One of the central dogmas of the Church of Extreme Net Neutrality (CoENN) is that quality of service on the Internet, or using the preferred nomenclature of the CoENN, “paid prioritization,” is the equivalent of a deadly sin.
The CoENN creed against quality of service states that paid prioritization of Internet traffic: (1) has never been contemplated by standards organizations like the Internet Engineering Task Force (IETF); (2) does not exist on the Internet today and, to the extent it exists anywhere, is probably being used nefariously by the pagans; and (3) if it did exist on the Internet, it would be available to and affordable for only a small number of deep-pocketed hegemons.
These iniquities of paid prioritization are spelled out in a recent filing at the FCC in which Free Press preaches the old time religion of the dumb network. But, like so many dogmas, this one turns out to be, well, not exactly true.
Which leads me to the letter we filed yesterday in the FCC’s Open Internet proceeding to correct the record with respect to paid prioritization. In a nutshell, we point out that, contrary to the CoENN’s claims: (1) the IETF documents clearly contemplate and permit differentiated pricing of Internet traffic based on the use of prioritization; (2) paid prioritization of Internet traffic is widely available to businesses today; and (3) such prioritization is often voluntarily purchased by small and medium-sized enterprises, including minority-owned businesses and community organizations.
I think our letter speaks for itself, and provides the details and facts on these three points in a succinct five pages, but I would like to elaborate briefly on a couple of points. First, the elders of the CoENN seem to have adopted a somewhat self-contradictory creed. At various times, Free Press has expressed the view that any router-based prioritization of Internet traffic is by definition harmful to unprioritized packets. Yet now they seem to have endorsed the use of DiffServ, which is a mechanism for router-based prioritization, as long as it is in the control of “end users.”
Free Press apparently does not recognize that content and application providers may also be “end users” of Internet access services. Indeed, to the extent that packets must be marked for prioritization at their origin, content providers may be the “end users” best placed to make use of DiffServ.
One sometimes hears from members of the congregation of the CoENN that the introduction of paid prioritization would enable ISPs to turn best effort Internet transmission into a “dirt road” and force virtually the entire Internet ecosystem to “pay extra” for prioritized transmission. (Query: why would ISPs require such an elaborate scheme to raise rates if they have the market power attributed to them by the CoENN?) Yet now Free Press seems to suggest that ISPs would restrict prioritization to only a few “deep-pocketed Internet giants.” While I enjoy the Da Vinci Code conspiracy theories as much as the next blogger, I do expect at least some superficial consistency.
Finally, Free Press has tried heroically to distinguish “router-based” prioritization from “geographic” prioritization (i.e., CDNs), based on unsubstantiated allegations that geographic prioritization does not result in any displacement of unprioritized traffic. For whatever reason, they apparently never subjected this particular doctrine of the faith to any empirical testing. Well, someone else did and found that geographic prioritization can be quite effective in pushing unprioritized packets to the proverbial slow lane.
The FCC must view with healthy skepticism the interpretations and opinions it receives on technical Internet standards, and how they are operationalized by ISPs, from an advocacy group with no demonstrable expertise or experience in such matters. When it comes to data-driven policy making, there is no room for faith-based initiatives, as Chairman Genachowski testified before Congress last year.