Mark Cuban made a blog post in which he confessed that
telcos charging different fees for various levels of QoS (that "haves and have-nots" as Jeff Pulver put
it) is not only acceptable but probably a good idea.
There are all kinds of problems with the logical
outcomes of Cuban's reasoning. But there is also a glimmer of genious in there. Pulver has addressed the crap side of Cuban's idea. Let me
now address the issue that may be underlying Cuban's sudden support for this idea (to his credit, Jeff Pulver did touch
on some of this, too--but man was he fired up, it almost read like he felt philosophically betrayed by Cuban).
The whole notion of charging more for higher QoS pathways is to prevent abuse of the QoS protocols that make entrance
into the fast lane possible. In other words, the carriers feel as though they would HAVE TO charge more for better QoS
to keep out the traffic that claims to be high priority but really isn't. The idea of universal, equal QoS
enforcement on the Net it Utopian, but it just isn't possible. People would abuse it just as they abuse e-mail
today. For that reason, the almighty buck steps in to make it costly for people to get into the fast lane. Once
again, the existence of the indecent penalizes the fortunes of the decent.
But what Cuban DOESN'T recognize
is that we already have two tiers of service for data connectivity on the global matrix. As the first tier, we have the
Internet, which is the "slow lane": commodity bandwidth at a reasonable price with best-effort
delivery. The second tier of service is already here, and it already has built-in QoS at a premium price.
That second tier is the PSTN, folks. Think about it.
People do pay a premium for TDM services on the
PSTN, which are dedicated, circuit-switched, guaranteed bandwidth. Inherently QoS-enabled. The FAST LANE.
This second tier already comes at a higher price. Moving such guarantees to IP is of great benefit to those who
want to totally replace the PSTN, of course--myself included. I don't want Bell to control the global transport
layer and data link layer.
We don't want to reinvent the PSTN on the Internet--but we DO need to find
a way to maintain Net neutrality while encouraging global QoS measures on the Net. And yes, the Internet NEEDS QoS
enforcement. But how can it be done without abuse of protocols (my fear and presumably Cuban's fear), and without
abuse of consumers (the net result of Cuban's proposed solution)?
For what it's worth, Jeff and I don't
appear to be the only ones who thinks Cuban is barking up the wrong tree, even if I do
understand WHY Cuban is barking. (Incidentally, Cuban runs a Net-video firm that would benefit greatly from Internet
QoS.)
