We select nearby links too eagerly with some devices and zoom levels.
<rdar://problem/16917843>
Created attachment 231568 [details] patch
Created attachment 231570 [details] patch2
Comment on attachment 231570 [details] patch2 View in context: https://bugs.webkit.org/attachment.cgi?id=231570&action=review > Source/WebCore/platform/PlatformScreen.h:69 > + float mainScreenRelativePPI(); Relative to what?
Created attachment 231585 [details] patch3 now with missing PlatformScreen implementation
(In reply to comment #4) > Relative to what? Relative to the regular iPhone PPI 163.
Comment on attachment 231585 [details] patch3 View in context: https://bugs.webkit.org/attachment.cgi?id=231585&action=review > Source/WebCore/page/ios/FrameIOS.mm:435 > + float scale = page() ? page()->pageScaleFactor() : 1; I don't know how we'd ever get here with a null page. > Source/WebCore/page/ios/FrameIOS.mm:-469 > - // We don't approximate the node if we are dragging, we instead force the user to be precise. Did the code related to this comment disappear some time in the past? > Source/WebCore/platform/ios/PlatformScreenIOS.mm:102 > +float mainScreenRelativePPI() This name still confuses me. It sounds more like a screenPPIFactor(), and a comment in the header should say what it's relative to.
(In reply to comment #7) > I don't know how we'd ever get here with a null page. Probably can't. > Did the code related to this comment disappear some time in the past? I think so. It didn't make sense where it was. > > Source/WebCore/platform/ios/PlatformScreenIOS.mm:102 > > +float mainScreenRelativePPI() > > This name still confuses me. It sounds more like a screenPPIFactor(), and a comment in the header should say what it's relative to. Will rename, thanks.
http://trac.webkit.org/changeset/168977