To answer your question, the getElementById was the native implementation. Although I’m not sure what prototype does with that situation, so the $ function could be a patched version. I used the latest stable version of prototype in this example (1.6.0.2).

Thanks for pointing out the bug, I was aware of it, but I usually just work around it by making sure my names and ids are all unique, although I never really thought of the meta description tag as pointed out by the bug.

see bug here:

http://webbugtrack.blogspot.com/2007/08/bug-152-getelementbyid-returns.html

Getting the result fast in IE is no good to anyone if it is the WRONG result!