Goodbye to a flat rate for broadband?

[erol]

Originally Posted by ianathuth

Erol, if you wish to use analogies, please use ones that are realistic and comparable.

I must say that is pretty rich comming from the person that introduced a _gunman_ into the train analogy !

Originally Posted by ianathuth

I am not so het up over the situation and if you think that the "mythical" 5% is not correct then what would your percentage be. It would seem logical to anyone that has an iota of commonsense and has followed these forums over the past year that a very small percentage of users consume a very large percentage of bandwidth and that they are primarily file traders using p2p or newsgroups.

You are _still_ confusing the usage of BW with the cause of congestion. I am not in the business of 'making up' figures, just countering the telcos 'made up' figures. So the reason you do not move to an ISP that would block P2P traffic and give you the connection you 'want' is that you are not the bothered really ? Strange thats not the impression I got.

Originally Posted by ianathuth

Your statement of "What has any of the above got to do with my disagreement that 5% of BB users do (or even physical can) create 60-80% of the congestion in peak periods ?" is posing a question that was not the initial subject of the thread. 5% of users CAN consume 60-80% of the bandwidth and this is very easily proved, just think about it for a moment and then tell me how they can't.

You still do not get it. Just because (if it is ture) 5 % use 60%-80% of the total bandwidth, that does NOT mean that they are responsible for the same % of _congestion_ in peak periods. Just because my 5% of beer drinkers might consume 60-80% of the total beer serverd in 24hrs, it does not mean they are responsible for 60-80% of the bar congestion. When they are in the bar during peak periods they cause _exactly_ one person's worth of congestion per person, anything more is _physically_ impossible. Just because a driver might do 1000's of miles on the roads, when there is congestion, they cause _exactly_ one cars worth of congestion, anything more is _physically impossible_. Yet if an BB users consumes 'higher than average' bandwidth, then when they are usinf it during busy periods, they somehow cause way more congestion that one connection can, at least according to you and telco propaganda. This is simply as _physicaly impossible_ as my two analogies and is at the root of my problems with statements that '5% are responsible for 60-80% of CONGESTION'. Just not possible. That's logic.

Originally Posted by ianathuth

Your solutions are theoretically correct, but practically have NTL for one got the financial resources to properly scale their resources and what should be the "price". If I or you had the solution to the problems we would be the ones getting £3,500 per day, not Barclay.

You are priceless. So NTL have got themselves into a total mess finacialy, wiped out 1000s of original shareholder money through bad management, strategic incompetence and acquisitional follies, so we can't expect them to scale their networks properly. And the man most responsible for this state of affairs ? Yep you got, it's the same BK, that you seem to think must have all the answers because he is paid £3500 per day ! Priceless.

I think that is about all for me on this thread now. I have done my best to explain why one person can cause no more than one person's amount of congestion (something so eveidently self obvious to me that I sometimes despair for others). I am obviously getting no where and to be honest I am not sure I can 'keep on' without slipping into the 'personal' which is something I do not want to do.

So thanks for the discussion so far. I like to say it's been fun but I try and avoid lying as much as I can. I'll just leave you with one final analogy from my part of the world.

You can lead a camel to water but you can't make him drink, bacause 5 camels have somehow expanded their body size, through some mytstical process, that allows them to take up all the space round the water that can normaly accomodate 50-100 camels simultaneously.

[ian@huth]

The problem with cable in particular is that broadband is running on a system that was not designed specifically for it. The planning and infrastructure was based on the model of a customer that does certain things, but that model is no longer valid. Times have changed and current days applications consume far more bandwidth than the planners ever thought. The more that todays customers use their available bandwidth, the sooner you get to a state of congestion. If NTL was to implement even higher service tiers then the congestion would arrive much sooner and there is a limit to what cablecos can increase the service tiers to without a very radical change in their infrastructure which they cannot afford.

[Stuart]

Originally Posted by erol

You are _still_ confusing the usage of BW with the cause of congestion. I am not in the business of 'making up' figures, just countering the telcos 'made up' figures. So the reason you do not move to an ISP that would block P2P traffic and give you the connection you 'want' is that you are not the bothered really ? Strange thats not the impression I got.



You still do not get it. Just because (if it is ture) 5 % use 60%-80% of the total bandwidth, that does NOT mean that they are responsible for the same % of _congestion_ in peak periods. Just because my 5% of beer drinkers might consume 60-80% of the total beer serverd in 24hrs, it does not mean they are responsible for 60-80% of the bar congestion. When they are in the bar during peak periods they cause _exactly_ one person's worth of congestion per person, anything more is _physically_ impossible. Just because a driver might do 1000's of miles on the roads, when there is congestion, they cause _exactly_ one cars worth of congestion, anything more is _physically impossible_. Yet if an BB users consumes 'higher than average' bandwidth, then when they are usinf it during busy periods, they somehow cause way more congestion that one connection can, at least according to you and telco propaganda. [/b]

As has been quoted before, each DOCSIS upload stream is about 2Meg. If you get ten 1 Meg users on one upload stream and they all upload at full speed, then that connection will be fully loaded.

As you may or may not know, TCP/IP requires acknowlegement signals sent back to the server before the server can send the next packet, so if you do get ten 1 Meg users on one DOCSIS upstream, and they are all uploading at maximum speed, then it will slow internet access for everyone else. This to me sounds like a congested connection.

You are priceless. So NTL have got themselves into a total mess finacialy, wiped out 1000s of original shareholder money through bad management, strategic incompetence and acquisitional follies, so we can't expect them to scale their networks properly. And the man most responsible for this state of affairs ? Yep you got, it's the same BK, that you seem to think must have all the answers because he is paid £3500 per day ! Priceless.

As you say NTL got themselves into this mess. Unfortunately,because of the mess, they don't have the money to expand the network. QED.

Having said that, BK is responsible for the mess, so they shouldn't reward him with £3.5K a day.

[erol]

OK I said that was about it from me in this thread but goalposts seemed to have moved a bit, and any pledges from me to 'shut up' are always a little suspect as I can not stop myself sometimes

Originally Posted by ianathuth

The problem with cable in particular is that broadband is running on a system that was not designed specifically for it.

I quite agree. In fact I have said so on many times in the past, such that the above could easily have been a 'quote' from me.

Originally Posted by ianathuth

The planning and infrastructure was based on the model of a customer that does certain things, but that model is no longer valid. Times have changed and current days applications consume far more bandwidth than the planners ever thought.

If commercial organisations make bad decisions and fail to plan properly for the future then that is their look out. The solution to this should not be for customers or the government to 'bail them out' imo.

Originally Posted by ianathuth

The more that todays customers use their available bandwidth, the sooner you get to a state of congestion. If NTL was to implement even higher service tiers then the congestion would arrive much sooner and there is a limit to what cablecos can increase the service tiers to without a very radical change in their infrastructure which they cannot afford.

This is all quite true and correct. At no point did I suggest that a 2mbs service would help any of these problems. All I 'wanted' in terms of this thread is for NTL (and others) to stop saying or implying that a single user can cause any more _congestion_ when things are overloaded than a 'single users worth' of congestion. This is clearly (to me at least) just not possible and yet it is a _constsnt_ refrain from telcos/cablecos and sw companies trying to sell 'solutions', in one form or another and i object to it, strongly.

[erol]

Originally Posted by scastle

As has been quoted before, each DOCSIS upload stream is about 2Meg. If you get ten 1 Meg users on one upload stream and they all upload at full speed, then that connection will be fully loaded.

As you may or may not know, TCP/IP requires acknowlegement signals sent back to the server before the server can send the next packet, so if you do get ten 1 Meg users on one DOCSIS upstream, and they are all uploading at maximum speed, then it will slow internet access for everyone else. This to me sounds like a congested connection.

believe it or not I actualy understand all this. Yes 10 users (ish) can saturate the upstream. When the 11th user comes on all 11 slow down _equally_ and so and so on as users more user come online. So in your senario each of the 11 users is causing 1/11th of the congestion. No more and no less. My problem is with the idea that if 1 of these users use the connection a lot when it is not overloaded, or his/her use happens to be a P2P app, then they suddenly are creating more than this 1/11th (ie the 5% cause 60-80% of congestion). They cause exactly '1 users worth', no more and no less. Similarly 5% of users can only cause 5% of the congestion. Just because of thier usage in non peak times may be high or of what they use it for does NOT mean they suddenly become able to cause more than 'one users worth' of congestion. See my point ?

Originally Posted by scastle

As you say NTL got themselves into this mess. Unfortunately,because of the mess, they don't have the money to expand the network. QED.

Having said that, BK is responsible for the mess, so they shouldn't reward him with £3.5K a day.

It's a free(ish) market jungle out there. If you make bad commerical decisions then you suffer the consequences. I hold no truck with the argument that users (or the government) should some how 'help out' to correct these problems caused by by commercial decsions. Either NTL are in a position where they can offer competetive BB products or not. If they are not then they should be 'allowed to fail and fail fast'*

* another great Isenberg piece btw at
http://www.netparadox.com/fccletter.html

[ian@huth]

Erol. when the 11th user comes online in the example that you have just quoted they do not all slow down equally. It depends on what they are doing.

Can't you see that there can be 100 users on line on a upstream card and they can all be suffering extreme degradation of service. You can then have 5 users come off line (5 p2p users thrashing their connections) and the remaining 95, still doing what they were previously, then don't experience any degradation of service. To me. in my way of thinking, those 5 users were causing all of the degradation. In your way of thinking it would be that the 5 users were causing 5% of the degradation and if they came offline there would still be 95% degradation in service.

[erol]

Originally Posted by ianathuth

Erol. when the 11th user comes online in the example that you have just quoted they do not all slow down equally. It depends on what they are doing.

Can't you see that there can be 100 users on line on a upstream card and they can all be suffering extreme degradation of service. You can then have 5 users come off line (5 p2p users thrashing their connections) and the remaining 95, still doing what they were previously, then don't experience any degradation of service. To me. in my way of thinking, those 5 users were causing all of the degradation. In your way of thinking it would be that the 5 users were causing 5% of the degradation and if they came offline there would still be 95% degradation in service.

So just what are these other 95 people doing then ? None of them are downloading anything? Or uploading anything ? Or watching streaming video or listening to web radio or looking at a flash based website or playing online games or any number of other things (that actualy require data to be sent up and down) ? If they are not doing anything that requires data to be sent up and down then they are not being effected and if they are they are contributing to the congestion. When I pull off the congested motorway I stop contributing my 'one cars worth of congestion' to the problem and I in turn am not effected by the congestion. When I pull back onto the motorway I am effected by the congestion and I contribute my 'one cars worth again'.

So now you can start telling me about 1mbs users that only want to send and recieve data at 500kbs being effected by those that want to use it at full speed but the argument is the same if you think about it (along with the obvious question , why have they got the 1mbs service an not the 600kbs one). Again in analogy terms you could argue that big cars cause more congestion than small ones - and you would be right, but in turn small cars are more able to nip in and out than big cars (or use bikes if you prefer) and thus the effect of the congestion on them is proportionately less that someone in the big car.

Originally Posted by erol

Again in analogy terms you could argue that big cars cause more congestion than small ones - and you would be right, but in turn small cars are more able to nip in and out than big cars (or use bikes if you prefer) and thus the effect of the congestion on them is proportionately less that someone in the big car.

This analogy falls flat in two ways that I can see.

First, have you ever been on a motorway in very heavy traffic with lorries filling two of the three lanes? Mr Warez Monkey is like a 40 ton truck complete with extra trailer, and there are so many of them on the road, that even if I'm in a Smart car I have limited room to manoeuvre around them.

Second, smaller cars are by no means guaranteed to be able to make best use of small gaps in the traffic. Small cars generally have small engines. It's the Jaguars and BMWs that have the power to exploit gaps in the traffic flow. And so at this point the analogy breaks down all together.

A contended service relies on the 20 or so people to whom the same chunk of bandwidth has been sold, not all trying to use it at once. It would be reasonable to say that, in any 24 hour period, my fair usage of the system would therefore amount to 1/20th of the time ... 72 minutes of download, if used in one constant block. This is a very rough sketch of the principle, but hopefully you get the idea. I accept this, because I accept the economics of it. If I want truly limitless service unipeded by other users, I would have to pay a lot more for it.

Now, I don't want a capped or metered service, but I don't want high petrol prices either. I just have to accept that fuel duty is the weapon of choice in a fight to reduce traffic on our congested roads. Similarly, if there is no effective way to beat the warez monkeys, then it is not unreasonable to discuss pricing them out of their illegal activities. This does not have to mean 'all pints are £50' as was suggested earlier; perhaps a pricing scale that only gets punitive with very excessive use.

Yes, the cost of using the service is the same regardless of how heavily you use it, but to make this the basis of a campaign to retain flat-rate broadband internet at all costs completely fails to take aaccount of basic world economics. It just aint like that.

[erol]

Originally Posted by towny

This analogy falls flat in two ways that I can see.

First, have you ever been on a motorway in very heavy traffic with lorries filling two of the three lanes? Mr Warez Monkey is like a 40 ton truck complete with extra trailer, and there are so many of them on the road, that even if I'm in a Smart car I have limited room to manoeuvre around them.

All anlogies have thier limits and can be pushed too far. However the point is in my analogy terms is that the 'size of your car' realtes to the 'size of your connection' (which tier of service you are on - 150k,600k or 1Mbs). I only introduced the idea of different size cars (to match the different sized 'connection') because I felt that was where ian would go next. It was probably a mistake to do so. The point is though that 'Mr Warez Monkey' can not be the equivalent of a 40 ton truck in terms of my analogy, because NTL do not offer 40Mbs connections. Simple as that really.

Originally Posted by towny

Second, smaller cars are by no means guaranteed to be able to make best use of small gaps in the traffic. Small cars generally have small engines. It's the Jaguars and BMWs that have the power to exploit gaps in the traffic flow. And so at this point the analogy breaks down all together.

Like I say all anlogies have their limits. Their real use is to highlight certain points to help in understanding / conceptualising an idea that is trying to be conveyed. However if the person 'listening' does not want to understand then they are bad, because they can allways be attacked because no analogy is a 100% representation. It does not try to be, just to highlight the point.

Maybe a better way to have tried to explain my point would have been to have suggested that a 500kbs user is half as affected by congestion than a 1mbs user, because they can only do half as much to start with.

[snip] *

Originally Posted by towny

Now, I don't want a capped or metered service, but I don't want high petrol prices either. I just have to accept that fuel duty is the weapon of choice in a fight to reduce traffic on our congested roads. Similarly, if there is no effective way to beat the warez monkeys, then it is not unreasonable to discuss pricing them out of their illegal activities. This does not have to mean 'all pints are £50' as was suggested earlier; perhaps a pricing scale that only gets punitive with very excessive use.

I understand you are happy at 'these peoples' illegal activites, but illegal activites should be dealt with by the _law_. You don't stop shop lifting by charging an 'entrance fee' to shopping centres. You stop it by arresting people.

There are already ISP's that charge on usage. To make all ISp's charge on useage would be a mamoth undermining of the potential of the Internet to be useful to people (doing _legal_ things). This is because the incremental cost of _provision_ of the service (IE that cost to NTL that goes up when it is used more is _tiny_ in relation to the fixed cost _of provision_ )

Originally Posted by towny

Yes, the cost of using the service is the same regardless of how heavily you use it, but to make this the basis of a campaign to retain flat-rate broadband internet at all costs completely fails to take aaccount of basic world economics. It just aint like that.

Now you have serious misread and misunderstood me here.

Its not that the cost to the customer is fixed, it's that the cost of _provision_ of the service (ie the cost to NTL to provided it) is almost entirely fixed (which is why the vast majority of net connection around the worls are fixed price, btw.)
This was exactly the argument we used when arguing for flat rate dial up and eventualy Oftel agreed and mandated a wholesale flat rate dial product on BT and flat rate dail up access in the UK was born. If anything the incremental costs on dial up were higher and less prone to halving every 12months (as the incremntal cost of provsion of BB halves) yet still the argument was won and sustainable unmetered dial up was a boon to internet access in the UK

(as an aside soon after it was mandated on BT, the cablecos also introduced it, as unlike anyone else they could have done at any time previously but did not. Remarkable situation with BT driving the competition vs companies that were designed to drive competition with BT)

If you read back my posts Towny I think you will have to agree that I have always been clear that it was 'cost of provision' and not 'cost to customer' that I was tlaking about ? An appolgy is optional

* I was just going to ignore this because it has been covered time and again. Try the link to the anticap site I provided.

[ian@huth]

Originally Posted by erol

There are already ISP's that charge on usage. To make all ISp's charge on useage would be a mamoth undermining of the potential of the Internet to be useful to people (doing _legal_ things). This is because the incremental cost of _provision_ of the service (IE that cost to NTL that goes up when it is used more is _tiny_ in relation to the fixed cost _of provision_ )

It is not as tiny as you may at first think. The more that the service is used the more there is need to balance UBRs. Because it takes so few users, particularly 1Mb p2p users who are thrashing their connections, to saturate the upstream channel of a card, it is so easy for all the users on that card to experience problems. What happens then? Many calls to Customer Services and Technical Support complaining of the degraded service. Each and every one of these calls costs NTL to answer them. This has a knock on effect of increasing the queues on the phone for these departments with increased frustration for both the customers and the agents answering the calls. Increased frustration amongst agents leads to the frustration being spread to subsequent callers to those agents and the likelyhood of increased absenteeism amongst the agents. All these items increase the costs of NTL. Frustrated customers with problems are more prone to say "that's it, Ive had enough" and so the churn rate increases and they lose customers that may be spending quite a lot on all the NTL services but leave them all because of their perception of the service levels of the company.

A last analogy for you :

I was on the M1 the other day and was involved in a massive traffic jam. The eventual cause I found to be a very large lorry spanning two lanes with a police escort and travelling very slow. That one lorry was solely responsible for the congestion.

[erol]

Originally Posted by ianathuth

It is not as tiny as you may at first think. [snip]

Well any calls to NTL that are not on freephone numbers generate income for NTL. As for the rest, if calls to CS causes the costs and impact you are implying then NTL would first be best of fixing their mail servers, their news servers and avoiding things like the balls up 'enforced use of dialer sw' on offnet dial up customers.

A last analogy for you :

I was on the M1 the other day and was involved in a massive traffic jam. The eventual cause I found to be a very large lorry spanning two lanes with a police escort and travelling very slow. That one lorry was solely responsible for the congestion.

As I have already explained , in terms of the way I was using the analogy (car size = connection size) NTL do not allow lorries onto their network at all.

Oh all right I will accept that a single user can cause congestion greater than the size of their connection - by going out and cutting the main cable from the street cabinet to the UBR. There, happy now ?

[ian@huth]

Originally Posted by erol

Well any calls to NTL that are not on freephone numbers generate income for NTL. As for the rest, if calls to CS causes the costs and impact you are implying then NTL would first be best of fixing their mail servers, their news servers and avoiding things like the balls up 'enforced use of dialer sw' on offnet dial up customers.

All calls to NTL concerning service levels generate an element of cost to them. It's a viscous circle in that the more that calls increase, the more it costs and the less there is to sort out the problems causing the calls.

Originally Posted by erol

As I have already explained , in terms of the way I was using the analogy (car size = connection size) NTL do not allow lorries onto their network at all.

Oh all right I will accept that a single user can cause congestion greater than the size of their connection - by going out and cutting the main cable from the street cabinet to the UBR. There, happy now ?

Replace the oversize lorry with a normal family size car that has a drunk at the wheel and is weaving about so much that nobody dare overtake him.

Let's face it, there is no analogy that corresponds to the situation we are trying to prove or disprove. Common sense and knowledge of the system are the answers but a person who refuses to accept any view but his own isn't using these to the full.

Originally Posted by erol

I understand you are happy at 'these peoples' illegal activites, but illegal activites should be dealt with by the _law_. You don't stop shop lifting by charging an 'entrance fee' to shopping centres. You stop it by arresting people.Now you have serious misread and misunderstood me here.

A civilised society relies not just on the force of law to maintain order; it relies very heavily on pressure from other areas to force would-be offenders to think twice about doing something illegal. The law is of course a valid means of dealing with warez monkeys, but in my view a pricing structure that makes their activities less attractive to them is also a fair and reasonable means of dealing with the problem. You and I would both prefer it if this was not so, but such is human nature.

Its not that the cost to the customer is fixed, it's that the cost of _provision_ of the service (ie the cost to NTL to provided it) is almost entirely fixed (which is why the vast majority of net connection around the worls are fixed price, btw.)

I snipped out the bit about reading back through your posts because some of the posts (not just yours) in this thread are so long they are not effective as a means of conveying an opinion. This is a discussion forum, not an essay-writing competition!

However, with due apologies for my loose use of language and terminology, I think you have misunderstood me. Regardless of whether it is cost of provision or cost to customer that is fixed (and I do appreciate the difference), what I am reluctantly advocating is a pricing structure designed to manage usage of the network, not a pricing structure designed to reflect actual costs, whatever those costs may be. As I said in my previous post, this country's tax regime is an example of how pricing can be used to encourage or discourage certain behaviour. I don't like it, but I see no effective alternative. The Law never dealt effectively with widespread copyright infringement pre- the digital age; I see no reason why it should work now, without help.

[erol]

Originally Posted by towny

but in my view a pricing structure that makes their activities less attractive to them is also a fair and reasonable means of dealing with the problem. You and I would both prefer it if this was not so, but such is human nature.

Fair to the paraplegic who uses the internet constantly for voice and video links? Fair to the low income earners ? Fair to those that are 'heavy users' and NOT engaged in illegal activity?

Originally Posted by towny

Regardless of whether it is cost of provision or cost to customer that is fixed (and I do appreciate the difference), what I am reluctantly advocating is a pricing structure designed to manage usage of the network, not a pricing structure designed to reflect actual costs, whatever those costs may be. As I said in my previous post, this country's tax regime is an example of how pricing can be used to encourage or discourage certain behaviour. I don't like it, but I see no effective alternative. The Law never dealt effectively with widespread copyright infringement pre- the digital age; I see no reason why it should work now, without help.

As far as I am concerned the Internet is a vital utility. It is essential to the future health of UK PLC. To price such a 'service of general ecconomic interest' on a what the market will bear basis and not a cost + basis, espcially when it is also a natural monoploy/duopoly would be nothing short of madness. Luckily the government and the EU seem to feel the same.

If the cost of provision was incremantal on usage then customer pricing on such a basis would be fair right an proper. If it is a fixed cost then the customer / user / citizen should have every expectation of being able to accrue the benefits of this (within the law).

Thats my view anyway

Originally Posted by erol

Really? So if every car user that is using his car for an illegal activity was to be removed from the roads there would be far less chance of road congestion would there? Or maybe the more you free up the road, the more cars that then come and fill the void until you are back where you started, regardless of whether the journies are illegal or not ?

60% (approx) of people on the road aren't using their cars illegally, though.

60% of traffic on most networks, higher on some, lower on others, is P2P traffic. We don't know whether it's legal or illegal P2P traffic, but come on, let's not be naieve it's hardly ever likely to be legal. I've prolly only downloaded one or two legal things via WinMX et al.

If 60% of the traffic was removed from the network then yes it would reduce overall network traffic would decrease and there would be at that point more available bandwidth!

Look at it this way, 60% of traffic accross most residential ISPs is P2P. If P2P vanished (which isn't likely!) then there would be 60% less traffic, assuming that userbase stayed the same and network capacity wasn't reduced.

Taking your car analogy again. Leave all the roads as they are, same amount of lanes etc. Now take away 60% of the cars. I imagine that even in rush hour there would be a lot less congestion, sure there would still be some, as there would on the 'net. But a lot less. 60% less, in fact!