Click here to show or hide the menubar.
Thread started by Dave Winer on Thursday, March 28, 2013.

Make a Twitter out of RSS

As developers rush to fill the Google Reader hole, it seems everyone is trying to reproduce Google Reader, almost verbatim. It's understandable. There are a huge number of fans of Google Reader who will be without Google Reader as of July 1. So when they tell you what they want, it's no surprise that what they come up with is basically, Google Reader.

I think/hope they will be well-served.

But there's an opportunity to do something different that might be equally interesting. Without asking users what they want, create a product based on a different hypothesis.

A famous platform vendor once asked me if the users were asking for ThinkTank, my first product, before I created it. He had a point. There were no users before there was a product. So they couldn't ask. Because they didn't exist.

The hypothesis: What if Twitter were, in every way, open.

A picture named rssTShirt.gifWhat would that look like?

1. First, it would look like Twitter. There would be a box at the top of the page that asks What's Happening, and below that a sequence of new items from people you follow, in reverse-chronologic order.

2. It would be easy to follow someone. When you're looking at their profile page, there would be a big easy to see button that says Follow. It would not say Follow in Product X or Follow in Product Y, etc. It would not open up a huge dialog with a list of products you could follow it in. It would say Follow. And when you click it you Follow that person. No questions asked.

3. It's a web app. No synchronization among different clients. Supply an API if you like. But like Twitter it's always reverse-chronologic, and it doesn't remember what you haven't read. Yes, some users won't like that. They will use the other kind of RSS app, the ones that synch between clients. (See the first paragraph of this piece.)

4. No lock-in. That's where it gets tricky. But it's a different kind of tricky than the people cloning Google Reader will encounter. Over there, you have to synchronize news items among lots of possible clients. Here you have to let the user tell you where she stores her subscriptions (and that will determine who gets the message when the user clicks on the Follow button). Beyond that, every developer could experiment with different scanning possibilities, and different presentations, and different whatevers.

This is my dream system. I would happily make components for it. I'm just one user, for sure, and you might have to think a bit more to do this one vs the straight Google Reader clone. But there's also more opportunity to innovate. And it won't just be a repeat of the last 8 years.

XML