Category: real time web

The Trace, The Scent, and the Link: Tracking the Moment

Published September 29, 2008 by cgerrish

Consuming the multicast, looking for traces of import, and then switching and focusing. Lyndon Johnson was famous for watching all three television networks at once during news broadcasts. But he didn’t consume each stream in its entirety, he was looking for cues to dig out the segments that mattered. He assembled his own narrative from this highly engaged viewing activity.

Politicians need to keep their finger on the pulse to be successful. Elvis Presley also watched all three networks at the same time. He was looking for cues to crack a different kind of code. He scanned the frequencies searching for the scent of cultural information, then quickly switched and focused.

This model was taken to the extreme in the film The Man Who Fell To Earth. David Bowie played a space alien who absorbed the local culture through a raw feed of all available broadcast channels.

The television remote control made switching simpler, but unless you could visually monitor each of the frequencies, you might miss the sign that signaled the necessity of a switch of focus. Cable television allowed the number of channels and networks to explode. Scanning the frequencies is no longer a job that can done by an individual. The Internet multiplied the possible number of channels into the millions.

Originally it was the VCR, and later the DVR and YouTube that made filtering and copying these valuable moments into a buffer for ready Network access a simple affair. Scanning the raw feed pouring off the network is now done through social media filters, perhaps most effectively by Twitter through communities of interest. A tweet containing a hyperlink is the most compact channel switcher, the most efficient pointer to items of interest.

These pointers we share through the Twitter feed point to locations in the cloud. We click and activate on-demand content that streams in to our computers. Today we think about the text, video and audio we access as a substitution for traditional broadcast and print media. But almost anything that can be expressed as software can be on the other side of that hyperlink. Here we are only limited by our imaginations.

One Comment

A Separate Reality: Identi.ca on the Brink

Published September 21, 2008 by cgerrish

Sometimes it takes a few days for the dust to settle, for all the threads to become untangled, and for the bright lines of an event to emerge. BearHug Camp defined the silhouettes of two alternate futures.

To paraphrase a politician’s recent comment, the fundamentals of microblogging are sound. The 140 character standard message length seems safe for the moment. But one senses there’s an uncomfortable feeling about the randomness of that specific constraint and its origin in SMS. Access to APIs and the ecosystem of multiple end clients providing and discovering unique new value propositions filtered from the fire hose of the full microblogging stream is pretty stable. But there’s a fear that access may be cut off, or that the economics of API access may change radically. System stability has improved measurably, but is still below acceptable major league standards. Real time messaging and track are still on the critical list, either absent or cobbled together as a pencil sketch (everything works for a small N).

Convergence on a unified microblogging standard is key to the foundation of a larger ecosystem, what Dave Winer calls a coral reef. Currently that convergence owes its existence to the mirroring of Twitter’s feature set. The distribution of power within the political economy of the system leaves this as the only avenue for progress. An open standard that departed from Twitter wouldn’t have a leg to stand on.

The first possible future belongs to Twitter. It’s a future where scaling a real time microblogging messaging system with track is key to success. The transition of the economic model of API access from free to one with some kind of usage tax will lay the foundation for a potentially dominant business model. As long as the tax is low enough and the volume high enough, Twitter will prosper and the friction they’ve introduced won’t slow down viral growth.

Oblique Strategy: Think Garden instead of Architecture

The second possible future belongs to a distributed network of players. The big scale required by Twitter’s architecture is redistributed to multiple players with different roles and responsibilities within a networked system. It’s not Identi.ca that competes with Twitter, but an ecosystem of sites that cooperate to provide the identical feature/function set along with a fertile ground for new innovation. But there’s a fly in the ointment, there is no ecosystem. Currently there are only unscalable instances of Laconi.ca that don’t connect to each other very well. In order for there to be viral growth in the Open Microblogging ecosystem the individual nodes actually need to form a network of connections. Today they don’t. The nodes aren’t nodes so they can’t grow as a network. Many aren’t competing against One.

There are couple of things missing from this garden:

Name resolution across Open Microblogging nodes
Inter-node real time public and direct messaging
Full network real time track (Aggregate XMPP Firehose)
Multiple clients for multiple devices

To the extent that these items aren’t at the top of the Open Microblogging project priority list, Identi.ca/Laconi.ca and Open Microblogging stand on the brink of an abyss. The “growth” of disconnected nodes is the illusion of growth. In a few weeks Twitter will turn on all services, introduce a small tax and the game could well be over. The acolytes of Open Source believe they will win the war because they have a structural advantage that over time will prevail. The metaphor that was used was “flipping the iceberg.” Except for the fact that there is no structural advantage and they don’t have a critical mass of users, nor a method to virally attract them. They’re living in a separate reality, their watches have stopped and their eyes aren’t on the prize.

BearHug camp showed us all the shape of the playing field, that the game was underway, and the ball was pointedly handed to the key players. Can they keep their eyes on the prize? From the opening gun, this game is being played in sudden death. The next move is crucial.

35 Comments

Continental Congress: Notes From the Bearhug Underground

Published September 2, 2008 by cgerrish

First Continental Congress

I’ve been trying to organize my thoughts about the terrain unfolding around the politics and technology of real-time microblogging. Here are my notes from the underground.

Sun Tzu said:

It is said that if you know your enemies and know yourself, you will not be imperiled in a hundred battles; if you do not know your enemies but do know yourself, you will win one and lose one; if you do not know your enemies nor yourself, you will be imperiled in every single battle.

1.0 Gillmor and Winer have called for a meetup, a camp, to bring the conversation about microblogging into the open.

1.1 As Winer noted, the users are ahead of the developers and product owners at this point in the timeline. That’s a unique opportunity.

1.2 Twitter prototyped an experience of the real-time web, giving us a glimpse of the power of IM and Track. It was withdrawn to attempt to insert a business model and reconfigure their infrastructure to support high volume real time messaging.

1.2.1 Track is the ability to follow keywords rather than specific people. This allows for vision into both sides of conversations happening outside a directed social graph and the ability to discover new relevant connections. When combined with a real-time feed it allows for the discovery of current events and conversations happening right now.

1.3 Previously, Twitter recognized the value of Track by purchasing Summize for a 10% stake in its future valuation.

1.4 When IM and Track were withdrawn, Twitter substituted an illustration of a whale for a user dialogue. Users were locked out of the dialogue and potentially about to be locked in the trunk.

2.0 Identi.ca begins to build a decentralized microblogging model that re-instates the real-time web, when combined with Dustin Salling’s Spy extensions.

2.1 Of the two services, only Identi.ca offers up the full XMPP stream to enable real-time Track. For the moment, this gives Identi.ca (Laconi.ca) the superior feature set.

2.2 To the extent that Laconi.ca deviates from Twitter’s interaction model it will destroy itself. With the exception of decentralization and Track, it has no advantages over Twitter.

2.2.1 It’s about users and network effects, not software features.

2.3 The addition of new features outside the current Twitter interaction model will not create growth. It will create confused users trying to understand the difference between Twitter and Identi.ca.

2.3.1 Some users are confused about the role of Open Source in the competition for users between Identi.ca and Twitter. If Identi.ca manages to grow to critical mass, it will have nothing to do with Open Source. Open Source is a good way to produce and maintain software; it doesn’t motivate users.

2.4 A bridge was extended across the divide between Identi.ca and Twitter. Messages, and half conversations pushed through into the Twittersphere, pointing back to the real-time web.

2.5 A two-way bridge would have the effect of flooding Identi.ca with Tweets and sucking the conversation back over to Twitter.

3.0 Twitter adds the “refers to” meta-data element. This is the piece of the puzzle that begins to radically change the shape of conversations unfolding through the Network. Tweets and Dents now addressable as URIs on the Network. Conversations potentially can be aggregated across platforms.

3.1 Gillmor proposes a three-party system including Twitter, Identi.ca (and other Laconi.ca installations) and aggregation and Track in the client (Twhirl, Gchat, etc.).

3.2 If Twitter has its own brand of Track and Identi.ca has its own brand of Track. Twitter demolishes Identi.ca. Only a three-party system will preserve Identi.ca and the ecosystem.

3.3 If neither Twitter nor Identi.ca have Track and real-time messaging. Twitter demolishes Identi.ca. Identi.ca’s highest priority has to be maintain its real-time Track advantage over Twitter. That window will close soon.

4.0 If you’re wondering why Twitter is winning, it’s because they don’t have to care about establishing a multi-party ecosystem for micro-blogging. They can simply wait until their competitors destroy themselves. (See quote by Sun Tzu at the top of the post)

4.1 All Continental Congresses start as a brawl. But as with Internet Identity, once the vendors understand they can’t, and shouldn’t own the space, cooperation is engendered.

4.2 Winer called Twitter a Corral Reef. Now we’d like the whole micro-blogging ecosystem to be a Corral Reef. It’s an opportunity happening right now in real time.

4.3 A strong bear hug at the right time, with the right players and the right conversation could create a solid foundation, a Corral Reef.

3 Comments

Small World Theory: 6 Degrees of Micro-Communities

Published August 25, 2008 by cgerrish

Just a short thought experiment: picture, if you will, the kind of network graph you’d draw to represent traditional broadcast and print media. Initially a very small set of one to many one-way relationships. All downstream, very little upstream– perhaps the letters to the editor section. Desktop publishing changed the look of that graph, as did the personal video camera, lighting up more broadcast nodes on the network, but distribution remained a challenge.

Blogs, Podcasting, YouTube and RSS changed the shape of the picture even more substantially. Distribution moved to the common platform of the web and the economics supporting a publishing node changed radically. More publishers light up on the network, but more importantly the means for two-way traffic is established as publishers talk to each other. Two-way traffic expands to a many-to-many relationships and micro-communities begin to form. All of this built on the back of HTTP.

Now think about who has real time broadcasting capability and draw a mental picture of that network graph. Think of the shape of the network, it seems to me the traditional model still dominates. Facebook, MySpace, Dogster, LinkedIn and others concentrated and increased the speed of communication transactions within communities– but they don’t generally achieve real time continuous message flow. Twitter, and more recently, Identi.ca have achieved message flow liquidity and have established themselves as primary markets. As the XMPP protocol starts capturing the imagination and islands of Laconica instances begin appearing, more real time nodes light up on the network. It’s early days and there aren’t a lot of dots to connect.

Whether or not those dots will be allowed to be connected is currently in question. Our ability to track those XMPP streams is even more fragile still. There’s a real time web emerging and we’ve yet to imagine how it will manifest. It’s something we’ll have to talk to each other about.

The power of real time micro-communities is broader than common wisdom would suggest. Each broadcaster in a real time micro-community is connected and messages to a different circle. We misunderstand the nature and power of micro-communities if we focus on the number of connections in a particular circle. Each circle is embedded in a network of circles, but the network of circles occupies a small world. And you already know this: the whole world is connected through six degrees of separation.

2 Comments