For years Digg has had an active comment community, where the comments are submitted and appear on the Digg landing page, rather than on the article linked from Digg. FriendFeed got into this game by making it possible to comment on content pulled in from multiple web services, where all the comments appear on FriendFeed, rather than on those services. Today, the tech blogosphere is debating a service called Shyftr that allows users to comment on the full text of blog posts, drawn from full text RSS feeds.

These are all forms of disintermediation on the web — disintermediation defines distribution on the web, made possible by RSS and hyperlinks.

The funny thing is that disintermediation is like a hall of mirrors — there’s really no end to it.

As Dave Winer posits:

There will of course eventually be the equivalent of Technorati, which assembles in one place, all the comments about each blog post.

In other words, somebody will come along and disintermediate FriendFeed, Shyftr, and all these other services by pulling all of the comments created on those services into yet another service.

How about a WordPress plugin that gathers all the comments about my blog posts on FriendFeed, Shyftr, etc. and displays them right here? That’s the beauty of data on the web — you can easily disintermediate the disintermediator.

Which is why disintermediation, i.e. CONTROL, is the wrong way to look at — it’s a very analogue media (i.e. old media) mindset.

The objective on the web isn’t to keep the data on your site — it’s to have an open exchange of data. It’s a wonderfully counterintuitive way of thinking.

If Shyftr or FriendFeed or anyone else wants to enable their users to comment on my blog posts, I say fine — so long as I can easily display any comment created on their services over HERE. Heck, I’d even give them a feed of the comments created here to display over THERE.

Think about it from the user perspective — too often these debates overlook what would best serve the USER.

The user is ill-served if there are conversations about a piece of content going on across multiple services, and the user has to go to each service to participate. Those participating in a conversation on one service are ill-served because they can’t hear what’s been said on other services.

What would be best for users is if all the services were connected, so that all the data appeared on EVERY service, and it didn’t matter which service I used to read or contribute — the data would propagate throughout network.

Remember, it’s the WEB — the network, right? Stop obsessing over YOUR blog or YOUR service or YOUR node — focus on enabling EVERYONE’S network. There’s only ONE web.

The pressure isn’t just on Facebook to open the flow of data — it’s on every web service and every publisher.

The winners won’t be those that control the most data — the winners will be those that channel the most data — and those that create the most value on top of the data flow.

Always remember Google.

Google makes billions of dollars by sending people away. Google doesn’t control any of the titles, URLs, or excerpts that appear on its search results pages — these are all determined by publishers. Google doesn’t control any of the pages where AdSense ads appear.

Google’s business model isn’t about controlling data or people — it’s about CHANNELING data and people. The more people and data that pass THROUGH Google, the more money Google makes. That’s why they offer a free blogging service and want to offer free internet service. That’s why they are evangelizing OpenSocial.

Google profits from liquidity in the system, from the free flow of people and data. Google dominates the web because they are the only company that really understands this.

You could say that Google’s control over the web is a function of its lack of control.

Go meditate on that for a while. I will, too.