<rdf:RDF
	xmlns="http://purl.org/rss/1.0/"
	xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
>
<channel rdf:about="http://interconnected.org/home/mini/_9517">
	<title>Interconnected Mini Links</title>
	<link>http://interconnected.org/home/mini/_9517</link>
	<description>Links alongside interconnected.org (Single link 9517)</description>
	<dc:rights>Copyright Matt Webb</dc:rights>

<items><rdf:Seq>
<rdf:li rdf:resource="http://interconnected.org/home/mini/_9517" />
</rdf:Seq></items>
</channel>
<item rdf:about="http://interconnected.org/home/mini/_9517">
<title>Bloglines API Documentation</title>
<link>http://bloglines.com/services/api/</link>
<description><![CDATA[<p>"Bloglines provides several mechanisms for accessing account information. The Notifier API is used to quickly gather a count of unread items in a Bloglines account. The Sync API is used to accessing subscription lists and unread blog items. The Blogroll API is used for incorporating subscription lists in other sites."</p><p><a href="http://inessential.com/?comments=1&amp;postid=2922">http://inessential.com/?comments=1&amp;postid=2922</a> -- NetNewsWire and the BlogLines API</p><p><a href="http://inessential.com/?comments=1&amp;postid=2923">http://inessential.com/?comments=1&amp;postid=2923</a> -- more on NetNewsWire and the BlogLines API</p><p>It's quite exciting having a system like this. Once it unfolds fully (sync bookmarks, multiple systems offering the same API), then two consequences:
1. Manage bookmarks in NNW, read in BlogLines on the road, get a subset read on your mobile phone. The mobile app is tiny because it doesn't even need to take care of fetching feeds, caching, or marking as read. Easy to build!
2. Distributing the load. The big services will sync together using something like NNTP, which is hard to set up, but that's fine because they're big (there will be a dozen or so when this starts, but people will have local mirrors, and heavy readers will also run NNTP). Weblog pings are used to populate the system, which relieves the load on the individual weblogs being fetched (spidering can't last that much longer, surely, given the scaling problems). Standard HTTP is used for clients to fetch feeds, which means the people who want the traffic get it, and HTTP works happily through all firewalls. Interesting mix of technologies to make a distributed system. I'll bet you a tenner it turns out like this. ETag-Vary or whatever it is, it's only a stopgap solution.</p>]]></description>
</item>
</rdf:RDF>

