bucketizing more tests, fixing v8/jscore differences by re-baselining

webkit/data/layout_tests/platform/chromium-mac/LayoutTests/fast/dom/Window/README

Index: webkit/data/layout_tests/platform/chromium-mac/LayoutTests/fast/dom/Window/README
===================================================================
--- webkit/data/layout_tests/platform/chromium-mac/LayoutTests/fast/dom/Window/README	(revision 0)
+++ webkit/data/layout_tests/platform/chromium-mac/LayoutTests/fast/dom/Window/README	(revision 0)
@@ -0,0 +1,24 @@
+This file documents why the various tests have been rebaselined.
+
+atob-btoa-expected.txt
+--------------------------
+window.btoa(window) calls window.toString first. KJS returns [object Window]
+and V8 returns [object global].
+
+element-constructors-on-window-exepcted.html
+--------------------------
+V8 has different output for HTMLElement.toString.
+
+new-window-opener-expected.txt
+------------------------------
+It is a time-sensitive test. Chrome passes all tests, Safari failed some,
+so the output expects some failures. Firefox sometimes passes all tests
+and sometimes fails some. Use Chrome's output as the baseline.
+
+window-property-shadowing
+get-set-properties
+-------------------------
+For security reasons, we do not allow window.top to be shadowed.  This
+is because Flash accesses top.location and allowing the user to shadow
+top.location might therefore break Flash security checks.
+
Property changes on: webkit/data/layout_tests/platform/chromium-mac/LayoutTests/fast/dom/Window/README
___________________________________________________________________
Name: svn:eol-style
   + LF