[webkit-reviews] review denied: [Bug 7855] REGRESSION: document.open() in a script added through document.write() causes a crash : [Attachment 7307] proposed fix

bugzilla-request-daemon at opendarwin.org bugzilla-request-daemon at opendarwin.org
Mon Mar 27 10:07:49 PST 2006


Geoffrey Garen <ggaren at apple.com> has denied Geoffrey Garen
<ggaren at apple.com>'s request for review:
Bug 7855: REGRESSION: document.open() in a script added through
document.write() causes a crash
http://bugzilla.opendarwin.org/show_bug.cgi?id=7855

Attachment 7307: proposed fix
http://bugzilla.opendarwin.org/attachment.cgi?id=7307&action=edit

------- Additional Comments from Geoffrey Garen <ggaren at apple.com>
I like the approach here, but I have some quibbles:

1. nodeSetsInited. Seems unnecessary. Also somewhat vague. Why not just:

ASSERT((!newNodes && !nodeExtraRefCounts) || (newNodes && nodeExtraRefCounts));

if (!newNodes) {
  // create them
}

2. Why re-create the newNodes hash table in clearNewNodes()? Seems like
unnecessary work. If it needs to be re-created, it will be, the next time a
Node object is constructed. I also think that creating the newNodes hash table
in only one place is more in line with your goal to streamline the code.



More information about the webkit-reviews mailing list