Bookmarking a page that has a PURL is difficult. The user must know about the PURL and must manually create the bookmark. It would be very nice if the user agent provided a means to bookmark the PURL instead of the real URL. To do this, the agent would need to know the PURL. While agents could figure this out while resolving a PURL, it would be much easier if each page specified its own PURL internally. This is done most easily by using the LINK element, with the 'rel' attribute set to 'purl'. What follows is an RFCish description of how user agents should deal with PURLs.PURLs
Wednesday, February 01, 2006
I don't understand why Persistent URLs are not more common. Using the LINK element in HTML documents to indicate a page's PURL is another tool that may make them more common.