When Comcast first unveiled its plans to not count XfinityTV traffic towards bandwidth caps, it said that this was permissible since it is a service separate from its high speed access, with the Xbox acting as another set-top box and with traffic travelling over its own separate IP network. But the closer the scrutiny gets, the less separate it all looks.
Indeed, several sources out there are directly challenging this now, suggesting that all Comcast has really implemented is QoS. In recent articles, both Dan Rayburn and Brian Berg have taken a detailed look at the actual data being transferred to find out Comcast’s implementation. Bits from Xfinity and bits from Netflix, MLB, or other over-the-top sources apparently differ only in their QoS tag. The former are high priority CS5, and the latter are low priority CS1. Others from deeper in the industry have told me similar details offline.
In other words, Comcast’s service is separate from its Internet Access service only in a virtual sense. There is no separate, parallel pipe when it comes to the last mile, the bits are mingled – just tagged differently. Comcast may count it differently, but it comes from the same pool of bandwidth. Maybe further upstream there is actual separation, but this is true for all traffic at some point – that’s the content provider side of things. Nobody cares about traffic caps anywhere except where the bottleneck is, and that’s where the traffic here is mingled and separated only by a QoS tag.
This doesn’t mean that Comcast is slowing anyone down or delivering traffic differently – you certainly *can* do that with QoS tags but from what I hear there is no evidence of that. All they seem to be doing with the QoS tags is using them to counting the bits differently, as in they’re not counting the XfinityTV ones against the cap.
Yet that’s precisely what the NBCU consent decree says they can’t do, so either they’re violating the order or they’re rubbing the DOJ’s nose in a loophole they had ready before the ink was dry.
As GigaOm’s Stacey Higginbotham characterized a letter from Senator Franken, the DOJ and FCC seem to be letting Comcast walk all over them. But whether the lawyers have threaded a way through the needle of the letter of the law is irrelevant in the grand scheme of things. If it is possible to simply slap a QoS tag on your own traffic and call it a new service separate from the Internet Access service, then the entire concept of net neutrality is already dead and we should just bury now it before it starts to attract flies.
Some would argue that’s how it should be. In fact, I myself am not sold on pure net neutrality, and have wondered openly if the 800-number model that AT&T and Verizon are said to be readying might offer something to the discussion if implemented in an open and accessible manner. Net neutrality as demanded by the most strident proponents is a political solution to an economic problem (that’s not a good thing).
But let’s at least not play pretend. Net neutrality with arbitrary, virtual self-exceptions is just a bit Orwellian. (All bits are treated neutrally, just some more neutrally than others…)
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!
Categories: Cable · Government Regulations · Internet Traffic
Join Our LinkedIn Group
I’m struggling to see a workable business model for SPs (wireless, MSOs, ILECs, etc.) going forward. The concept of an Open Internet hinges on the erroneous concept that broadband networks and the “open and free” Internet they carry are cash cows. They are not. Especially for those carriers who don’t happen to own HFC or FTTx plants and/or are competitive with those who do.
No one, to my knowledge, is making money on $30 for 30 megs. They are going to make that money elsewhere. Be it on providing their own content, aggregating others, or differentiating their networks through SLAs, QoS, or any other method possible.
The carriers are becoming the Harrison Bergeron of the new economy with the FCC and consumer advocacy groups playing the Handicapper General. If they compete too well, they get laden with regulations and rulings and orders. If they don’t compete well enough, they are squandering the networks tax-payers supposedly built.
Here’s the cold hard reality. Carriers in the US built and sold data networks as loss-leaders. The model was built on their subscribers purchasing more than they needed, using less than they had, and having all of the other services a carrier would offer (such as voice and video) to pay for it. It turns out that model was flawed. Now they are trying to correct the model in an environment where their their core products are commodities and suffer from competition that is unburdened by regulation and the realities of things like amortization and investor value.
Here’s a thought; let the consumer decide if they don’t want caps and pay-to-play internet. Anyone see VZW going bankrupt due to all their customers fleeing to Sprint’s “truly unlimited” network? Anyone see Apple’s walled-garden model failing under the burden of curated content and pay-to-play App Store?
I know its easy to see SPs as money-grubbing and anti-competitive. But the FCC and, frankly, consumers, need to see them for what they are. Businesses trying to make money on the only real and relevant asset they have.
Why not simply stop selling bandwidth as a loss leader, and change pricing to match the real costs? Bill for the pipe, and stop overcharging for voice and messaging bits etc… Then who cares about wireline voice losses.
So raise prices while competitors (like Comcast and AT&T) don’t? AT&T spent $7B on U-Verse. I’m thinking they’d be willing to price below smaller competitors for a little while to win the business. Would you pay $30 for 12M when you can get 30M for the same price ? At least, until they drive the smaller ones out of business and then raise rates.
Innovation and competition will have to happen outside of “speeds and feeds”.
Last time I looked Comcast still owned the Comcast network. They can charge per day, per week, per month and require dollars or rocks of gold. If they want to match market prices for one service (e.g., voice) but promote another (ISP) who cares? Should we force McDonald’s to charge for ketchup? Can we force them to let you eat Wendy’s at a McDonald’s because of some vague notion of hamburger-neutrality?
I am all in favor of hamburger neutrality.
“The former are high priority CS5, and the latter are low priority CS1”
The DSCP markings do not mean that one has priority over the other. It just means they are tagged with DSCP which can be used for priority AND routing, OR accounting, OR security, or any combination.