Their was a 3-part blog post back in 2004 that always stuck with me as I have observed the evolution of the Internet. And I’ve found myself often referring to the term “Stock and Flow” when discussing the real-time web. This just came up again in a comment that I wrote (but not yet submitted) on Chris Messina’s blog post entitled “What can dogs tell us about the real-time web?” – http://factoryjoe.com/blog/2009/09/16/what-can-dogs-tell-us-about-the-real-time-web/ –
So let me officially refer to this insightful article authored by Common Craft’s Lee Lefever.
Common Craft Blog
Introduction to Stocks and Flows in Online Communication (parts 1, 2, 3)
By Lee Lefever
http://www.commoncraft.com/archives/000593.html
http://www.commoncraft.com/archives/000599.html
http://www.commoncraft.com/archives/000601.html
Probably the most interesting aspect to the discussion of the real-time web is not speed of human consumption but rather speed of machine consumption. Correction. This is not necessarily true. People are filters too. Very good filters! In fact, I once started a blog called SpreadTheMedia.org in 06 and the core idea was “Spreading Human Filtered Content”. I valued the human touch that would bubble up content for others to consume. What I meant to point out was the difference between the speed of consumption and the speed of filtration. Machines equipped with smart algorithmic tech will do much of the real-time filtering…. Google, Bing and the other usual suspects as well as new entries to the market. These engines will need to do anti-spam, relevancy, source ranking etc. And in mere seconds so that the intended consumers can get their filtered data.
Rapid Failure, Rapid Intelligence
In a recent post here – http://vocal.ly/p4l – I touched on this stuff a bit.
The notion of rapid failure came to mind as a way to compare how content and content sources need to go through some checks and balances and if their is a rapid failure log recorded, then the filter becomes more efficient and therefor more intelligent. It was not a perfect analogy but the point is clear… Nobody wants to get spam, worms, hoaxes, rumors or any inaccurate data. Nobody wants to spend time fact-checking. The system should handle this for us. And when it does, it will have a nice ripple effect on the infosphere. This doesn’t come without some issues that will breed skepticism and controversy, however.
Trust Analytics
It’s going to be important to rely on the intelligence of the real-time data crunching machines as well as people who act as filters themselves. But in both cases, their will be flaws and corruption contaminating results. It is probably inevitable that their will be stocks formed from the flows that are biased info repositories that will cater to certain minded folks. Many people are set in their ways and like to believe what they want to believe. What i’m getting at here is…. Their will always be layers of filtration and many networks of real-time streams. No different than TV, really. Some will watch Fox, some will watch CNN. I see that i’ve narrowed this down to news gathering. Suppose that makes good sense though. “News” makes up much of the real-time stream. Even if accompanied by a comment, it’s mostly a link and a headline.
So how do look at this filtering point? How does it not become any different than the faulty filters we have already in place on current mediums like TV, Radio? I am reminded of the recent US Presidential Elections and the CPD that make the Debates nearly impossible to qualify for if you are a 3rd party candidate. To me, that is a filter that does us no good. Or the fact that the MSM also does not give fair airtime to candidates unless they are the front-runners, thus maintaining the front-runners status for the duration of the elections. That is a filter that does us no good. Or how about subtle changes to bills that are not reported on so that they can slip through without an outcry? These are just a few examples of bad filters that exist in the world.
Will the web’s real-time filters be an improvement? How many filters will there be? How will we look at these filters? How will they compare to traditional streams of content? Will the Flows be more important than the Stocks or vise versa? At the end of the day, does the landscape change at all? Does The Real-Time Web bring us more noise or can it bring us more signal?
It will be interesting to watch this play out.
And to clarify, The Real-Time Web will be more of an umbrella term as it bleeds into more things. The popularity of Microblogging (Twitter) brought Real-Time content to the forefront, even though twitter isn’t really any more real-time than email or RSS set to poll the server every minute. But it’s the micro chunk data that caught on and the illusion of real-time that got people talking more about it.
Twitter as a service is people powered. People often post links to new pages on the web… news articles, blog posts etc. These people are probably first alerted of this new content via RSS or Email. Someone inevitably is first to re-post content to twitter alerting their followers and anyone who re-posts… the gossip propagates quickly. This is what some will refer to as Real-Time Web. And that’s acceptable.
Other’s will point out that Real-Time Web is how content from one service immediately flows to other services. A great example is Twitter data instantly hitting Friendfeed.com. Twitter does have a data “Firehose” that they let some partners tap into. Usually the technology involved in capturing the data is called long-polling or http streaming. So Twitter is involved with Real-Time technology but it only truly is exposed via the Firehose, not the consumer side where people are using twitter clients to monitor their streams (thats just like email or rss readers).
The Real-Time Web is also effectively now related to RSS and Atom feeds with the advent of rssCloud and PuSHb. Now we can be notified faster when sites with Real-Time RSS feeds are updated. How Stocks and Flows will apply to this will be seen in the next breed of RSS Readers where you can organize your subscriptions more intelligently using filters and groupings etc. These types of features already exist but will become more important with Real-Time Feeds.
I’ll stop this here since i’m sure this is getting a bit unfocused. Check out Common Craft articles and Chris Messina’s Dog analogy!