UI Guy Tim Powell has been trying to recreate the super spiffy SnapBack function of Safari.

I thought I would weigh in on this a little since I have been using this excellent little browser full time since I fixed the login bug in b2.

The mechanism employed is ingenious; basically the moment you click on a link within the content of a webpage Safari tags the original page as root. Then as you navigate from link to link a tree is built that allows you to instantly return to root via the SnapBack icon.

As he noted, using a bookmark or inputing a new URL into the URL field resets the SnapBack feature.

It functions this way so that it works with any search engine, allowing you to return to whatever has been flagged as root, and since you can manually assign root as I mention below it really becomes a powerful feature. I would imagine if you could through javascript assign an indicator or flag to a history entry, and then scrape the history later for that flag and then load the corresponding history entry you would have a functional SnapBack.

One thing to note, you can manually assign a page to be the SnapBack root via the History menu, or by utilizing the key combo option-command-M, and you can activate the SnapBack by using option-command-P, and of course the search results SnapBack is stored in a different cache and to activate that SnapBack you use option-command-S.

I just thought I would throw this out incase it could help Tim in his quest (worthy as it is) to bring this feature to mozilla.