Monday, December 15, 2008
Bogus WSJ Story on Net Neutrality
My bullshit detector was triggered by paragraph five, which reads
One major cable operator in talks with Google says it has been reluctant so far to strike a deal because of concern it might violate Federal Communications Commission guidelines on network neutrality. "If we did this, Washington would be on fire," says one executive at the cable company who is familiar with the talks . . .Yeah, right, the cable guys want to preserve Network Neutrality, while Google wants to violate it. That **would** be a boy-bites-dog story, if it were true.
Google's FCC counsel, Rick Whitt explains that Google simply seeks to do edge caching, just like Akamai, Amazon, and several other companies. The idea of edge caching is to locate frequently used content closer to the people who access it. It makes accessing the cached content faster.
Importantly, since the cache must be connected to the Internet by a big, fat, stupid pipe, the company doing the caching can, in principle, buy pipes from any carrier. Indeed, if it is concerned with up-time, local congestion, or avoiding single points of failure, it will buy connections from several providers. [See my "Buy as Many Nines as you Need"]
Also, in principle, carriers can let any edge cache access its network. Indeed they have a duty to do so, under the Doctrine of Public Callings, the doctrine of common law that underlies common carriage and network neutrality that has been in effect for about 900 years now.
Net Neutrality only becomes an issue when a carrier picks and chooses which cache to supply pipes to.
The argument the WSJ seems to be making -- and they don't make it very well -- is that when Google has an arrangement with carriers to provide a cache it advantages its access. However, it has always been the case that Google pay a carrier more for a fatter pipe to its content. Edge caching is another case of that, no matter in which building a caching platform might be located.
In other words, if Google does edge caching it buys access. It's the same as when I, as a residential customer, pay $34.95 for one megabit DSL service or $49.95 for 3 megabit DSL.
The concern of Network Neutrality advocates is not with access but with delivery. The fear is that Internet connection providers would charge for expedited delivery of certain content to the end user, and in so doing would put themselves in the business of classifying which content gets enhanced delivery. Since they were charging for expedited delivery, they'd get more revenue for improving the enhanced delivery, so the only network upgrades would be for the enhanced service. Non-enhanced would fall further and further behind. Plus the power to decide what gets delivered might, indeed, be powerful, and power corrupts; just ask NARAL.
Since the edge caching Google is proposing is about access, not delivery, there's no problem.
A second concern of Network Neutrality advocates is that deals for "enhanced delivery" could be used to advantage some application providers and disadvantage others. So, for example, [hypothetically, but see this, this, this, this and this] Internet telephony might work better with Comcast telephony than with Vonage if you're a Comcast customer -- but this would have NOTHING to do with the Internet access that Vonage purchased from its suppliers.
And it has NOTHING to do with Google's edge caching proposal.
A third concern of Network Neutrality advocates has to do with barriers to entry and innovation. The concern is that the Internet has two very innovation-friendly properties, Transmission-without-permission and Rapid-market-discovery that foster an unprecedented rate of innovation. If barriers to either of these appear, the Internet becomes less fertile. Suppose for a moment I had to pay each ISP that had a reader of this blog as its customer, and every month I got hundreds of bills from around the world. I might decide to limit the distribution of my blog. Or if I didn't pay some of the ISPs, I'd lose the readers they connect. In any case, my ability to discover my market would be impaired.
Again, this has to do with delivery. It has NOTHING to do with Google's edge-caching proposal.
There's a fourth concern. It has to do with the amount of Internet capacity that some apps -- mostly video -- use. Internet providers on both the access and the delivery side are concerned, and they assert the right to "manage" their network. The concern over management is inversely proportional to the capacity of their network. So, on one side we have Verizon, with its new FIOS fober-to-the-home network, saying it doesn't need to cap or manage its network to deal with video. On the other side, wireless carriers often institute severe limits on monthly capacity AND on which applications you can use. In the middle, you have telephone companies that provide DSL but not fiber, and cablecos that only allocate a small proportion of the capacity of their coax to the Internet -- here you get the most intense concern for the "coming Exaflood" of bandwidth, a bogus claim that, they say, can only be solved if there's more investment than these individual companies can afford. And, of course, the screaming about Internet video is loudest from the companies most vested in the old video entertainment model.
I have some concern here that since the caching servers that Google proposes to install, and their associated access pipes, are expensive, and they do provide an advantage to Google-as-Video-Provider, that it may make it more difficult for other video platforms to get a start. But I must say it is the same concern that I have for new search providers and new on-line advertisers. Once you get big like Google is big, you're a barrier.
But this has very little to do with the Internet, and nothing to do with Network Neutrality.
Also, the concern over the specialness of video is severely misplaced, because the cost of high-speed connections is falling. Remember when Internet connections were so slow it was hard to make a VOIP call? Not anymore (except for special cases such as outlined above). Soon the Internet will be able to do 6.5 billion channels, so everybody on Earth can watch what he or she wants whenever he or she wants, affordably and without delay. The technology already exists, and has for decades. It is now rolling out . . . slowly. The gating factor is not availability or affordability, it is the extent to which incumbents need to protect existing revenue streams and the depreciation schedules on existing equipment.
Yes, even mobile networks could soon provide gigabits per user, just as soon as we (a) re-think spectrum regulation in terms a bit more modern than 1927 technology and (b) put fiber everywhere so radios can whisper rather than shout.
So, even as we acknowledge the carriers' need to manage their networks, let's be mindful of whether they're managing for scarcity or abundance.
Finally, it is funny that two of the WSJ story's five named sources, Rick Whitt and Larry Lessig, either don't remember saying what the Journal quotes them saying or feel their words were twisted and taken out of context. How can this be good journalism? This is not the first time I've been in an expert position to see the WSJ engage in gotcha-reporting [link].
Disclosure: Google has sponsored several isen.com events, such as F2C: Freedom to Connect, in the past. Currently Google is not a sponsor of the next F2C, which will be held in Washington DC on March 30 and 31, 2009, but we're in discussions about it. My sponsorship relationship with Google has no causative relationship to anything I've said above; rather Google has been an isen.com sponsor because WE BOTH SHARE pre-existing positions on issues like Network Neutrality. Verizon, mentioned above, sponsored F2C 2008, but is not likely to sponsor F2C 2009; again, they sponsored F2C because it was already in synch with isen.com on issues of global climate change and network capacity.
Further disclosure: An isen.com sponsor might disagree with isen.com, and vice versa, but I'll still accept their sponsorship if there's no quid pro quo, implicit or explicit, about my positions on any issues.
Even further disclosure: Chris Rhoads, one of the authors of the WSJ story, once took me out to lunch, and I've had several cordial phone calls with him since then, but I think he did a crappy job on this story. Chris, you blew it. Completely.
More disclosures? I can't think of any, and I'm trying, but it's not because I'm hiding anything. [Cause for this nervousness here.]
Technorati Tags: AndrewOdlyzko, Broadband, Cableco, Cellco, Censorship, Competition, End-to-End, F2C, F2C2009, FIOS, FON, Framing, FreePress, FTTH, Internet, isen.blog, LawrenceLessig, NationalBroadbandStrategy, NetworkNeutrality, Organizational Culture, Press, Stupid Network, Telco, Verizon, VOIP, WirelessNetworks, WSJ
So far I understand this is pretty much what Google is doing: they pay for a special service which smaller competitors wouldn´t get. You may say it is not a matter of net neutrality because the "net" is not affected, but it is a barrier to market access.
The fact is that Google has carved out special exemptions in its peculiar and unique definition of "network neutrality" to suit its selfish corporate interests. You should not be surprised by this, much less be in denial about it. This is what corporations do in the brand of capitalism currently practiced in our country. They are not legally allowed to have morals or scruples or even make sense; all that matters is how many pennies they can say that they're earning for shareholders (but in fact deliver to overpaid execs). So, it's no wonder that while they send millions to DC lobbying groups (such as Public Knowledge) who cry that "little guys will be stuck in the slow lane on the Internet if we don't regulate!", the truth is that they want to stick their competitors in the slow lane.
It's ironic that even though you have adopted "freedom to connect" as a mantra, you seem to forget that ISPs *are* the Internet. Google is just one of the hucksters that hawks wares on it, though it happens to be the biggest and most privacy-invading. (Google, now that it has acquired DoubleClick, is by far the largest source of spyware tracking cookies on the Net.) I, as an ISP, am interested in getting my users the best possible service and access at the most affordable price. Is Google? Nope; its customers are its advertisers, not Internet users. So, whose interests are more likely to be aligned with those of the end user?
Have you hugged your packet today?
Caching speeds up delivery, so you're making a distinction without a difference.
"Suppose for a moment I had to pay each ISP that had a reader of this blog as its customer, and every month I got hundreds of bills from around the world."
In other words, Richard, there is plenty of significant, material difference between access and delivey.
By locating servers inside the ISP's CO, Google is paying for QoS, which has been forbidden in nearly every Net Neutrality bill. They aren't paying retail for QoS, they're paying a toll that permits ISP retail customers to access Google content faster, smoother, and more reliably than they can access Netflix.
This is sound engineering, but it contradicts the rules proposed by Markey, Snowe-Dorgan, and Wyden. As I opposed all those bills myself, I don't have a problem with what Google is doing. But I do have a problem with companies and other folks saying one thing and doing another.
So yeah, this is a big deal.
It is unfortunate that the quality of the Journal's reporting has fallen so precipitously as is evidenced by this story.
Of course, that was the objection many had before the sale.
Think about it. I’ll bet that Google will be able to place an edge cache at the site of any ISP it wants for free. Why? Because YouTube and its related services consume SO much bandwidth that the ISP would be crazy to say no. The ISP would surely save big on its backbone connection. Terabytes per month on YouTube alone. That's money in the bank right there. And service would be faster, too.
But would the ISP allow just any content provider to put a cache at its sites, for free or even for money? Doubtful. Caches take up space and power and require access for maintenance. If a small Internet startup were to call your local cable company and ask for "co-location space," the person there would probably say, "That’s not a product we sell to the public." That is, if the person who answered the phone at the cable company even knew what it was. Or was even in the US. (Most of the time, when you call a cable company, you get a telemarketing operation in some foreign country. Good luck getting through to the actual office in the States.)
And of course, would-be competitors of Google won’t be big enough to offer the same bandwidth savings to the ISP and won't be able to buy space on Google’s private edge caches.
So, in what way is this neutral? Google can get its servers into places where CoolNewInternetGarageStartup.com can’t. Therefore, Google is indeed getting preferential access to infrastructure; it’s just co-location space instead of pipes. What's more, it has a big advantage there, because co-location is much more difficult to obtain than bandwidth. You can get any Internet carrier to sell you a pipe. But co-location space at ISPs, which is more cost-effective than buying pipes, isn’t necessarily even available to you unless you’re Google. So this is really, really non-neutral and anticompetitive.