Anatomy of a URL

A typical URL (uniform resouce locator) looks like this:

The host identifies what remote server you are connecting to.  If you choose domain, host, path, Bookdog sorts bookmarks firstly using the top-level domain (TLD).  If the TLDs are the same, it then sorts by 2nd-level domain, then 3rd-level domain, then 4th-level domain, etc.  Note that Bookdog is working from right to left, which is opposite to the way we normally alphabetize or sort regular words.  This is because the leftmost key ("www" in this example) is in fact the least significant.  It is only a host (a fish) in the domain "apple" (a harbor), which in turn is in the domain "com" (an ocean).  Working from right to left in host names is, thus, sorting with the most significant key first, as is normally done.  The idea here is that it will it group, for example, all bookmarks to web pages in the "apple.com" domain together.

More information on urls can be found in RFCs 1738, 1808, 2368 and 2396, available here.