NNSquad - Network Neutrality Squad
NNSquad Home Page
[ NNSquad ] Re: [OIA] Fwd: [IP] The myth of title ii reclassification
I want to be clarify one point that Erik made. SS7 is not IP. What is true is that SS7 and IP have similar roles in being the basis for two approaches. They both use packets but very differently.
SS7 is the defining protocols for the “Intelligent” telecom network. It’s concerned with assuring that the appropriate resources are available and sets up a voice circuit. (There are also other services but I want to keep things simple). The is designed to carefully manage scarce recourses. If the resources are not available you get a circuit busy signal. This is QoS – quality or nothing.
IP is a very different concept. It is about best efforts packet delivery. No promises – if you want a voice call give it a try and if you’re lucky it works. We get confused because given sufficient quantity we have enough packets available so we can act as if it’s a guarantee. This is reasonable because the SS7 guarantee is hedged – there can still be failures and the voice path may have segments not under control as with analog and shared segments.
Most of the technical description is premised on the SS7 and that’s the world of the FCC regulations.
IP is very different. Concepts like out-of-bound might exist in some protocols but are not fundamental. Packets are just packets. Since the application and meanings are outside the network they don’t really fit into the telecom model. Hence all the machinations in trying to fit a protocol where packets have no intrinsic meaning and no guarantees into the telecom model in which all the meaning and roles are defined in the network.
There is severe cognitive dissonance. And plenty of billable hours for lawyers. I can’t comment on the title stuff. It has nothing to do with IP per se – it’s all about pretending nothing has changed when everything has.
There's a ton of network engineers in this business who play lawyer; it is as dangerous as lawyers playing network engineers. What I find, as a lawyer, is (a) generally lawyers will concede network engineering issues - that's the great thing about cross examining engineers, even trained AT&T "engineer" professional witnesses - they will, eventually, tell the truth (but it takes a long time); economists, generally, are not worth the trouble of cross examination; and (b) engineers, particularly in the policy wonk arena, tend to have extremely strong opinions of what the law means. They really have trouble grasping the fact that the law of the law is knowing that (a) everything is relative; and (b) the real law is knowing which details matter. It makes them bonkers to think that the law, particularly regulatory law, is stuff we make up as we go along. This can be frustrating for everyone - lawyers included - but you have to be Zen about it b/c regulators (and incumbents), and pretty much everyone at some level or another, make up law all the time. That this occurs is not a bad thing, but there is such a thing as balance. I wrote that post in order to demonstrate the scope and depth of present imbalances, and thus, the futility of incrementalism.
On Thu, Apr 15, 2010 at 12:22 PM, Alexander Goldman <email@example.com> wrote:
I called Ou a network engineer who has no understanding of RF on the wispa.org website. That's my opinion. Like myself, I think he's trying to get out of journalism. Unlike me, he does have a network engineering background -- but not a recent network engineering background.
On Thu, Apr 15, 2010 at 2:10 PM, Erik Cecil <firstname.lastname@example.org> wrote:
I think Ou completely misses the point. Here's my take: The Myth of the Myth of Title II Reclassification: http://www.erikcecil.com/2010/04/myth-of-myth-of-title-ii.html. My point is largely the same as Mark Cooper has made in other places: the framing of this entire debate is skewed. In fairness to Mark, I think my points go a little deeper and touch more broadly than where he thinks things should be.
Long story short, Bob Frankston is about 1/2 right. The framing is screwed. I pick up where all of you seem to run right into impasse: squaring the framing problems with the daily operations and money of these networks, which is where I earn and have earned my daily bread for the past 15 years.
Here's the text of the piece if you want to engage the details:
Well, finally, the telecom name game hits the big time. Having cross examined no end of Bell company witnesses, I'll examine George Ou's apparently pro-ILEC spin by juxtaposing his apparent "analysis" against the whole of the Internet. (I say "pro-ILEC" (aka pro-Bellco) because it appears Digital Society is a half-way house for Progress and Freedom Foundation (who routinely supports "deregulation" of the AT&T and Verizon variety) judging by their personnel: http://www.digitalsociety.org/about/)
Ou takes issue with the proposition the FCC should regulate the transport components of the Internet in a short article entitled "The Myth of Title II `reclassification'". He refers to some FCC orders without naming them and then sets up and proves a straw man argument. Setting aside a certain lack of legal depth, his analysis also fails to account for industry history, evolution of the networks, interconnection arrangements (uh, across every platform - cable, wireless, wireline, satellite, and so on), tariffs, facilities, towers, or money behind them, much less the practicalities of actually running these business or litigating the cases he seems to cite with such confidence. Rather than dive into briefing this - b/c thorough treatment would require briefing, why don't we just examine the facts. Secondly, he was a network engineer, now reporter, not a lawyer. So to keep this fair, I'll take off my lawyer hat - and go right into the deep end of the network stuff to illustrate how, on the facts alone, this kind of reasoning is ludicrous.
First, let's not get lost in the name game:
If we run it to the lowest common denominator – the signals travelling across coaxial cable, hybrid coax, copper, fiber optic, TDM, mobile wireless, fixed wireless, handsets of every stripe, CDMA, GSM, LTE, 4G, satellite, class V switches, class VI switches, cable head ends, CDNs, etc. – we know that electronic communications networks – however labeled (Internet, Public Switched Network, cable, broadcast, etc.) involve people using computers to communicate with each other. Computers process "information". Information, to a computer, exists in one of two states: on or off. Your 2 gigahertz computer processor, for example, is a switch that essentially turns on or off 2 billion times per second. So it all boils down to 010010001010100101010101010s (and programming and money, but hey, let's see what's really going on underneath the name game).
Setting aside gaping holes in the applicable legal history, which, again, we don't need to address here as the facts, I think, are delicious, let's just see how that plays out against all of the "transport" and "information" distinctions the article asserts are so utterly clear.
We're not done yet; it gets better:
Secondly, let's just take one real world example:
Somehow the incumbents, and apparently Mr. Ou say Title II never applied to the "Internet". Well maybe ... they are hyper-technically correct. But, again, to avoid confusion, recall that certain special rules apply to certain 1010101s that carry voice information. Sometimes people use the Internet to communicate by voice. Sometimes they talk to other people who use computers to connect to the Internet. Sometimes, however, their friends are away from their computers. So they call the telephones. In that case 1010010101s go from a computer to other computers that "convert" the 10101010s from IP to other formats - whether TDM, CDMA, GSM, etc. In some cases the 101010101s from your computer will have other numbers associated with them - a telephone number - let's say you buy Vonage in Washington, DC (202 area code). You move to San Francisco (415 area code). You take your vonage phone with you. You call your grandmother, across the street. Your grandmother buys old fashioned telephone service from AT&T. You buy DSL service from AT&T. AT&T will charge the carrier who takes your 101010101s from the "Internet" (IP format) and converts those to 101010101s that can talk to it's old fashioned telephone switches (TDM format - this is old copper landline) "long distance" fees. But you can only charge "long distance" if you apply a tariff (tariffs are creatures of Title II) to that call.
The money in these charges is several tens of billions of dollars per year. AT&T, Qwest, Verizon, Windstream, Embarq now CenturyLink, etc. has sued all kinds of carriers who convert 1010101s from IP format to 01010101s in TDM format to connect with their old fashioned switches and won hundreds of millions of dollars. At the same time these carriers use IP in their backbones, deploy all sorts of IP-enabled gear, but use their old TDM operations as cash registers to clean out everyone but themselves. Go figure.
So of course Title II has never been applied to the Internet. "Reclassification is a myth." OK. Sure. Clear as day.
Fortunately for us, AT&T, Verizon, Qwest, Windstream, CenturyLink, and apparently the good people at Digital Society, want to keep the system just the way it is right now. Mmmmm. Glad the "free market" guys are completely in alignment with all of the "free market" incumbents who want to make sure that selective deregulation stays in place.
On Wed, Apr 14, 2010 at 7:40 PM, Joly MacFie <email@example.com> wrote: