Policy: Re-enabled plugin still disabled

webkit/glue/plugins/plugin_list_mac.mm

Index: webkit/glue/plugins/plugin_list_mac.mm
diff --git a/webkit/glue/plugins/plugin_list_mac.mm b/webkit/glue/plugins/plugin_list_mac.mm
index e7a23375859904c12fdb211195ce98770048a24c..d3a1ccb4f79124c613829f6fa1d3430343e981f3 100644
--- a/webkit/glue/plugins/plugin_list_mac.mm
+++ b/webkit/glue/plugins/plugin_list_mac.mm
@@ -98,8 +98,9 @@ bool PluginList::ShouldLoadPlugin(const WebPluginInfo& info,
   // (we're loading plugins hierarchically from Library folders, so plugins we
   //  encounter earlier must override plugins we encounter later)
   for (size_t i = 0; i < plugins->size(); ++i) {
-    if ((*plugins)[i].path.BaseName() == info.path.BaseName()) {
-      return false;  // We already have a loaded plugin higher in the hierarchy.
+    if (plugins->at(i).path.BaseName() == info.path.BaseName()) {

[Bernhard Bauer, 2010/12/20 22:30:28]

Is there a particular reason you're using at here? In most other places, we just
use [i] (and if you're worried about index mismatches, use DCHECK).

[pastarmovj, 2010/12/21 16:53:02]

plugins is a pointer to a vector here so to use [] op I'll have to write
(*plugins)[i]. I have seen examples in the code where pointers to arrays have
been used with at instead of dereferencing and using []. Should I change it
here?

+      // We already have a loaded plugin higher in the hierarchy.
+      return false;
     }
   }