RSS needs: redirection
Jon Udell points out the need for a way to redirect an RSS feed. He was subscribed to the RSS feed for Olivier Travers’ Webvoice blog, which used to be powered by Julian Bond’s RSS scraping service at voidstar.com/rssify.php. Now that Blogger Pro produces RSS directly, Olivier’s feed is at webvoice.blogspot.com/rss/webvoice.xml, but Jon accidently subscribed to Julian’s (mostly silent, these days) feed instead. A nice example of the need for feed redirecting, and the confusion that can result from serving things through another service.
This is why Syndic8.com implements my ideas on Redirected and Duplicate status for feeds. You can check the feed’s info page and see that it’s been redirected. Or make a web services call to do it programmatically. Put your list of feeds into a personal list and then programmatically pull down a status report on them.
But as to feeds themselves doing redirection, that can do it right now. HTTP supports a whole range of robust status codes on a page. You can redirect, move temporarily, indicate dead, really dead and gone and a number of other choices. The problem is several of the RSS readers out there don’t pay attention to this.
So before we reinvent the wheel here, how about taking a look at the functionality already provided in the HTTP specs?
This is also where an RSS feed using RDF modularity could simply setup rdfs:seeAlso and other attributes. But hey, what do I know, I’m only suggesting people actually research the stuff…
HTTP redirects are cool if you can do them (and if they are supported: note to self – set up a 301 page and subscribe to it in various RSS readers and evangelize the authors that don’t follow it), but in this case, it doesn’t apply: Olivier can’t exactly set up a redirect on a query string sent to a script on someone else’s site.
rdfs:seeAlso is a good start, but AFAIK it doesn’t offer any way to describe what’s at the other end of the seeAlso: is this pointing to the new location of the feed, or the FOAF file, or an RDFed manifest of other feeds from the site, or what?
I’m not clear on what Oliver can’t do. If one susbscribes to the wrong URL, well that’s just like dialing a wrong number. Now if that URL ’moves’ and the scraper can’t figure it out then the scraper needs to be improved. The source of which are available so have at it. We have an XML-RPC interface into Syndic8 that would let you programmatically look it up. Go to town!
Yes, evangelizing the reader developers has been successful. Slowly but surely they’re getting up to speed on all this stuff. It takes time and most of them are doing it as an extension of their own needs. As a result some stuff is slow to improve. But they do generally want to get around to it.
If you want the sophistication of automatic stuff you’re definitely going to want to use a framework.