compiling links and images from RSS

["jdklub" <pseudonym@mindspring.com>]

There are a number of RSS feeds available that include HTML markup 
in the description field. The most popular markup (based on my un-
scientific tests) is the <a> tag, while the most interesting is 
often the <img> tag (which, sadly, few use in weblog posts). There 
are debates as to whether or not such markup is legal, and if so, 
how it ought to be encoded. But that's beside the point. It's there, 
so why not have fun with it?

I've started slurping links and pictures out of RSS feeds to see 
what the results would look like. Here are the results. 

Links:
http://fyuze.com/links/

Images:
http://fyuze.com/images/


I've found it entertaining to click through the out-of-context links 
to see where they take me. Given a larger number of feeds with 
embedded markup, one could produce an RSS blogdex. More 
interestingly, if the feed base was large enough one could query to 
find out who is linking to who (ala Metalinker, or is it 
Metablogger?). That is, you could pass it the URL to your latest 
blog entry and it could return a list of sites linking to you. Sort 
of like MT's TrackBack, but without the required ping. I hope to 
implement this when the feed base grows a bit. 

The image list is small, since few people include images. However, I 
think it has a lot of potential to create interesting composites. 
Imagine what such a list might have produced on 9/11/2001. There 
also exists the possiblity of building the list via category (just 
news photos, just european news photos, etc). Again, not enough 
feeds included images to make this very useful at present.

And it seemed only fitting to produce RSS versions:

Links:
http://fyuze.com/api/rss/?op=links

Images:
http://fyuze.com/api/rss/?op=images

If you want to participate, simply add your site to the list of 
sites fyuze collects at http://fyuze.com/customize/suggest.php and 
make sure the markup is included in your RSS feed (encoded, 
unencoded, or as CDATA).