Jambe wrote:Could you elaborate on this?
I mean, what are your fears/hopes wrt the whole situation, what should ISPs, transit providers, the FCC, etc, ideally do?
Let's start with a practical suggestion: The FTC should disallow any Comcast/Time-warner merger. Period.
The FTC should not allow the giant ISPs to become anymore gigantic through acquisitions as opposed to attracting new customers. This isn't just ideal, it's something completely within the Government's power and the Government should absolutely prevent it without question.
Onto my ideal world: ISPs shouldn't be in the business, either provided for "free" (e.g. OpenConnect) or provided by fee (e.g. the recent Comcast-Netflix deal), of hosting content providers on their internal network for the sole purpose of reaching their internal customers. That situation is what I refer to as a ghetto. I'll explain more later.
I only favor the idea of Netflix paying for internal hosting as opposed to Netflix getting it for "free" via "OpenConnect" because:1) OpenConnect isn't Open.
It is simply a big player trying to get a special deal only for themselves solely by virtue of their size and clout. That's 100% the type of anti-competitive, 'Big Bad Business" behavior everyone loves to condemn. It's not magically different because we like Netflix and hate Comcast. It's a hideous misnomer of a clearly monopolistic action.2) Money is fungible.
If Netflix was hosted for free, that means that Comcast alone bears the cost. If Comcast isn't allowed to charged for "tiered service", that automatically
, and irrefutably
means that the cost will just be amortized across all of Comcast's current internet subscribers, regardless of whether or not those subscribers equally subscribe to Netflix. In other words, by saying that Comcast should host Netflix for free and that Comcast can't specifically charge customers for using it, you've just turned Netflix's costs into the generic cost of connecting to the internet for Comcast subscribers!
Which is unfair because it 1) means people who don't use it pay for it and 2) Netflix makes a larger profit because I doubt they're going to drop their fees to match. In effect, we're talking about a transfer payment between non-Netflix using Comcast Customers and Netflix shareholders, and people, ludicrously
, see this situation as PRO-CONSUMER and ANTI-MONOPOLIST! 3) "Free" doesn't exist.
TANSTAAFL. I'm sick of calls that regulators should enforce utterly absurd impossibilities, or that people should support the same. It's dangerous, misguided, and actually somewhat insane. I'm not trying to take this into R & P, but we already have IRS agents making people swear oaths on penalty of perjury that they didn't make common sense market decisions that completely fall within the purview of Judge Hand's famous dictum on tax avoidance. Leaving aside the infeasibility of any such certainty, or the inherent ambiguity, this facile approach to regulation should be "considered harmful."
But, ultimately, I am opposed to the idea paid or unpaid. I'm just saying that if it is going to happen, it'd better be paid. However, I don't think it should happen *AT ALL* because:4) It damages the "internet" and turns it into a bunch of ghettos.
The internet is literally defined as a "global system of interconnected computer networks." What Netflix is trying to do damages the "internet" because it removes much of the need for those interconnections in the first place. If all the big drivers of bandwidth cut special deals for intranet hosting, that means the "internet" doesn't grow, just the Verizon network, the Comcast Network, or the $BIG_ISP network. Instead of numerous different providers, commercial AND residential, constantly growing and increasing their connections to each other, the individual networks will just become beefier and jam-packed with internally-hosted content providers. The following is far-fetched and ridiculous, but it illustrates a real temptation: What if Netflix ceases to bother to have an internet presence at all, and is only available via the various ISP intranets? What if they then offer to help alleviate IPv4 Address Space Exhaustion by claiming some IP range (maybe even a currently non-routable one) and then just letting each individual network independently route to it? That's
the sort of "ghetto" I am talking about: where members of each network are essentially stuck within that network, with numerous restrictions on their ability to get out of it. It's a troublesome trend, rendered terrifying
because many of the people who think they are fighting against it are literally helping it come about!5) It is anti-competitive.
This is related to 4, but more specific. You see, Netflix is using Comcast (or vice versa) to leverage Comcast's captured residential customer base. This means that commercial ISPs, like Cogent, are at a competitive disadvantage because generally speaking commercial customers (unlike residential customers) have a lot of choice in how they get their internet. It isn't because Cogent is inherently inefficient, or any other market force, but rather because Cogent doesn't have any way to chain its customers to its services like Comcast and the other big ISPs. Netflix and Comcast shouldn't be able to effectively steal Cogent's business because they can take advantage of customers who don't have much of a market choice. That's not only bad form, but unhealthy for the internet at large. 5A?
The continuing bifurcation of "residential" versus "commercial" internet annoys me, because I like the idea that there shouldn't be any (well... port 25, etc... yeah ok) difference between the two. Indeed, I (and many others on this forum) regularly rely on that concept. The diminution of commercial-class internet towards ISPs doing it directly for their customers will likely increase the forces encouraging the emergence of internet-access "classes" in which residential customers have a diminished and more restricted "content-receiver" class of connection. This is a greater trend and more complicated, but what we are talking about here will definitely help tip the wrong side of the scale. No thanks, that would really suck for me.6) It removes an essential positive externality.
When everyone is increasingly building interconnects with everyone else, we all benefit from the enhanced connectivity. Netflix isn't doing it for us, but it's obviously a very big driver behind the the growth of those interconnects. Unlike a negative externality, where the people bear the ill-effects of business processes that have nothing to do with them, a positive externality is where people bear the good effects of business processes that have nothing to do with them. The growth of internet infrastructure is a boon for everyone, a net benefit for society. It's hard to appreciate, understand, or even scope how huge that effect is, but I think it's MASSIVE and that we probably won't properly appreciate it until it's gone (which hopefully won't ever happen). Just an example: What if, a decade or so ago, the predecessors of Netflix had done something similar in regards to the networks of their day? What if AOL's model of self-hosting had won out (remember keywords?) and all the big sites of the day the internet were individually hosted within each ISP network?
The internet, as we understand it, would likely not exist. Heck, Netflix probably wouldn't exist
(it started with mailed DVDs, but only and explicitly as a bootstrap to the "watch instantly" concept. Hence the name NETflix, which hasn't changed since its inception)!
Jambe wrote:Given what little I understand, this deal seems to paint dark skies over the transit business (and/or it incentivizes further consolidation of backbones and ISPs).
Yes, it does. But that's only more
true if Netflix managed to get the same deal for free (OpenConnect)!That's my frustration.
People think the problem is that Netflix had to pay! NO! The problem is the deal, paid or unpaid, which paid actually being slightly better