This use case is based on Google Chrome's prerender capabilities introduced with Chrome 13. We use two pages for this use case. This is page #2 of the example. See Page #1 for the full explanation.
If you clicked the link to this page from Page #1, you should see page performance results show up below. It may take a while if this is the first time you're doing the test since testing your bandwidth takes about 6 seconds. You can also click the link to Page #1 to see the same output on Page #1.
In the box below, if rt.start is set to cookie, you should see the following numbers:
If rt.start is set to navigation (or something else), then t_done is the same as t_prerender. If t_prerender is not set, then this page wasn't prerendered, and t_done is the actual perceived load time.
The latest code and docs is available on github.com/lognormal/boomerang