Time Warner bandwidth caps arrive (updated)

[Adam8281]

FWIW, I just e-mailed my TW people here in Raleigh, sent a link to the arstechnica article, and told them I'm not in favor of caps, and would most likely cancel my service if they imposed caps here. Hopefully a lot of people will contact TW as this thing goes forward, and they'll change course.

 

[legoman666]

Originally posted by: spidey07

Originally posted by: Wheezer
so how much bandwidth does online gaming consume on average....any place to look that info up?

Very little. Just assume 384 kbs and do the math from there.

its less than that. I remember from when I played CS:S that it barely used more than dialup. It was around 70-100k. Of course, different games will use different amounts....

 

[frostedflakes]

Originally posted by: spidey07

Originally posted by: frostedflakes

Originally posted by: spidey07

Originally posted by: frostedflakes
I understand that, the point was the ISPs are screwing over a majority of the users, and only 0.4% are screwing over the ISPs. How can they not come out ahead in this scenario.

Also, most of the related costs would be relatively constant no matter how much bandwidth a person uses. Billing, for example, would be the same, as it's dependent on number of subscribers, not how much they use.

I wouldn't hold onto that .4% number. That's quite below industry norms. norms being 5-10% of the customer base.

On the topic of operational expense - the big chunk of any network operator. That IS constant, provided you do not add any more customers or any capacity. But that's not realistic as capacity and customers are added all the time and the operational expense (and associated wholesale bandwidth costs, staff, etc) rise. The good part is this is just normal business and rising operational costs are watched by profit to make sure overall profit margins are maintained.

But when operational expense squeezes too much into profit (slim profit margins as they are) then adequate forcasting/planning and fee structures have to be evaluated as the main topic of the OP illustrates.

5-10% is a lot closer to what I would have expected, I was just using the 0.4% because that's what Frizto was using to justify his argument.

And I'm sure once operational costs are added margins do get much smaller, but I doubt things are as bad as they'd like us to think. The large cablecos are still making billions in profit, how bad could they be doing? Maybe in the long run it'd be better to pull in a smaller profit for a few years and invest heavily into network upgrades. Just make sure to assure stockholders that if they stick with them, this will give the company a huge edge in the market, leading to higher profits in the long-run.

Of course, there's probably not much incentive to create a better service. ISPs will impose caps, and most people will live with it, because there aren't really any good alternatives available. Maybe once WiMax penetration is higher cablecos will be forced to actually be competitive.

OK - you need to look at overall profit margins, not dollars. It seems like you understand basic business so I'm glad you understand the business of business is making money.

Your suggestion is exactly what many cable companies are doing because of competition. The whole "network upgrade" is a gamble to spend money to make money. Just like the gamble Verizon is making with FiOS. I do like your line of thinking however and hope you are gunning for an officer level position.

You give me more credit than I deserve, I really don't know a lot about the business side of this stuff.

I agree that network upgrades are important, but I also mentioned a few posts above about network management. I'd think this would be one of the most efficient ways to increase capacity of the network (even though technically it isn't increasing the capacity, but rather reducing the abuse). Cut off that small percentage that is using so much of the network's bandwidth. Seems to me like a pretty cheap/easy way to free up *a lot* of bandwidth.

Or even better IMO (but harder to implement) would be some kind of automated dynamic management that takes into account numerous factors, such as long term usage over the last few days/weeks/months, account type, network load, etc. to make the best use of available bandwidth, something similar to but more complex than what Comcast was discussing in the link Rubycon posted.

Of course it's easy to talk about this stuff and completely gloss over the logistics of actually implementing it.

 

[legoman666]

Originally posted by: Fritzo

Originally posted by: Anubis
caps not good, way to low



http://arstechnica.com/news.ar...width-caps-arrive.html

40GB is HUGE!!!!! We service 200 employee corporations that only use 30GB/month. Turn off Bit Torrent and it won't be an issue.

I work for a big ISP- only .4% of our customers go over 10GB per month. The ones that do are impacting the network for other users, causing the need for more infrastructure and bandwidth. If those .4% are causing the other 99.6% slowness, would you expect an ISP to :

A) Charge higher fees to everyone
B) Make the heavy users pay more

I know which one seems fair to me.

Just from newsgroups alone, I have downloaded on average 164gb/mo over the past 13 months. That doesn't count browsing, youtube and what have you.

 

[Adam8281]

As I posted a few pages back, I think I'm a pretty average bandwidth user for a techie. Here are my stats for the past year or so:

May 2008: 35GB
April 2008: 37GB
March 2008: 11GB
February 2008: 39GB
January 2008: 42GB
December 2007: 28GB
November 2007: 24GB
October 2007: 33GB
September 2007: 21GB
August 2007: 23GB
July 2007: 9GB
June 2007: 46GB
May 2007: 5GB

So in my situation, I would get screwed by a bandwidth cap. My monthly average is only 27.15GB per month, well below a 40GB cap. However, there were two months during this past year when I went over 40GB and would have to pay extra for the overage. So even though I'm using LESS bandwidth than would be allotted overall, the ISP would end up making MORE money off me. Great scheme for them. Granted, at $1 per GB over the limit, I'd only owe $8 on the year, but still, that's got to add up to a lot of money in aggregate, and many of those users, like me, probably have a monthly average that is BELOW the 40GB cap.

 

[spidey07]

Originally posted by: Chiropteran

Originally posted by: spidey07
More proof also that the all you can eat residential pricing model is going to end. The heavy users kill any profit. If you really need that much bandwidth/capacity then just pay for it.

Except unused bandwidth doesn't benefit anyone. This isn't gas or electricity or water where unused portions can be saved for other customers in the future. If it's not being used, it's just lost.

Hypothetical situation.

ISP has 95 users who use 5gb/month, and 5 users who use 50gb/month. All users currently charged $10/month.

Currently, the ISP makes $1000/month from 100 users.

ISP decides to set a 10GB bandwidth cap, and charge $1 for each GB over cap. The 50gb/month users all quit, and 15 of the "regular" users also quit just because they don't like thei dea of being charged the same amount of money for a limited service that was previously unlimited.

Now the ISP makes $800/month from 80 users, and has a bandwidth surplus. But WTF good does the bandwidth surplus do? The ISP can't save it, or package it, or resell it. The actual result is a loss of 20 customers and zero real gain.

The stupid ISPs will try caps, and fail. The smart ones will gain a lot of market share.

I know it's confusing to people, but read the article earlier. It does a much better job of explaining it than I can. You can't try to think of it in such terms. You can make the simple calculations as you did but the model and costs/expense are the same of all ISPs.

The SMART ISPs will try caps and succeed. I don't know how to make this any more clearer. I build ISPs/networks for a living for the past 15 years. Dr. Pizza already summed it up pretty well in very clear terms that folks can understand.

If one doesn't agree with this approach then go get a business loan, build your ISP and stick it to those evil ISPs. I'm serious - if your model is so much better then go do it. Make your business plan, get the funding and go do it.

 

[frostedflakes]

Originally posted by: legoman666

Originally posted by: Fritzo

Originally posted by: Anubis
caps not good, way to low



http://arstechnica.com/news.ar...width-caps-arrive.html

40GB is HUGE!!!!! We service 200 employee corporations that only use 30GB/month. Turn off Bit Torrent and it won't be an issue.

I work for a big ISP- only .4% of our customers go over 10GB per month. The ones that do are impacting the network for other users, causing the need for more infrastructure and bandwidth. If those .4% are causing the other 99.6% slowness, would you expect an ISP to :

A) Charge higher fees to everyone
B) Make the heavy users pay more

I know which one seems fair to me.

Just from newsgroups alone, I have downloaded on average 164gb/mo over the past 13 months. That doesn't count browsing, youtube and what have you.

With all the money you're saving by pirating, surely you could afford to buy a business connection.

 

[spidey07]

Originally posted by: frostedflakes
You give me more credit than I deserve, I really don't know a lot about the business side of this stuff.

I agree that network upgrades are important, but I also mentioned a few posts above about network management. I'd think this would be one of the most efficient ways to increase capacity of the network (even though technically it isn't increasing the capacity, but rather reducing the abuse). Cut off that small percentage that is using so much of the network's bandwidth. Seems to me like a pretty cheap/easy way to free up *a lot* of bandwidth.

Or even better IMO (but harder to implement) would be some kind of automated dynamic management that takes into account numerous factors, such as long term usage over the last few days/weeks/months, account type, network load, etc. to make the best use of available bandwidth, something similar to but more complex than what Comcast was discussing in the link Rubycon posted.

Of course it's easy to talk about this stuff and completely gloss over the logistics of actually implementing it.

Don't sell yourself short. Your line of thinking is sound. You really should look into network planning as what you are stating is exactly what I do. By planning I mean it's all about the money, it always is.

Not a "evil corporation" planning, but a "let's try to predict trends and invest appropriately". Just like any business. The traffic management tool you are describing is not yet perfected - get to it.

 

[legoman666]

edit

 

[mrSHEiK124]

Originally posted by: MrDudeMan
Most people don't use nearly 40gb per month.

40gb per month is 133mb per hour for 10 hours a day every day. I played WoW last night for 5 hours while watching YouTube videos on my other monitor basically the entire time and I still didn't use 100mb.

I hope every ISP does this so the assholes who are using all the bandwith can start paying for it. I haven't even come close to 40gb in a month even when I was pirating software frequently. I haven't downloaded anything in the last year other than free map packs or updates and I still barely hit 5gb per month.

Are you fucking serious? You have a 1 Mb[it]ps connection or something? I can easily pull 1 MB[yte] per SECOND with my connection, that would send me over your 133 MB hourly limit in a little over two minutes.

 

[Special K]

Originally posted by: Special K

Originally posted by: spidey07

Originally posted by: Special K
The rate of progress of internet connections doesn't seem to be nearly as great as the improvements that have been made in HDTV and general PC hardware. I know they are separate industries, but of all the consumer electronic industries, high speed internet seems to have made the least amount of progress.

Not really. In general capacity/bandwidth follows moore's law roughly doubling every two years. Now it may not be apparent from the access layer (your home) because by somewhat limiting at that layer you can stave off more serious capacity problems. But it does double every two years, both usage and capacity.

So basically the "backbone" that is feeding my area of subscribers may have increased dramatically in speed over the years, but rather than advertise higher speeds for the end user, they instead use that speed increase to put more people on that "backbone"? That would be consistent with what I have observed - i.e. that advertised speeds in the areas I have lived really haven't changed much over the past few years. Sorry if my terminology and use of the word "backbone" is off, I do not work in the ISP industry. I am just trying to understand what is going on.

spidey, can you clarify my post above? Is that how the ISPs are making use of the new bandwidth, which is why the individual speeds seen by the end user don't seem to have increased much over the years?

 

[Fritzo]

Originally posted by: Xavier434

Originally posted by: Fritzo

I'll be glad to explain. Say you have a 100Mb pipe from an ATM. If you have 10 people at 10Mb running full bore 24/7, it's going to screw everyone else using that pipe.

This doesn't seem to be an issue in Japan. Why is it an issue here? The previous explaination made by another poster doesn't make sense since ISPs can simply purchase less total bandwidth to distribute if there are less people to sell to.

I'll gladly explain this as well. Japan is an island with a land area of around 233000 square miles. They also started building their communications infrastructure in the last 50 years or so.

The US land area covers nearly 5700000 square miles----and has a very spread out population. On top of this, much of our communications infrastructure is 50, and even 100 years old in some places. It would bankrupt any company to upgrade this much area, so deployment is slow and expensive.

So:

Japan= Smaller, more concentrated population + more modern infrastructure = more available bandwidth

US= Wide area, less concentrated population + antiquated infrastructure = less readily available bandwidth. To provide more means costs have to go up, or networks can be optimized. Optimization is free and only singles out people that use networks in ways for which they were not designed.

As the US infrastructure grows, caps will go away, but for right now, they're going to become necessary. Other countries have the luxury of being able to build new, where we have to tear down and rebuild.

 

[Fritzo]

Originally posted by: zanejohnson
this makes me love my connection even more, we almost went with time warner too, instead we went with the ethernet drops in room that the apt complex we live in offers... 30 bucks a month, no bandwidth caps whatsoever.... symetrical speeds up and down... varys throughout the day from 2.5mbps (lowest i've seen yet) to 6-7mbps (highest i've seen yet)

Oh reeeeely?

 

[Fritzo]

Originally posted by: JeffreyLebowski

Originally posted by: Fritzo

Originally posted by: Anubis

Originally posted by: Fritzo

Originally posted by: Anubis
caps not good, way to low



http://arstechnica.com/news.ar...width-caps-arrive.html

40GB is HUGE!!!!! We service 200 employee corporations that only use 30GB/month. Turn off Bit Torrent and it won't be an issue.

last month i downloaded a grand total of 1 torrent for less then 400 megs

i used 65 GB of bandwidth on a 8mb connection, no news groups no illigal shit

its too fucking low

Then you're a power user. Pay more.

Why, you're company said, 6Mbps unlimited downloads. OK, so that means. I GET UNLIMITED DOWNLOADS. If you can't live up to your part of the contract, then to bad. But if I can't then I should have to pay all kinds of fees right?

OK, MY company doesn't say unlimited downloads anywhere, and there's not may places that say Unlimited Downloads anywhere anymore.

 

[Fritzo]

Originally posted by: spidey07

Originally posted by: Xavier434

Originally posted by: Special K

Originally posted by: Gothgar
I've been with TWC for a while too, and if this goes live anywhere in California I will be a DTV/ DSL customer over night, fuck that shit.

I probably on average don't even get close to that cap, but it is mostly just principle.

What if all the other ISPs adopt similar plans? Then we're all screwed

This is why we need more competition. Corporate greed and the desire to steal "the other guy's" customers will result in price plans and deals which benefit the consumer a lot especially considering that placing limitations like bandwidth caps gives the competitors a much larger weapon to play with.

ISP is a highly competitive market and most broadband customers have a choice.

You're free to go and start an ISP if you feel the profit margins are so high, because they are not. Mostly because it is so competitive.

In fact it is such a mature market that these innovative pricing models are being tried because there is so little money in the all you can eat model for such a cheap price. Most of the decision makers I talk to in the industry all say the same thing "all you can eat is going away, we just can't keep losing money like this and have to do something."

On average, companies make around $3/DSL account and cable companies make around $6/account. DSL infrastructure is more expensive to maintain and implement. Out of that money, they have to pay employees, health insurance, and maintenence. You're looking at about $1-$1.25 profit per account. The big money is in the service selling, not the connection itself. They make their cash off of extra software, speed upgrades, etc.

So, if you really want to screw your Internet provider, get the lowest priced plan you can get, install any "free" software that comes with the service, and run torrents 24/7. You'll end up costing them money

 

[spidey07]

Originally posted by: Special K
spidey, can you clarify my post above? Is that how the ISPs are making use of the new bandwidth, which is why the individual speeds seen by the end user don't seem to have increased much over the years?

I really would love to answer your question. But I can't do a whiteboard in text format. That's what makes my position so difficult, those that work in the industry understand.

Your question is one of "the last mile". Laying cable and gear to service that last mile from provider premise to the customer costs a lotta dollars. And the return for that capital investment is hard to prove if it isn't there.

It all goes back to subscriber density per route mile of fiber.

 

[Fritzo]

Originally posted by: spidey07

Originally posted by: frostedflakes
I understand that, the point was the ISPs are screwing over a majority of the users, and only 0.4% are screwing over the ISPs. How can they not come out ahead in this scenario.

Also, most of the related costs would be relatively constant no matter how much bandwidth a person uses. Billing, for example, would be the same, as it's dependent on number of subscribers, not how much they use.

I wouldn't hold onto that .4% number. That's quite below industry norms. norms being 5-10% of the customer base.

On the topic of operational expense - the big chunk of any network operator. That IS constant, provided you do not add any more customers or any capacity. But that's not realistic as capacity and customers are added all the time and the operational expense (and associated wholesale bandwidth costs, staff, etc) rise. The good part is this is just normal business and rising operational costs are watched by profit to make sure overall profit margins are maintained.

But when operational expense squeezes too much into profit (slim profit margins as they are) then adequate forcasting/planning and fee structures have to be evaluated as the main topic of the OP illustrates.

I got that figure from the last ISPCon in Chicago 3 weeks ago. This was a big topic of discussion there because a small number of abusers are really hurting ISP's running on paper thin profit models. It's accurate. 10% of customers average above 20GB/month.

 

[SoulAssassin]

I just realized I'm old because I remember when internet access was charged by the minute. Anyways, we should be evolving to the point where internet access is a ubiquitous commodity. Hopefully people affected by this have other ISP choices and will speak with their wallets.

 

[mugs]

Originally posted by: SoulAssassin
I just realized I'm old because I remember when internet access was charged by the minute. Anyways, we should be evolving to the point where internet access is a ubiquitous commodity. Hopefully people affected by this have other ISP choices and will speak with their wallets.

What's worse, I had to dial long-distance because I had no local ISPs. :Q At the time the long distance rates were $.25 for daytime, $.15 for evening and $.12 for late night.

 

[Fritzo]

Originally posted by: SoulAssassin
I just realized I'm old because I remember when internet access was charged by the minute. Anyways, we should be evolving to the point where internet access is a ubiquitous commodity. Hopefully people affected by this have other ISP choices and will speak with their wallets.

We will be there, but we don't have a way to finance the network needed to do that yet. People don't seem to understand that. I'm going to say 10-15 years down the road when we get more fiber ran and more wireless infrastructure is in place caps will no longer be needed. Keep in mind that even if caps are in place, they're most likely only going to nail you if a pattern is in place (ie- if you go over one month, you'll be watched for a repeat the second month).

 

[n7]

All i have to say is...

40 GB?

Wow, what a joke.

 

[Evadman]

$1 a gig? For that I can ship a hard drive to china and have them type 1's and 0's to fill it, ship it back, and still come out ahead. Why would I pay $1 a gig when I can buy the physical medium for about 12 cents?

 

[Rubycon]

Originally posted by: Evadman
$1 a gig? For that I can ship a hard drive to china and have them type 1's and 0's to fill it, ship it back, and still come out ahead. Why would I pay $1 a gig when I can buy the physical medium for about 12 cents?

Just buy a Velociraptor. Same price and much faster plus it works anywhere there is power. :laugh:

 

[Aikouka]

I was talking to a friend of mine that works at TWC and he couldn't tell me much about it for my area (I think he's not allowed to ), but he said not to worry about it. That bastard better be right...! The thing is, I don't see why it would even be necessary in my area? I've never had any degradation of my service nor have any of the people that live near me that I know have had any. I could understand implementing this in trouble areas where there's a lot of abusing and it's hurting legitimate users to help give people what they pay for (although, one should understand that cable Internet is a "community" type of thing, so degradation at peak times shouldn't be out of this world ).

 

[ManBearPig]

I will switch our service over to another company as soon as this "test" arrives here.