Join 3,512 readers in helping fund MetaFilter (Hide)


The Year of Net Neutrality
January 16, 2008 5:29 PM   Subscribe

We're only two weeks into the year, but net neutrality issues hit the ground running. The FCC already has three different inquiries open. (also) (previously) The 700 Mhz auction threatens to disrupt an already converging telecom industry. AT&T's post-merger commitment to net neutrality ends this year, and they plan to test the filtering waters, despite recently opposing the practice. And today, a leaked memo revealed that Time Warner will test tiered internet services soon. The Internet as we know it, and communications in general, might be headed for some major changes in 2008.
posted by spiderwire (20 comments total) 4 users marked this as a favorite

 
I wouldn't throw TW's tiered pricing in the same basket as AT&T's filtering. TW's tiers, at least as I understand them, may really not be a bad thing. I personally think that tiered bandwidth plans would be better than the "all you can eat (but not really)" plans that we have now, where you're implicitly promised something that the network isn't capable of delivering and they'll disconnect you if you try to use.
posted by Kadin2048 at 6:22 PM on January 16, 2008 [1 favorite]


MHz.

/pedant
posted by ZenMasterThis at 7:12 PM on January 16, 2008


So will the adverts count against our limits?
posted by JakeEXTREME at 7:13 PM on January 16, 2008 [1 favorite]


2008 ... The Last Year of Net Neutrality?
posted by wendell at 7:16 PM on January 16, 2008


Um, bandwidth caps != network non-neutrality. In fact, it is net neutrality. You get charged based on bandwidth, not based on content type. That means, if you take up more then your "fair share" of bandwidth doing p2p, then you have to pay more.

Compare that to comcast, which filters and degrades BitTorrent traffic.

The only problem is that a lack of competition means TW can really gouge people who go for higher bandwidth packages.
posted by delmoi at 7:31 PM on January 16, 2008 [1 favorite]


Imminent death of the Net predicted. Film at 11, but only if you've signed up for our Real-Time ActionWeb Video-Streaming Live News Feed.
posted by sfenders at 7:31 PM on January 16, 2008 [1 favorite]


I think I'd be fine with tiered pricing if the tiers are set reasonably.

Of course the problem is what is reasonable? Some people burn bandwidth with bittorrent, but lots of others burn it by telecommuting, running citrix-hosted apps over VPNs, participating in online conference calls with video, shared screens, or shared whiteboards, etc.

Or even if they're not working, more and more games are downloaded and updated over the net, as are more and more legitimate sources of movies and music.
posted by Tacos Are Pretty Great at 8:08 PM on January 16, 2008


Maybe someone who works for a telco or backbone provider can answer this: Is bandwidth at the ISP-to-ISP level getting cheaper over time? Is there a "Moore's Law" of bandwidth?
posted by arto at 8:51 PM on January 16, 2008


Ridiculous. If TWC goes to tiered I can just go to the AT&T, Comcast, FiOS or the other couple of choices I have in the area. I'd gladly pay more for those really high speed networks, like FiOS, which is suppose to be completed relatively soon. Charge more for speed, not for amount used.
posted by geoff. at 8:54 PM on January 16, 2008


Or even if they're not working, more and more games are downloaded and updated over the net, as are more and more legitimate sources of movies and music.

Right, it seems anti-competitive to me. I have Vudu which sends me HD movies over the Internet. I pay for these, but it will cost me even more via the bandwidth cap tax? But I don't get penalized for purchasing the same movie over the TWC HD on Demand service?
posted by geoff. at 8:56 PM on January 16, 2008


Where are you, Geoff? Because here, I haven't found a good alternative to TW.

The real problem has always seemed to me to be municipal contracts, which can be good or bad, but discourage real competition among providers.
posted by klangklangston at 9:17 PM on January 16, 2008


Maybe someone who works for a telco or backbone provider can answer this: Is bandwidth at the ISP-to-ISP level getting cheaper over time? Is there a "Moore's Law" of bandwidth?

If you mean backhaul and middle-mile lines, there seems to be some consensus that by the end of last year we'd essentially eaten up most of the bandwidth glut that was built up in the late 90s, and the major Tier 1 ISPs have been laying more fiber recently. Verizon, especially, stands out for having made a huge investment in expanding their network, to the point where they apparently neglect most of their copper loops almost completely now (even in terms of support) and are going whole-hog with their FiOS offerings.

I'm not sure that the "Moore's Law" analogy applies to major pipes, since it's a fixed infrastructure cost (and some of it can be upgraded at the ends by better fiber multiplexing, etc.) -- there is a need to increase capacity, but my feeling is that it doesn't scale nearly as fast as processor power.

Ridiculous. If TWC goes to tiered I can just go to the AT&T, Comcast, FiOS or the other couple of choices I have in the area. I'd gladly pay more for those really high speed networks, like FiOS, which is suppose to be completed relatively soon. Charge more for speed, not for amount used.

Caveat emptor on this one. I hear a lot about FiOS having serious problems in many areas, and it is, after all, Verizon/UUNet/The Borg. These are the guys that just raised FiOS rates from $99 to $109 for a bunch of customers and called it a "decreased discount amount." Double-plus what?
posted by spiderwire at 9:52 PM on January 16, 2008


I predict this will work out as well for the incumbent telcos as media consolidation and price-fixing has worked out for the RIAA. I hope they do it, it's just the shot in the arm municipal broadband and community wireless initiatives need to get some real public support behind them.

Yes, please, by all means do it.
posted by George_Spiggott at 9:55 PM on January 16, 2008


The United States is already horribly backwards when it comes to bandwidth and service. I managed to find 12mbps downstream for about $55/month, which was quite amazing, but the upstream was 1.5mbps and that's still way beyond any other residential service I've seen here. If "tiered service" means offering symmetrical high speeds for a premium price, sure! However, I'm afraid we're looking at any decent service being priced higher, "gaming" and "business" packages marked at a serious premium, and basic speeds effectively being throttled to the cheapest of basic DSL.

I just have to wonder how this will play out internationally, in regards to traffic that involves the US.
posted by Saydur at 9:58 PM on January 16, 2008


Um, bandwidth caps != network non-neutrality. In fact, it is net neutrality. You get charged based on bandwidth, not based on content type. That means, if you take up more then your "fair share" of bandwidth doing p2p, then you have to pay more.

That's an oversimplification. You're right that "net neutrality" is often in the eye of the beholder, but overage charges are not the same as per-byte charges, and they're a big change from a per-pipe charge, which it's looking like we've lost already -- it's a fait accompli. This is worrisome for a whole bunch of reasons, but I included it because it's part of a pattern, and just one of the different flavors of traffic-shaping and cost-reallocation that's being floated right now. More combinations are in the pipe -- more on this below.

Compare that to comcast, which filters and degrades BitTorrent traffic.

Well, that's an interesting statement. I think you're right in that the big division today that the net neutrality debate isn't really addressing is the difference between metering and filtering. Both are, in a sense, non-net-neutral in that they're not the content-agnostic, per-pipe setup we're used to.

Metering (per-byte) charges certainly aren't neutral if you're a company that's dependent on high-bandwidth content delivery or you compete on style, and without prioritization, it doesn't really give you an increase in QoS. It also clearly tips the playing field against distributors who distribute through peer-to-peer networks, like, e.g., the way Blizzard does with many of their big productions over their custom BT client. Skype uses a constant low level of bandwidth for load distribution, so this potentially hurts them bigtime (and the fact that a major threat to the voice telcos is getting the shaft like that is no accident).

It also disproportionately hurts consumers who use a wider percentage of their pipes, and those aren't all BT users and pirates -- lots of legitimate users just have high-bandwidth needs. Another way of framing the per-byte/overage system is as a way for ISPs to shift the costs of bandwidth allocation to their dependent users -- rather than having to keep a full pipe open for all customers at all times, they're essentially penalizing you for using your own pipe more efficiently. They don't want you to use your full upstream/downstream all the time, even though you pay for it. They want to let the casual user have burst capability, then reclaim all that idle space and use it instead of paying for infrastructure upgrades, and offset the costs onto the minority of users who aren't actually trying to get what they pay for.

Now, admittedly that's not nearly as problematic as filtering in many ways, but there are two sides to that debate as well that aren't entirely clear. On one hand, there are players like AT&T and especially Verizon who've spent billions upgrading their networks to provide improved QoS and who are now pissed the Comcast and Cox are "faking" improved QoS without paying the infrastructure price by doing traffic shaping -- and I'm sure they're not happy about TW's move either. Verizon's put a lot of capital into building a network that can actually carry the data that TW and Comcast are trying to suppress on the sly, and they're not happy about it.

If you're a content provider, the setup you prefer probably depends on what sort of content you provide, but there are a couple of bottom-line facts that apply no matter what.

First, any of these changes is almost certainly only the first step in transit providers taking control over the networks, which is letting the genie out of the bottle. If they have the right to shape traffic, it'd be foolish to think they'll stop with tiering. Their goal is reclaiming the gateway to the network and locking down interconnection, just like Bell did with the phone system. It puts an industry that's built around burstable pipes under the thumb of the big providers.

Second, both filtering ("fake" QoS like Comcast does) and prioritization ("real" QoS like Verizon wants) are problematic, though again depending on what kind of content provider you are. Both probably entail deep-packet inspection, which should make consumers uncomfortable for a number of reasons. Both tend to privilege established players who can pay for big pipes, which is bad news for anything peer-to-peer -- it'll tend to ossify the network by enabling big-money players to pay to drown out the upstarts (e.g. the big telcos can pay for their VoIP, but tiering hurts Skype, which piggybacks on its users idle bandwidth; again, that's not an accident -- Verizon and AT&T very much want to isolate and suppress VoIP as an anticompetitive means of preserving their legacy voice business).

Frankly, filtering is probably preferable to players like Google, because it tends to preserve the current pay-per-pipe ecosystem. A content-prioritization network is seriously dangerous, and the overage setup is just the first step. The sort of pay-for-service system that Verizon is dreaming of means that (a) established players would have enormous power to suppress new entrants to their markets, simply by paying for better connectivity, and (b) all content providers will be at the mercy of a few big ISPs (and there aren't many nowadays) -- providers will essentially be forced to pay protection money to prevent their traffic from being deprioritized below their competitors: e.g, Google will have to pay a premium to ensure that their search results stay as fast as MSN's.

This isn't a "necessary" outcome -- what we're talking about here is nothing but wealth redistribution; Verizon, et al. just want to skim off Google's profits without actually contributing anything that creates value. They're building the networks already; what they're trying to do is to blackmail major content providers by forcing them to pay premiums in order to not receive a relative downgrade in service compared to their competition.

delmoi is right to a degree, but the distinction isn't as clear-cut as per-byte versus by-content, and that's the thing we have to start wrapping our heads around. The entire telecom market is heading for some major upheavals, and as soon as we start talking about "fair shares," we should start asking: well, who gets to say what a "fair share" is? Bell pulled this same trick for decades: they artificially suppressed service upgrades to prevent innovative uses of their network, gold-plated the network in other places to justify price-supports, cross-subsidized their competitive services with windfalls from their monopoly services, and played havoc with standards and interconnection, and no one was able to touch them for years as a result. When big telcos start talking about "fair shares," remember that these are the guys that sued to prevent the first modems on their network, on the pretext that it would fry the phone lines, while conveniently preparing their own, identical (well, technologically inferior) devices that had just been a bit late getting to market.

They've been playing this game for about a century now, and they're very, very good at it.
posted by spiderwire at 10:48 PM on January 16, 2008 [2 favorites]


it's just the shot in the arm municipal broadband and community wireless initiatives need to get some real public support behind them.

Mmhmm. Before you go build your muni broadband network, ask yourself how you plan to get backhaul and who from, and why they'd want to give it to someone who's basically trying to poach their current and future market. Or, ask whether you'll be able to stay competitive once Verizon gets around to offering cheap fiber triple-play services straight to people's homes.

Even Google can't solve the backhaul problem, and there's a damn good reason for it.
posted by spiderwire at 10:54 PM on January 16, 2008


I actually think tiered pricing for bandwidth works well for people who don't use the internet alot (like my parents), allowing them non-dial-up service at a decent rate.

For me, it's can be a bit frustrating to pay $40-$50 a month for the net, but at least the money isn't going to AT&T.
posted by drezdn at 7:03 AM on January 17, 2008


Has AT&T Lost Its Mind?
posted by homunculus at 11:28 AM on January 17, 2008


The crux of the issue is that a whole lot of users want speed but don't really use that much transfer. They're interested in having an always-on connection that loads web pages fast, but are only actually using that connection for a few minutes a day on average.

These people don't really benefit from "per pipe" pricing, and would probably see big savings as a result of a transfer-based tiered pricing model. Obviously, heavy users tend to get hurt, and will probably get reamed as long as there's little to no last-mile competition.

Since ISPs pay for their connectivity to higher-tier networks based on speed and transfer, it makes sense that someone downloading TB per month would cost, and consequently would pay, more than somebody who just wants to check their email pretty fast. In a competitive market, you'd expect the bandwidth premium to eventually fall to whatever the 'wholesale' cost of transfer is to the ISP. But until we make some pretty sweeping changes to telecommunications law, that's not going to happen.

Instead, ISPs are going to use the pricing tiers to drive prices up to the maximum that different users are willing to pay, rather than letting them fall to something that approaches the actual cost of delivery. We can try to prevent this through regulation but to be honest I don't think it'll work -- with that much potential for profit they'll just mess with the regulatory structure to warp it however they want.

The only solution that I can see is competition, and I'm just not really sure how we're going to get it at the end-user level.
posted by Kadin2048 at 8:06 PM on January 17, 2008 [1 favorite]


FISA Debate Lookin' Good for The Telecoms?
posted by homunculus at 6:05 PM on January 23, 2008


« Older Word Sandwich: The mystery five-letter word is som...  |  Don't send in the clowns.... Newer »


This thread has been archived and is closed to new comments