I think he's tying speed and data more closely than is justifiable for most people.
Many people (I'd guess maybe even most) want higher speeds not in order to use more data. They want it in order to use the same data as they now use but with less waiting.
Comcast just upped me from 220 mbps to 500, for example. That is not going to have much affect on my data usage. It will just mean that when, for example, I decide to play an old MMORPG that I haven't played in a long time I won't have to wait as long to download all the patches I missed. It won't mean that I'll play more games.
Same story when a year ago they upped me from 120 to 220. And a year before that when it went from 90 to 120.
I see a natural inflation in data usage because the services I use are inflating the data supplied to my requests. Streamed video sizes increase, I actually have had to use my cloud back up (3 times unfortunately) in the last 3 months so I've had to download 250gb * 3 because I actually had a cloud back up, web page sizes increase from ad-bloat, my hard disk is larger, so my back up size needs increase.. And so on, and so on, and so on...
The speed increases are just a way to eventually tip you into data overages if the data cap does not increase proportionally to the speed increase.
10Mbps is more than I need. But their slowest plan is 25Mbps.
The Romanian model of "micro ISPs," ISPs serving a few blocks here and there, is also illegal in most U.S. cities. The U.S. places a very high value on universal coverage, both of lower-income people and rural people. In nearly any U.S. city, the only way you will get permission to build Internet service to anyone is if you agree to also serve low-income communities or rural parts of town, even if doing so isn't profitable.
I used to live in an apartment building with both cable and fiber. Nobody subscribed to the fiber because the fiber provided didn’t have a franchise to offer television service. It was nuts—I was the first person on a floor of almost a hundred apartments to get the fiber service. I found out because the ONT had been misconfigured when it was first installed years before. People just didn’t care—they wanted their cable package with ESPN. It’s why Google Fiber offers a television package—a broadband only service is not competitive in most US markets.
Now we have 3 major ISPs and they all offer similar services.
The fact I pay 75 dollars for the same service doesn't sound so absurd after accounting for that.
And like consumer electronics, which easily flow across borders, which is why an Xbox costs the same in the USA and Romania, ISP is a local service that relies on local labor costs.
Despite the source the following breaks down the profit margins based on the telco's own information that they provide.
Government-sanctioned monopolies, and regulation to prevent competition from having a fighting chance. Basically regulation in the wrong places, while looking the other way in areas that could actually use more regulation.
Jon Brodkin has to write a "ISPs are evil, support net neutrality" article literally daily, but data caps are very simple: A way for Comcast to increase the cost of service to heavier service users, rather than raising prices for all customers equally.
A 1TB cap feels reasonable.
Data caps only exist in the context of an ISP covering their ass for overselling. "Fair" and "reasonable" are just a matter of perspective, one I can never sign onto when they're the fine print on misleadingly marketed product.
If anything, Comcast's past behaviors warrants not giving them the benefit of doubt.
edit: I guess a downvote indicates that I'm wrong. The lowering of price is definitely due to the overwhelming generosity of Comcast. /s
> And, their Gigabit plans have no cap at all
Comcast could make the argument when they were overprovisioning their nodes like crazy the network sucked, but higher order modulation and more channels in newer DOCSIS standards, fiber being run everywhere to nodes has really improved the situation.
Less of a cash grab, more of a bandaid to handle misallocated resources.
The final argument could have been made for peering costs, but Comcast is really big now, so I doubt they have to pay much if anything for transit.
Throttling is a reasonable response to the power-law distribution of network usage combined with flat-rate pricing.
At a certain point on the curve, a marginal user will cost more to service than they will produce in revenue. This means (a) you service them at a loss or (b) distribute their loss over everyone else, i.e. increase prices. The problem with (b) is as you raise prices, people at the other end of the distribution (the unusually light, i.e. profitable, users) will start de-camping. Balancing these factors is difficult, and almost necessitates capping or throttling.
The dumb move was promising unlimited data for an unlimited period of time.
Now that doesn't mean data caps themselves are the right solution to the "problem" (conceptually!), I'm just speaking to change the frame of mind of this.
Imo, what we should be demanding is an SLA for normal customers. Cap speed or data, I don't care, but give me what I'm paying for at all times.
As I see it right now you're getting capped on bandwidth, total data, and absolutely no guarantee about any of it working. It's the worst of all aspects.. for customers, great for Comcast I'm sure.
Users feel like, "Hey I paid for a 1Gbps link, I should be able to saturate 1Gbps 24/7".
Providers know that 99% of users don't need 1Gbps 24/7, so to reduce the cost of the service, they sparsely provision.
Caps of some form, either on speed or on bandwidth (or N Mbps speed cap after Y GBs of bandwidth) are the way they keep users within the limits of the network.
Ie, if you're overselling by a large margin than you'll need to be capping a lot of people. Not the same amount as the "large margin", but a significant portion of your users need to not be accessing the pipe at 5pm.
This is why I think they should just get rid of caps, and instead give us a meaningful SLA. They won't however, because they have massively oversold their network.
: I don't know if they've actually oversold their network, it's just clear that there is data troubles based on user traffic. Personally I think Comcast is purposefully limiting traffic based on marketing strategies. Ie, when Google rolled fiber into Portland, suddenly Comcast's "available speeds" massively spiked to stay competitive with Google. Which indicates that they had a lot more than they were selling.
If you're selling XMbps for $Y then I should be able to use XMbps, regardless of total usage. That's the unit that they have chosen to market, "speed of connection".
The problem is that I have to pay for this this two ways. As it stands, they're going to charge me for a 125Mbps connection, then they also get to charge me more if I use that 125Mbps for more than 18 hours in a month. To reverse your statement: Why pretend that I have XMbps of data if I can't use it?
The rest of your statement makes a lot of sense, and the SLA is what's really missing. If that was addressed in addition to / in lieu of neutrality it would be a net positive.
You're also paying a fraction of what it would cost to get a line with an SLA, even in a commercial building with several competing providers.
Because from a networking point of view, data caps is an artificial thing.
> Nothing is unlimited, so why pretend you can get an unlimited amount of data?
Your connection is not unlimited anyway; the max you can get out of it is MAX BW * seconds/month.
Since the providers don't have guarantees about the max bandwidth, it can vary depending on the demand from the other users, I don't see why you need an artificial data cap.
How so? If I limit you to 1MB a month, good luck clogging my pipes. I said it's not an effective method of approaching the problem, so I'm not defending them at all.
> Your connection is not unlimited anyway; the max you can get out of it is MAX BW * seconds/month.
Not true, if you have a data cap. Hence why data caps theoretically could serve a purpose. Again though, I'm not defending them.
> Since the providers don't have guarantees about the max bandwidth, it can vary depending on the demand from the other users, I don't see why you need an artificial data cap.
I agree. My point however is that complaining about data caps is besides the point entirely. We're lacking an SLA, we're lacking any way to ensure that we are getting what we're paying for.
Data caps provide no real benefit to the idea that they sell of reducing network congestion. The only way they could is to have a very aggressive data cap, resulting in a large X% of your users not even having internet access due to being capped. This of course, is basically impossible, as people would flip out.
By focusing on data caps we are, in my opinion, focusing on the red herring. Rather than getting what we're paying for on bandwidth, we're focusing on how much total data we can download.
The point of a data cap is to get users to self-manage their usage, not to block a large percentage of users.
Rather than getting what we're paying for on bandwidth...
I think by and large we get exactly what we pay for. Have you noticed that business class service, which has SLAs & generally better-delivers agreed speeds & bandwidth at all times, costs a lot more?
Residential users are paying for sparsely provisioned links, and that is what they are getting. But most do not realize that.
> How so? If I limit you to 1MB a month, good luck clogging my pipes. I said it's not an effective method of approaching the problem, so I'm not defending them at all.
Is your argument that "data caps" are not an artificial thing because you can have.. "data caps"?
In that context yes I agree, they are artificial.
Comcast offers service at speed tiers. When you pay for a speed tier you would expect to be able to use that speed a reasonable amount of time without having to spend extra. Nobody is expecting to max out their 150 Mbps connection 24/7 but that same connection starts hitting additional fees after 14 hours of maxing that connection in a month.
Data caps are a clear money grab because they need new revenue for their dying TV business. A business dying because of their failure to innovate and constant gouging of customers due to monopolistic practices.
Before Comcast had a punitive financial threshold they used to throttle the heaviest users. Now they simply charge those users, while presumably the majority of people are cognizant that they should use some discretion to avoid the fees.
Another user opines "it’s still limited by your maximum throughput and the number of days in a month" and this is an argument that seriously rubs me the wrong way because it's effectively a tragedy of the commons type argument -- I love having blistering fast internet when I need to download something, etc. But I realize I don't have a committed 500Mbps across the internet, and not far from me it's a shared resource.
I actually think charging for use makes a lot of sense. There’s no reason my monthly bill should be the same as my neighbor when they use it for nothing but email.
As a thought experiment, what if the caps went away tomorrow? Is it possible that the network would become saturated? If so, that is to say the caps might be reasonable. If not, how do you know that's the case?
I'm not any happier with Comcast than anyone else; I recently moved from 1Gbps service to 250Mbps service with them, and I was always bumping up against the 'cap' on my 1Gbps service. I want a better provider. But nothing in the article proves that the caps are useless nor a money grab.
One can disagree with the case they're making. But the headline was (now changed here): "Comcast disabled throttling system, proving data cap is just a money grab." The article covers both the fact that Comcast has disabled their throttling system, and why the writer thinks the remaining data cap is a money-grab.
Whether you agree that the data caps are money-grabs or not is completely irrelevant to whether the headline is (was) appropriate. Your argument seems to be with the article, not the headline.
The headline is like the label on box, and your argument seems to be with the contents of the box.
That's on Comcast to handle as a part of building and operating a network. It should not be the concern of the customer.