Learning about Trackbacks
Today I learnt a lot more about Trackbacks - I had as I'm hoping to implement
them in ProjectDistributor.
In a nutshell, you can embed an xml fragment (RDF) in your pages that other
sites can look for. This fragment tells the other site that, if they'd
like to notify you to tell you that they have some content which references
your page that they should "ping" another page within your site - probably
something like http://mysite.com/TrackbackHandler.aspx?anIdReferencingThePageThatTheyAreTrackingBackTo
When they ping that page, you do a database write (probably after a bit of
validation that they are not a spambot) to record the trackback details.
So, trackbacks only work if the referencing site has the Trackback pinging
framework implemented at their end. Why is that you ask? Well, at
their end, they need to initiate the initail page scrape which looks for the
embedded RDF node in your page. They would do this when a page is saved on
their site, they would grab any Hrefs out of the body and, if they are external
they would do a page scrape and do the scan for the embedded Rdf node.
So, to re-cap... if you want your application to be the beneficiary of
Trackback data you need to:
- Put an RDF node in each page which specifies a Uri for the page and a page
which can be reached to record the trackback notification
- Create the page to record the trackback notification - probably easiest to
write an HttpHandler to handle requests to this resource.
- Code the HttpHandler to record the Trackback notification
The best article that I saw explaining this topic was: http://www.wwwcoder.com/main/parentid/262/site/2344/68/default.aspx
... there's also ScottW's article about what goes on with trackbacks under
the hood of .Text: http://scottwater.com/blog/archive/2004/02/06/Trackbacks.aspx