You may notice a new link on the side of the blog to my new stalker feed. This feed is an RSS feed which combines all the items from all of my websites. So if you want to watch my blog, podcast, listal, del.icio.us, digg, Flickr, Last.fm and all that other stuff bunched into one feed, you’ve got it. Now you can monitor pretty much all of my Internet activities in one place.
I first attempted to find a free program I could download to run this on my own web server. That doesn’t seem to exist in any good and usable form. So I tried to write my own. This proved to be more difficult than I first realized. Since nobody seemed to format their feeds in a standard way, it was turning out to be a daunting adventure in XML parsing.
So in the end I combined two different services to perform the task. The first service is FeedJumbler which does a serviceable job of combining feeds. My only complaint is that it destroys a great deal of the extra data in the feed items. The other service is, of course, FeedBurner. It does a good job of taking the jumbled feed and making it a feed that is useable. It also adds in my del.icio.us feed and my flickr feed without killing the extra data.
I hope one day soon I will be able to use FeedBurner for this entire operation. So many people want to combine many RSS feeds into one, but most services that perform this function are meant for people who read rss rather than those who publish it. Not only that, but so few of them actually parse all of the XML properly causing the inevitable loss of useful tags.
I know from coding a solution why this is. RSS sucks. From the user perspective it’s great because programs that automatically create RSS, like WordPress, and programs that read RSS, like Firefox, already exist and work great. From a developer’s perspective it is a nightmare. I was not able to find a single RSS library in existence that actually handles anything that might be in a feed. No matter what tags from whatever XML namespaces appear in a feed, it should not be ignored or removed. People obviously are pretty good at coding software that writes RSS which is chock-full of semantic web happiness, but nobody has written anything that can make use of all of it let alone just hold onto it. Every library took the RSS data and destroyed at least some of it to make it easier for developers to handle it programatically. Then they made it incredibly difficult to re-write a new RSS feed.
I’m obviously a big fan of the uses of RSS, but not until we have intelligent RSS handling libraries which can actually make use of the full potential of RSS will we see more new and exciting things like podcasting.