Comcast XFINITY → Re: Bandwidth Limits - All discussion here

That sounds like you are making a few leaps of faith there. Abuse is handled on a case by case basis from what I've seen. If many users start complaining about slow speeds off a certain node it is passed over to abuse for investigation. Or conversely if the market engineers notice that a node with only 50 subscribers is routinely running hot on their capacity reports it is passed over to abuse.

I don't think the top .01% are kicked off every year for abuse or even get "the call." If that were the case I would imagine we would see quite a few more posts here complaining about it. It seems to me that it would be a waste of resources to investigate every sub who went over 250gb UNLESS they were actually causing an issue.

That was my sense a few years ago, but lately it seems more like "status quo." How confident are you that it's still been running case-by-case?I think the number is low-ball. As someone said previously, it really is a very useful number for someone who has just received "the call."

If it turns out that your sense on this is right, My God! What an overreaction this all has been!

Fairly confident.An overreaction on who's part? The small faction of people who made a stink about getting kicked off or Comcast itself?

Lawsuits like this:

»jacksonville.bizjournals ··· y38.html

Probably encouraged Comcast to publicly state a cap. Not to mention all the scrutiny the FCC investigation drummed up. The FCC wanted Comcast to be as transparent as they could about their network management policies so I think the cap is partly a result of that.

I don't agree at all. I got the call (and part of the reason was their fault-service didn't work properly) and after thorough investigation, I was told that no one complained, and that I was NOT negatively impacting my node. As one guy from "Quality Assurance" (what a misleading name) told me, "you made the list". He said it was his job to call, not allow me to speak to anyone else, and not to give his name. If I didn't like it, then I should sue them. He said some other things too. I had Comcast triple play, switched it all, and what happened to me got others to switch just for the point of how they acted. I will never go back to them, even though I am back to 3 MBPS DSL. I don't care.
You should do internet searches and just see how many people get the call.

Comcast wouldn't have so many people complaining to the FCC, FTC, etc, if they were honest about things, so remember that also. It isn't up to others either to judge how much is ok and how much isn't. Unfortunately, the ignorant point of view is that "If you use xyz amount you are doing illegal things". And yes, Comcast took this view with me, until I informed him that I showed EVERYTHING I was downloading to the tech when he came out to fix my service problem. After all, we had to test it to make sure he fixed it. Funny how he also did not take me up on my offer to have them come over any time they wanted without warning, and I'd give them full access to the computer to see what was done. They did a lot of ASSUMING, and you know what they say about that.

How much did you download?

I'm not denying that people get kicked off, again I just don't think 14,000 people are getting "the call" each year. Most of the evidence we have is anecdotal and comes from a very vocal minority.

Logically why would you want to waste resources on non-issues (i.e people using lots of bandwidth but NOT adversely affecting the system?) To get bad PR, make sure you hire extra people to handle all the abuse calls? I suppose it could be true but I have to think Comcast is just trying to protect its bottom line.

It's 14,000 per MONTH. They told me the lists and calls are monthly, at the percentages quoted often. Comcast is trying to avoid getting their infrastructure into the 21st Century, and they don't want people streaming movies and everything else that would hurt their own business. My total was a little north of 250. In the AUP it says they can do this if you are harming your neighbors (paraphrasing). They admitted I was not and they had zero complaints.

YES! If you're right, then everyone is overreacting.

Think about it. If warnings and disconnections really only happen 0.01% of the time (1 per 10,000 customers), then Comcast has shot itself in the foot and got a lot of people excited over nothing. If true, then it doesn't have a 250 GB cap!

If the only people that get a call are people who Comcast can demonstrate are operating their Internet connection in a way that unduly impacts others, then there is no bandwidth cap. The bandwidth measure is just an "early indicator" not unlike possible other indicators of accounts with atypical patterns that would tell an investigator who might be a likely source for user-to-user interference.

I don't think you're pulling my leg, but I do think that you've got the wrong impression about what has been going on -- especially recently.

Big case in point: Dave Winer ... this is the guy who, without exaggeration, is deserving of the most credit for everything we now call "Web 2.0." He's not some kiddie with a BitTorrent fetish. I find it unlikely that he was doing something nefarious. It is most likely that he rang the invisible bell simply by transferring podcasts from one server to another.

Wait -- there's a lot of things you can blame on Comcast, but the mass panic has been incited by the media like your friends at Free Press and a certain news reporter here on this forum. You've taken a handful of cases, given them a nationwide megaphone, allowed them to make wild claims, and are surprised there's been a reaction? When people get caught by exception, they get angry about it -- and that's the message you're pushing forward. When was the last time you heard someone talk about getting a speeding ticket where they didn't talk about how they were unfairly picked out by the cop? Most of us don't take responsibility or blame very well, so when you have an out like Comcast having some questionable practices, it's going to be in our nature to shine as much of the spotlight on that as possible.

Writing articles about how the infrastructure actually works is boring. Nobody wants to hear that your local cable/DSL/FTTH segment can't move an unlimited number of bits every month, and due to the shared nature of it all every bit you use on the wire is a bit that your neighbors can't use. You get more press coverage if you talk about how big bad Comcast is screwing the average little user. The only problem here is, the users you're holding up as examples aren't average users.Exactly. Nobody represents the average broadband user more than Dave.

Free Press had nothing to do with this. They got phone calls in response to this story. It looks like this story was a response to the Florida AG, who apparently was working on this in a cave somewhere.That's not the same. In this case, the cop is giving out SUSPENSIONS for running cars off the road based soley on the output of a radar gun.No, they aren't average users. They are exceptional users of bandwidth, and I've only held up two specific users as examples -- one is Dave Winer, and the other one's story is still being written and -- like Dave -- he was using his bandwidth in a completely non-nefarious way.

(The disclosure of the invisible cap actually is a better thing for users like these, because now they can know something before making a purchase decision. No cap would be best, however.)

Comcast gets the reaction they do because of their actions, it is not fair to blame that on everyone else. If they were so concerned about that, they should have thought of that before they handled things as poorly as they have. Lying about throttling when they had been caught is just one example.
I don't consider it a handful of cases, it's been going on for years and has been a cumulative situation. 1+1 only equals 2, but do it long enough, and eventually the numbers get pretty big.
Think about how different things MIGHT have been had they admitted to throttling when caught, and come up with a better defense, reaction, and solution. Same thing with the caps. Telling people there are none and then saying there are but we can't tell you because it changes every month doesn't fly. Yes, that's what was said to me, in the span of 2 different questions.

This is a discussion of straight quantity -- who cares what they're using the bandwidth for? It's not like the bits don't get counted on the wire if they're using their bandwidth to save kittens.Face it, the only way we're going back to no caps is if there is usage based billing, or we have speed downgrades. Every broadband ISP that advertises 'unlimited' service is lying. They don't have the bandwidth to deliver 100% use to every subscriber simultaneously, and if everybody started having usage patterns of the people you are describing they would go out of business because their infrastructure costs would be greater than their subscriber revenue.

I don't disagree that the marketing department (what I think of as the "front side of the house") has done a miserable job.The claim is 14,000 people per month. Can you find just 100 posting on the Internet from unique users out of that group of users? You can go back over the last year on that search if you want. With the numbers people are throwing out, Comcast should have disconnected 168,000 subscribers in the last 12 months, so finding 0.05% of them shouldn't be all that difficult. (they are heavy Internet users, afterall)

I still fail to see the actual cost of sending bits over a copper wire (or any wire for that matter).

They bought the equipment. The lines are going to be "live" no matter what. No matter if you use it or not.

The only real cost comes in use of electricity, paying your employees to mismanage the traffic, pay their "rent" to their service providers (after all, their customer traffic will need to connect to the internet backbones, right? Who owns that?) and any state/federal taxes.

And for the argument of pay-per-use, I don't think they'll ever truly charge by how much you use. Rather, it'll start at a flat rate up to a certain amount of traffic, then anything over would be designated as pay-per-use. Like hell they'll have millions of subscribers have a $7.50 monthly bill.

I have never counted numbers, but have searched enough and have seen posts and blogs on this going back at least a few years. 100 would be quite easy I believe, but that doesn't really matter. I have a few bookmarked, I only did to the more interesting ones.

The cost isn't in moving bits. The network involves more than just routers, CMTS head-ends, and nodes; there's a ton of people that go into keeping things functional. Those people get a paycheck whether you use 0bytes or 250,000,000bytes.

The costs are largely based on the percentage of infrastructure you use. When you're on a shared 38mbps cable segment, when you are using your connection at 8mbps that capacity is not available to everyone else on the segment. As long as there is idle capacity there isn't a problem, but when enough people start to use the network heavily enough the impact is typically felt by all users on the node. (the new QoS implementation helps address some of that)

Everyone pays roughly the same amount for Internet access, but the usage is greatly varied. This leads to a Tragedy of the Commons situation.

The Tragedy of the Commons is the title of an influential article written by Garrett Hardin, first published in the journal Science in 1968.[1] The article describes a dilemma in which multiple individuals acting independently in their own self-interest can ultimately destroy a shared resource even where it is clear that it is not in anyone's long term interest for this to happen.

Central to Hardin's article is a metaphor of herders sharing a common parcel of land (the commons) on which they are all entitled to let their cows graze. In Hardin's view it is in each herder’s interest to put as many cows as possible onto the land even if the commons is damaged as a result. The herder receives all the benefits from the additional cows but the damage to the commons is shared by the entire group. If all herders make this individually rational decision, however, the commons is destroyed and all herders suffer.

Source: »en.wikipedia.org/wiki/Tr ··· _commons Original credit pfsmith in post »Re: Just Like the Electric Company!

There are real limits on how much capacity the infrastructure can support today. It's been 2 years since the DOCSIS 3.0 standard was ratified, and we still only have a single CMTS vendor that is fully supported with both upstream and downstream channel bonding. In the meantime, upgrades are expensive, and there is no financial backing to implement them because under flat rate pricing they get the same income whether they invest $0 or invest millions in their backbone. (provided they can keep the performance high enough that most subscriber don't defect)

Broadband access is like public transportation. It's engineered for cost, and it's impossible for it to ever be everything to everyone. It is possible, however, for it to be good enough for the overwhelming majority.

With respect to that, then, why is it such a big deal for Comcast to tell a user that takes full advantage of that line when nobody else in their node cares to? If the impact isn't felt and no complaints are being made, does Comcast still have the arguing point? I don't think so. If they have to lie about it actually causing problems, I wouldn't be inclined to change my usage habit. But on the other hand if they were truthful and provided clear information that others are in fact being impacted because of my use, I would then be better inclined to care what was happening.

The caps are purely a financial stop-loss maneuver. They know what their monthly operating costs are per DOCSIS segment, they know what their target profit margin is (as does anyone who reads their financial statements, since they're a publicly traded company), and they've calculated out what percentage of the infrastructure each subscriber can use before they start losing losing money on that subscriber. I don't know Comcast's specific costs, but I think it's safe to say that their break-even point is somewhere under 100GB per subscriber.

It has become a game of "chicken" to see how far they dare stretch the truth before it breaks -- and maybe this is it and some correction has to happen. That said, I'm not ready to say that the business of bandwidth aggregation has broken down. And any speed "downgrades" we might get will, in actuality, be speed we never had in the first place. This 250 GB cap may be a substitute (a speed tier is just a "per second" bandwidth cap).

We need not set the goal that they "deliver 100% use to every subscriber simultaneously." That's unreasonable. However, they must be reasonably able to deliver the service levels that they are selling based on usage patterns and network upgrades -- and that's the part of this issue that is falling down right now.

At the end of the day, this is what's happening: Even if they don't know it, people are buying into one share in two pools of bandwidth. Before they do, they ought to know how many subscribers are in the upload pool and what bandwidth is available in it, how many subscribers are in the download pool and and what bandwidth is available in it, what the community's usage patterns are.

Now -- one way to proceed for ANYISP is that it can sell shares of that, if it wants to. But it has to update the buyers as to what they're buying every month. If they've done that, then the deal is fair. The ISP can't really underperform when it does that. It just makes the connections and sits back. If my neighbor and I were sharing a T1, it's probably the arrangement we would make between us.

If ANYISP wants to relieve the buyer of the burden of understanding all of that, then they can sell a "tier" and package certain expectations together -- that's useful. However, it breaks down if the ISP acts unethically and sells tiers that it can't responsibly expect to deliver. If it underdelivers, then the ISP should incur a cost -- either in bearing the additional expense to deliver what it overpromised, or by rebating the difference back to the customers. Selling tiers is what the ISPs have been doing, and it works, and has worked, for many many years.

I maintain that people haven't changed all that much. There is nothing new here. We've always had extreme users of bandwidth and we've always had light users of bandwidth. And the trend is not increasing, it is slowing down. With those facts, I can't rationalize that the years of "unlimited" are over.

If comcast customers only use 3 or 4 gigs a month. and only about 1 percent of customers use 250 gigs or more then again somebody is lying. This isn't possible for the following reasons. comcast being the 2nd largest isp in america is saying there giving a 250gig cap to all their customers every month to be fair, but only 1% of these custoemrs will be effected by it. that's about 1million people.

If this is true and only 1percent of comcast customers come anywhere close to this cap, then it's truly unfair to band customers for a year who break it. because these 1 million or so customers aren't truly effecting other customers. lets do the math. their offically claiming that there network can support over 3.5 billion gigs of bandwidth a month.

3.5 billion a month. i'm taking 14millions customers times 250 gigs. buy all standards that's alot of bandwidth. here's the twist. comcast claims 99% of them don't use anywhere near that much. so again lets do some math. take 13 millions times 20gigs a month. 260 million gigs a month.
that still leaves you with over 3.4billions gigs left over.

Comcast is claiming that 1% of their customers will use 250gigs or more. so lets alittle more math shall we. take 1 million times 250 gigs. 25 million gigs.

now comcast claims on average 4 or 5 gigs are used by there customers, i gave them them 20 gigs a month. now lets add shall we

260 million plus 25 million 285 million gigs a month is the average amounth of bandwidth comcast is planning to see. So if there claiming their networks can handle 3.5 billions gigs a month and there only planning to use 385 millions gigs. then honestly speaking their lying about their networks, and their lying and the need to have such low caps. or their lying about how much bandwidth the average customers uses. or their lying about the reason why they have these caps.

comcast is taking advantage us of, And i believe i just proved it.

I need to ignore marketing and think about what I am getting for my money. Essentially, I have a shared connection on which I can transfer ~200GB /month at inbound burst speeds up to ~16Mbit/sec (8MBit/sec sustained), and ~2MBit/sec outbound . Even when my neighbors are beating up on their part of our shared connection, I am never more than 50MS from my first hop, and reasonable bandwidth is available to me. That is, unless a truck takes a telephone pole out, which has only happened once, and they fixed that quickly. I pay $67/month for this. My other internet option is dialup. A T1 or fractional would be expensive for my location. Frontier may someday make DSL available to me, but at 5GB/month, I wouldn't last a week. I believe that residential internet customers are of no value to Frontier. Until Verizon eats Frontier and sets me up with FIOS, Comcast is my only choice for anything better than dialup. So far, they have been reliable for me. I prefer network management (protocol agnostic throttling) to hard caps, but I prefer a hard transfer limitation to an invisible one.

Of course, if I could get and afford an OC-48, I would have one.

Exactly right.

While the bandwidth cap does have some problems with it, that you can now go through the above process is useful to consumers.

While your version is a far cry from the current marketing used by all of the ISPs (not just Comcast), ISPs should take a look at your version and see that the truth doesn't really read so badly and that customers deserve to know what they're really offering.

Robb