As RSS burrows deeper and deeper into the heart of mainstream computing, the next-gen technology for processing the Web is having some growing pains. If you believe the more florid news headlines, its all about the battle for bragging rights between RSS and Atom. Head for Google or Yahoo if you want more on that rathole—what they call RSS makes no difference to me, or you. How well it works does.
Apparently, Ive been “feeling overwhelmed by thousands of unread items in NetNewsWire.” At least, thats what the normally circumspect Jon Udell
Of course, Dave Winer then pulls this particular sentence out and trumpets it on
Scobles typing speed frequently causes sonic booms, while decrypting my handwriting has overcome Moores Law. Luckily, Scoble really wanted to talk about his rant on
Next morning comes an apparently serendipitous email from NetNewsWire author Brent Simmons, pointing me at a new bug-fixed version of the beta Ive been testing. In minutes Im eagerly trying the new embedded browser functionality out on
Oh, great. Now Im in hot water with the developer of my favorite RSS information router. Maybe I should accept one of those Orkut invitations before I run completely out of friends. Or get myself a lawyer. Why, oh why did I not open my big mouth and not say any of this in the first place? What was I not thinking?
Confessions of an RSS
OK, heres the deal. My name is Steve Gillmor. I am an RSS addict. I have 4,624 unread items in NetNewsWire. Why so many? Because I have 400 separate feeds and some of them (the New York Times, Yahoo, Scobleizer) emit hundreds of items a day. Why so many unread? Because what I dont read wont get deleted. Why is that important?
What I really want is a persistent, controllable store of RSS data. Not just the abstract, or summary, data, but the full text and graphics, even scripting data, executables, and enclosures. Couple that with embedded browser rendering (Safari on the Mac) and add the ability to cache the Web pages of RSS feeds that dont support full-text.
Now add authoring system services with WYSIWYG features for dragging and dropping quotes, URLs, graphics and formatting. Safari doesnt support XML yet, but Mozilla does—and its cross-platform. Heres where Jon Udells vision suddenly crystallizes. If we have the full text, we can convert the HTML to XHTML and use XPath and XQuery to create whatever view is most appropriate to the consumer.
For me, the view thats most important is the one that reflects my interests—and the interests of those I consider most important. Some of that data already exists in NetNewsWire in an OPML file called MySubscriptions.xml—what RSS feeds I subscribe to, and in what order. The file could easily be augmented with additional data—what items I read, and in what order. Technoratis Dave Sifry calls just such an extension attention.xml.
Once I have the attention of the people I value the most, I can mine that data for insights on what they—and therefore I—most want to read and respond to. Its not the number of unread messages thats important; its the order in which I read them.
Come to think of it, using Udells XHTML strategy could render a dynamic attention-based view without touching the client code at all. That view could be exported as an RSS feed or even encapsulated as part of the current feed. And the same feed can be saved and viewed on the client when disconnected from the network.
So youre right, Brent. Generalizing about NetNewsWire based on Steve Gillmors use of it isnt fair. Its essential. As John Edwards reminded us, objects in the rear-view mirror may be closer than it seems. Or to paraphrase comedienne Carol Liefer, enough about me; what do you think about what I just read?
Discuss This in the eWEEK Forum
eWEEK.com Messaging & Collaboration Center Editor Steve Gillmor can be reached at [email protected].
Be sure to add our eWEEK.com Messaging and Collaboration feed to your RSS newsreader: