This is a very simple idea but it may not have occurred to people.

It’s a secret I’m sharing with 27K-plus people who follow me on Twitter and Facebook.

If you look at my tweetstream, you’ll see I push a lot of links. I don’t have millions of people following me, but the people who do follow me are the kinds of people who do three things: 1. They click on links, 2. they tend to read, and 3. they retweet.

I read news all the time in my realtime-aware RSS river. Not many people use it, but I do. And as I said earlier, I push a lot of links.

If there’s an interesting tech or political story that’s carried by two or more pubs, whose story do you think I’m going to push? Here’s the obvious answer —> The first one I see.

If I subscribe to your feed and it doesn’t support rssCloud, it’s going to hit my desk at the top of the hour along with all the other static feeds. But, if you’re like the small number of sites that have realtime feeds, like CNN, GigaOm and TechCrunch, I’m going to read your story right now.

I’ll have pushed your link, on average, a half-hour before I even see theirs.

When Mark Rizzn Hopkins says this stuff really works, he knows what he’s talking about. In absolute numbers you aren’t going to find very many people who are hooked into realtime feeds. But they’re exactly the people you need to reach.

This advantage won’t last long because eventually all the tools will automatically support this. WordPress does now, that’s how GigaOm and CNN got on board so quickly. I’ve met with a number of pubs who say the numbers aren’t big enough yet. Imho, they are looking at it all wrong.

Dave Winer
Berkeley, CA

PS: If you have a realtime news feed let me know!

It’s still Christmas for a few more hours and as my thoughts return to development (the last two weeks of the year are often my most productive) I’m hoping that Santa is getting ready to deliver a new API for Facebook. My hopes are up because when Facebook bought FriendFeed, they brought on board a couple of guys who know how to craft an API that works. And from what I hear they’re working on APIs for Facebook now.

What I look for in an API is simplicity, efficiency, and having all the bases covered. I don’t want to hit dead ends but I also want to reach “Hello World” in minutes, maybe even seconds. The FriendFeed API sure qualified on both counts.

So here’s what I’m looking for, Santa — in a new Facebook API.

1. It should be brain-dead easy to hurl a link at Facebook the same way I do it for Twitter. I have a simple entry point in my environment called twitter.newpost, it takes a string and a bunch of optional metadata. If I want to say Hello World to Twitter it’s just twitter.newpost (“Hello World”). And the XML that goes over the wire isn’t much more complicated.

2. These days the Facebook timeline is so much like Twitter’s, you might as well do what WordPress and Tumblr did, and implement the Twitter API. Obviously some of the limits would have to be relaxed because Facebook items aren’t limited to 140 characters. I know this is bound to be controversial, but maybe it’s just the nudge the guys at Twitter need. Besides, the advantage of supporting the Twitter API is that the lovely folk at Tweetie, Twidroid, Tweetdeck, etc will all work with Facebook. Then Joe Hewitt can get busy working on other stuff!

3. The FriendFeed realtime API. Please implement it on Facebook scale. It would change the planet.

4. Metadata everywhere. I want a fully open architecture for adding metadata. No need to get approval from anyone to add a few extra bits to a message. This will unleash huge classes of applications that flow through Facebook’s servers. A platform with the richness of desktop operating systems running at Internet scale.

5. I want to hook up RSS feeds to Facebook. You should poll them at least once an hour, and support all common realtime protocols, including rssCloud, Pubsubhubbub and SUP. If something else pops up support it too.

That’s what I want for now. Probably a million things come after this, but let’s start here!

Dave Winer
Berkeley CA

I was happy to read Richard MacManus’s article about stagnation in the RSS reader market. He said the market is in “disarray” but I believe that’s the wrong word. Disarray would imho be a good thing, because it would mean users have lots of choice, there would be competition, we’d be learning what works and what doesn’t. The problem is that Google Reader dominates the market, fully, and its view of RSS is wrong.

Fundamentally, Google Reader views RSS as email. But…

RSS is not email. It can be confusing because there are feeds that are like email. For example, GMail offers a feed of your inbox. I’ll concede that feed not only is like email, it is email. A reader for email feeds probably must behave like an email reader.

There are other feeds that you never want to miss a post on. If a member of your immediate family has a blog, you probably want to see every post. Or your boss or teacher or best friend. For those feeds you need a simple feature in your reader that allows you to click on the name of a feed to see all the posts from that feed. But that should not be the primary view of RSS, because the vast majority of RSS feeds, the ones that make RSS really unique, are not like email.

What makes it unique: RSS syndicates news.

News. Stuff that’s new. When you want to find out what’s new you don’t want to know everything, you can’t. The world is too big. There’s too much happening. If you were to get a true readout of the number of stories you didn’t read, just today, it would number in the millions. It’s a pointless number. As if it would mean anything if you got the number to be zero. All it would mean is that you spent every waking moment reading, and you had no idea what any of it meant. It wouldn’t make you smarter, happier, worth more, have more friends, get laid more often, go to heaven or become a saint. Reading every story is a meaningless concept.

I don’t know why I have the impulse to find out what’s new, and I’ve been spending decades thinking about it. But 99 percent of the time I sit down at the computer it’s what I want to know first. What just happened. Not what happened five hours ago, what just happened.

In the old days when news was distributed on paper, reporters had to chunk their stories into bundles because they were printed and distributed as collections. It was a physical reality. But if you went into the newsroom and saw how they got their news you would get a big clue about how RSS readers should work. The teletype. It scrolled stories on a printed roll, the newest story first. When you saw a story you were interested in you ripped it off. It meant that your colleagues wouldn’t get it, but that’s the reality of news. There’s nothing wrong with an incomplete view. It’s the natural order of things.

Anyway, Bloglines, Google Reader, Newsgator, etc all embarked down the wrong path, the view that news is email. Sure they added contortions that made it possible to go from story to story in reverse chronologic order. But the primary view, the one that isn’t hidden, is the one that defines the product.

Ask Richard about this. When ReadWriteWeb started, it was a Radio blog, a product I developed at UserLand. Radio was both a news reader and blog authoring tool. The biggest decision in designing that product was which view should be on the home page. We could have gone either way. I went with the blogging tool, forever cementing its position as a writing tool. Had I gone the other way, the market would have seen it very differently. That’s the way these things work.

But to see the market as stagnating you have to overlook the 800 pound gorilla running around the room throwing things all over the place. Twitter. We’re using it every day. If you follow more than a few dozen people you miss tweets all the time, and so what. Twitter doesn’t tell you how many you didn’t read because it doesn’t matter. If it’s important it will come back around again. If not, well no one can know everything.

Twitter found a way to put both the authoring tool and the reading tool on the home page. Had I cracked that nut in 2002, Twitter might have happened a few years earlier.

Dave Winer
Berkeley CA

Realtime RSS is the concept behind rssCloud and PubSubHubbub.

It does exactly what RSS does, but faster. Often much faster. And faster is better!

When your reader subscribes to a Realtime RSS feed it registers with the source of the feed. When there’s an update the source sends a short message to your reader saying there’s some new stuff.

Like RSS it’s really simple, but also really powerful.