ * Api/WebPageClient.h:
+ * WebKitSupport/InputHandler.cpp:
+ (BlackBerry::WebKit::inputStyle):
+ * WebKitSupport/InputHandler.h:
+ (InputHandler):
+
+2012-10-19 Mike Fenton <mifenton@rim.com>
+
+ [BlackBerry] Move all keyboard type calculations into InputHandler.
+ https://bugs.webkit.org/show_bug.cgi?id=99847
+
+ Reviewed by Yong Li.
+
+ PR 227381.
+
+ Move all keyboard type calculations into webkit and avoid passing type
+ information to the webview.
+
+ Reviewed Internally by Nima Ghanavatian.
+
+ * Api/WebPageClient.h:
+ * WebKitSupport/InputHandler.cpp:
+ (BlackBerry::WebKit::convertInputTypeToVKBType):
+ (WebKit):
+ (BlackBerry::WebKit::InputHandler::setElementFocused):
+
+2012-10-19 Jacky Jiang <zhajiang@rim.com>
+
+ [BlackBerry] Page viewport aspect ratio not kept on navigating back
+ https://bugs.webkit.org/show_bug.cgi?id=99760
+
+ Reviewed by Yong Li.
+ Internally reviewed by Konrad Piascik
+
+ PR: 222437
+ This is to fix the second issue on PR222437 that the viewport not kept
+ when navigating back from google after rotation to bloomberg.
+ We were trying to restore the history contents size and scale of
+ bloomberg which has virtual viewport. The scale would be recalculated
+ based on the new screen size. However, we would still use the current
+ contents size if it was greater than the history contents size in
+ FrameView::setContentsSize which made the contents seem to be scaled
+ too much.
+ Actually when we relayout the contents after rotation change, the
+ current contents size which is based on the virtual viewport has taken
+ the screen size into account. Shrinking the contents size will be
+ expensive. So we can ignore screen size based scale calculation here.
+
+ * WebCoreSupport/FrameLoaderClientBlackBerry.cpp:
+ (WebCore::FrameLoaderClientBlackBerry::restoreViewState):
+
+2012-10-19 Mike Fenton <mifenton@rim.com>
+
+ [BlackBerry] Add input logs to InputHandler::ensureFocusTextElementVisible.
+ https://bugs.webkit.org/show_bug.cgi?id=99837
+
+ Reviewed by Antonio Gomes.
+
+ Add input logs for InputHandler::ensureFocusTextElementVisible.
+
+ Reviewed Internally by Nima Ghanavatian.
+
+ * WebKitSupport/InputHandler.cpp:
+ (BlackBerry::WebKit::InputHandler::ensureFocusTextElementVisible):
+
+2012-10-19 Andrew Lo <anlo@rim.com>
+
+ [BlackBerry] Should suspend AnimationController when WebPage becomes invisible.
+ https://bugs.webkit.org/show_bug.cgi?id=99844
+
+ Reviewed by Yong Li.
+ Internally reviewed by Arvid Nilsson & Yong Li.
+
+ Suspend animation controller when the tab becomes invisible,
+ in addition to suspending scripted animations.
+
+ * Api/WebPage.cpp:
+ (BlackBerry::WebKit::WebPagePrivate::setVisible):
+
+2012-10-19 Mike Fenton <mifenton@rim.com>
+
+ [BlackBerry] Don't attempt to auto scroll on input for Fixed Position elements.
+ https://bugs.webkit.org/show_bug.cgi?id=99833
+
+ Reviewed by Antonio Gomes.
+
+ PR 195024.
+
+ Don't attempt to scroll fixed position elements.
+
+ Reviewed Internally by Arvid Nilsson.
+
+ * WebKitSupport/DOMSupport.cpp:
+ (BlackBerry::WebKit::DOMSupport::isFixedPositionOrHasFixedPositionAncestor):
+ (DOMSupport):
+ * WebKitSupport/DOMSupport.h:
+ (WebCore):
+ * WebKitSupport/InputHandler.cpp:
+ (BlackBerry::WebKit::InputHandler::ensureFocusTextElementVisible):
+
+2012-10-18 Jacky Jiang <zhajiang@rim.com>
+
+ [BlackBerry] Page viewport aspect ratio not kept on navigating back
+ https://bugs.webkit.org/show_bug.cgi?id=99760
+
+ Reviewed by Rob Buis.
+ Internally reviewed by Konrad Piascik.
+
+ PR: 190469
+ This was added in r127628 which was trying to fix some rendering
+ defects on maps.google.com and news.qq.com, however it was very harmful
+ and caused regressions.
+ zoomToFitScale() depends on the current contents size. During history
+ navigation and device rotation, if zoomToFitScale() is based on the
+ final contents size which is expected, there will still be a scale
+ calculation like this afterwards
+ screenSize.width()/screenSize.height()*zoomToFitScale() which will
+ result in a wrong scale. During the history navigation, if
+ zoomToFitScale() isn't based on the final contents size, then we will
+ get the wrong scale as well for sure.
+ This issue will be fixed by removing the wrong calculation.
+ The original issue of r127628 was fixed by the patch of PR225981.
+
+ * WebCoreSupport/FrameLoaderClientBlackBerry.cpp:
+ (WebCore::FrameLoaderClientBlackBerry::restoreViewState):
+
+2012-10-18 John Griggs <jgriggs@rim.com>
+
+ 2012-10-18 John Griggs <jgriggs@rim.com>
+
+ Crash in InRegionScrollableArea constructor while loading page
+ https://bugs.webkit.org/show_bug.cgi?id=99733
+
+ Reviewed by Antonio Gomes.
+
+ Check scrollLayer and avoid constructing InRegionScrollableArea when it is NULL.
+
+ * Api/InRegionScroller.cpp:
+ (BlackBerry::WebKit::InRegionScrollerPrivate::calculateInRegionScrollableAreasForPoint):
+
+2012-10-17 Jacky Jiang <zhajiang@rim.com>
+
+ [BlackBerry] Possible redundant zoomToInitialScaleOnLoad() in WebPagePrivate::setLoadState
+ https://bugs.webkit.org/show_bug.cgi?id=99624
+
+ Reviewed by Rob Buis.
+ Internally reviewed by Arvid Nilsson
+
+ PR: 225981
+ Several reasons to eliminate the ancient unnecessary
+ zoomToInitialScaleOnLoad() in WebPagePrivate::setLoadState committed:
+ - At this point, we have only committed the data we received.
+ - For the new load, we haven't started parsing the tokens and building
+ the DOM tree and the first layout will happen after that.
+ - For the load restored from cache, the first layout will happen after
+ setLoadState committed which is called by
+ FrameLoader::dispatchDidCommitLoad().
+ - zoomToInitialScaleOnLoad() can still be called after the first
+ layout finished.
+ - zoomToInitialScaleOnLoad() is expensive and it will boost the load
+ performance after the removal.
+ - Removal of this can fix some weird rendering defects on yahoo.com,
+ maps.google.com, news.qq.com and so on during load and rotation
+ mentioned on the PR.
+
+ * Api/WebPage.cpp:
+ (BlackBerry::WebKit::WebPagePrivate::setLoadState):
+
2012-10-17 Mike Fenton <mifenton@rim.com>
[BlackBerry] Prevent focus zoom on non-userscalable pages.
View Lorry Controller Status