That's helpful.  Now, two things strike me.  First, it's news to me.  That's not just because the media do not take much note of the complexity of telecommunications, as "mere" technology or business news.  It's not just, too, because it affects the media interests themselves. It's also because it's been almost impossible for Edwards to get the coverage accorded Clinton and Obama . . . or at times even Gore, who at least perpetually fascinates the OpEd regulars of The Times by smirking, presumably at them. This media bias toward the candidates whose "angle" they understand is shameful, as it alters Edwards's chances and thus the democractic process.

Second, it suggests a reason for Brooks's assault on Gore the other day apart from the main point well made by Jim Sleeper, that it's the GOP preemptive ad hominem attack on Gore again.  Brooks, like many print columnist, is naturally going to be upset about claims for the Internet as an alternative, even if I myself find them exaggerated.

John 

http://www.haberarts.com/

It is difficult, to say the least, to get a fair hearing into arguments for media reform and regulation when those who dominate the means of communication remain comfortably deregulated. Further, it will likely be difficult to find any clear presentation of communications issues via the communications industry as long as it is in the best interest of its corporate shareholders to avoid the conversation altogether.

By the way, does anyone know what ever became of HR 3302, the Media Ownership and Reform Act of 2005?  As far as I can tell, it's still bottled up in committee.

But if Edwards wins, it will be a different FCC. Likewise, probably, if Obama wins. It's more iffy if Hillary wins, because she's the closest to big media interests.

Imagine, Edwards was on Gore's short list for VP. What a pair they would have been. And Edwards would be our next PResident after Gore. This is a huge defining issue. Only visionaries get it. I don't see either Clinton or Obama holding back the telecoms. Edwards gets it.
"A thing moderately good is not so good as it ought to be. Moderation intemper is always a virtue, but moderation in principle is always a vice." Tom Paine

Fair enough.

Not surprised Edwards is out front on this. He and Elizabeth are big proponents of the tubular internets. They understand it and what's at stake.

QoS is probably valuable (given a proper implementation, which is a nontrivial task) in the context of a single organization's communication network. The typical example is anyone's home network, I can imagine QoS being valuable to help manage latency for isochronous media streams, while deprioritizing ordinary packet flows, such as backups or surfing. In this case, however, there is one single unambiguously defined owner of the home net.

The Public Internet, on the other hand, has no single owner; it's more like a public utility (I hope). Ideally, one pays for access to bandwidth on an equal footing with other payers. Organizations with high bandwidth requirements pay proportionally for their higher bandwidth.

Among the questions about QoS and similar technical means for differentiation, there is a question of how the effectiveness of such mechanisms are preserved across diverse network providers. The definition of network as a lossy FIFO is, by now, a fairly well understood technical problem. QoS and differentiation would increase the complexity of bandwidth peering arrangements dramatically. (Hey, this might be cool! Maybe the industry would see some growth and rehire communication and protocol guys like me...)

Net neutrality is not so much about QoS as it is about content. Right now, bandwidth providers *say* they want to provide differentiated service without reference to content. That is a nice thing to say, but all the major providers today have a financial interest in specific content or services (Time-Warner is one example, there are others).

Net neutrality is an attempt to prevent bandwidth providers who have a financial or other less well-defined interest in particular content from manipulating the marketplace.

As long as the company providing the bandwidth has a financial stake in particular content, the temptation to manipulate the marketplace is irresistible.

A bandwidth marketplace using QoS and service level agreements for differentiation makes lots of sense, but only when the bandwidth providers have zero interest in content.

I watched C-Span last summer when the net-neutrality hearings were being held in Congress. Both sides of the issue made 'good' points.

However, just a little glimpse into the tactics the large corporations will be using to win the day:

Most of them argued very convincingly that they needed special access for which users could be charged because "No one is going to invest the huge amount of money necessary to move internet communications forward unless they have some way of recovering that investment."

Funny, Google doesn't appear to be doing too badly.

QoS is appropriate when used in a network controlled by one entity, like a corporation or a government agency. That entity can decide which traffic is the most important, and which traffic is less important. The interests of the entity are the only considerations. If the accounting application gets higher priority than the manufacturing application, and that's a problem, there are ways to sort it out.

QoS is also appropriate when used in a network composed of multiple entities with clearly defined communication agreements. For example, between businesses or universities, there may be private networks set up for specific applications. The use of a multi-entity private network is generally defined by a contract or other written agreement. The various entities have mutual interests in maintaining the health and functionality of the network. When additional applications are added, they may get a higher or lower priority on the network. The network-using entities can make decisions, resolve problems, and assign responsibilities based on their existing agreements.

QoS is not appropriate when multiple entities with no relationship (contract, service agreement, etc.) connect to the same network. The Internet is probably the only real example of this. I have a contractual relationship with my ISP. Google has a contractual relationship with several ISP's. But Google and I have no relationship. We both plug in to our ends of the Internet cloud, and we expect to be able to talk to each other. The beauty of the Internet, as originally conceived, is that anyone can connect to it and use it for just about any purpose.

Once you introduce QoS, the cloud isn't neutral any more, it becomes almost infinitely more complex. In a private network, this is a finite problem where there are a finite number of network endpoints with owners, and all the owners have a relationship and defined responsibilities. On the Internet there are millions (billions?) of end points, thousands of network providers, most with no relationship to each other. As davai put it, much more succinctly, those network providers with more power will be tempted to exercise that power, diminishing the value of the shared resource.