Pingback and embedded metadata in (X)HTML
More on Pingback vs TrackBack on Hixie's Natural Log.
How embarassing - he points to my referrers as a typical list. :) Mine are crap. Look at how Mark Pilgrim handles referrers. Yesterday I was working at making my referrer tracking harvest titles, clean out false links, and collapse redundant backlinks, but I'm far from perfecting that. But, at the same time, I agree: Referrers are not enough. They're one source, the most noise-ridden but the most effortless on the part of the outside contributor. But you can only do so much with almost nothing. :)
I think, when it comes down to it, my only issue with Pingback is not a Pingback-specific issue at all: How to harvest machine readable metadata from a web resource. This applies to my referrer links, Pingback, and TrackBack alike. TrackBack has a bit of a solution, with embedded RDF, but that's got its own issues. Ian suggests a few things to me in comments, such as harvesting the title from the HTML title tag (a no brainer), and then harvesting further data from DublinCore-based data in meta tags in the page. I've seen this last convention only once before, in the geographical data consumed by Syndic8.com.
Is this a pretty common convention? I've not seen it done much, but the I obviously have not seen everything or a large chunk of anything. :) If this is a known convention, it makes me happy and I think it would answer a question I asked back in May.
Update: Duh. Yes, it's a known convention. It's even got an RFC: RFC2731: Encoding Dublin Core Metadata in HTML Simple Google search. Sometimes I can be so daft. :) Now I just have to start using this more - and I wonder why more people aren't using it? Most likely because there's been not much in it for them.
shortname=ooocoo