Caps: Innovation Killer? > Physically impossible

Physically impossible

Ummm, I was referring to physical deterioration, not performance deterioration. All well-used networks have periodic congestion (though the ISPs have not proven yet that congestion is a problem on their networks), but that doesn't cause the network infrastructure to become "used up". The point: it doesn't cost them any more to deliver 100 GB than it does to deliver 1 GB and the traffic, no matter how much there is, doesn't cause the cables/interfaces/routers/switches any more wear and tear than if there were no traffic at all.

Some things wear out with use--the more you use them, the more wear and tear... you have to repair or replace them, which costs money. Some things don't wear out with use; use them as much as you want--there is no wear and tear. Network traffic doesn't increase the wear and tear on network equipment; the only thing that "damages" network equipment during normal operation is the electrical power used to keep it up and running in the first place--at some point this will cause failure, but likely not before something is scheduled to be replaced due to upgrades (to provide faster equipment). So, the point is that no matter how much bandwidth you're using, the cost to the ISP for delivery is the same.

Wow... Can't argue with that "logic"

/sarcasm

When you continue to miss the point, why bother. Try saying something relevant, or just say nothing at all.

The point is irrelevant. Physical degradation has nothing to do with networks. This is like saying if the water pipe or electrical wire degrades when I use it excessively, then I should pay more.

Usage and capacity degradation is the issue.
--
"Believe only half of what you see and nothing that you hear." - Dinah Craik

No, the issue is money. (As I said... twice**, there is no physical degradation.)

My point is that (and I'll use Comcast as the example ISP) Comcast designs and builds a network segment to support, for example, 1,000 customers. The cost of maintaining equipment (failures and upgrades) is built-in to the pricing (the "cost of doing business"). Comcast increases the customer base and the tier speeds such that the segment is no longer supporting 1,000 customers but 5,000 or even 10,000 (over-subscribing to the max). Naturally, this will tend to slow down each and every customer (during "heavy traffic" periods, when everyone is "on"), preventing them from getting the service that they're paying for. So, instead of upgrading the network to support the larger number of users at the higher speeds (introduced for nothing more than as advertising to get new customers), they implement ways to get more money from their customers and/or ways to slow down usage (contrary to what they sold their customers originally). There's nothing wrong with the network--it works just as well as it ever has (according to how it was designed); it doesn't cost any more to run it now than it did when it was built. Enterprise class routers (and other network hardware) are designed to run at 100%, 24/7--they don't fail because they have to handle more traffic than they used to when they were installed. It doesn't cost Comcast any more to deliver 1 GB than it does to deliver 1 byte--the cost of running a network is not dependent on how much traffic there is. You turn it on, or you don't; of course, there's no point to having it if you don't turn it on. (Yes, there is an increase in electrical power usage for additional traffic, but it's negligible as far as cost goes.)

If Comcast starts charging you more based on your using the network more (metered billing, which is, no doubt, next on their list of things to do), then they're charging you for something that you've already paid for, aka "rip-off". And it's not like they're going to charge you less if you use it less, giving you a break for having modest needs. They simply want more money in their pockets and less in yours, without having to actually do anything to justify your having to pay them more.

(** And, now, three times.)

You forget two critical points to network design and usage.

1 - The over subscription model that every ISP uses at the metro, core and edge. There is no such thing as "dedicated bandwidth"
2 - The ever increasing bandwidth / customer that happens year over year with richer content, increased speed tiers (many times without an increase in bill), and more applications.

You are correct that it does not cost more now than when it was built, but the fact that you are missing is it is constantly being built to keep up with not only new customer growth, but existing customer usage growth.
--
"Believe only half of what you see and nothing that you hear." - Dinah Craik

Actually, I haven't forgotten either of those 2 points. The first I've already covered elsewhere in this thread (as have others--your statement is correct, by the way). The second is what every customer is already paying for when they pay their monthly bill now.

I've said this before, too, but let's say it again (since it doesn't seem to be clear yet): the only justification (as far as I'm concerned) for charging more based on higher usage is if there is physical deterioration of the network due to more usage of the network. Simply using more bandwidth does not justify higher pricing. Customers don't pay for bandwidth; they pay for time, as in, it takes less time to complete a task if your speed is higher. Customers pay for access. Metering works for electricity--the ISPs' favorite comparison to justify metering, because electricity gets used up and has to be generated, which costs more--we don't pay simply for the delivery of electricity, we pay for the production of electricity by way of what we use. This doesn't work for Internet access except as a means for charging customers more for something they've already paid for, namely, the network and its maintenance. If ISPs want to "manage" users who consume more than others, then they should go to a priority model: the more you use, the lower your priority. If the network isn't busy, then it won't matter; if the network is busy, then it matters--you wait longer for your tasks (downloads, uploads, page loads, whatever) to complete if your priority is lower (because you've been a "hog"... whatever that means). And if the ISPs truly wanted a well-managed network, then they would abandon speed tiers (aka speed caps), too, and simply sell higher priorities to those who are concerned about faster deliveries (when the network is "congested"). The longer it takes for a user's task to complete, the more likely it is to interfere with another user's task(s) which starts later. But I think ISPs see this approach as meaning less profit, which I don't think is the case--they could spend less money on upgrades because the network would be seen as performing adequately over a longer time-frame (less revenue, maybe, but less expenditure, too).

The issue is money... more of it for the ISPs while they do nothing to improve their service for the customers who are paying for that service.

Glad we agree on point 1. There are still people in this world that think DSL is "dedicated Internet bandwidth".

On the second point, yes we pay for over subscription, but we also expect the network capacity to be there when we want it. When Mb/Sub usage increases the oversubscription ratio needs to be lowered to address capacity. This is where the cost increases to upgrade the network to lower the oversubscription ratios.

Usage does = cost... And why ISPs charge / Mb in the commercial world. Broadband speeds these days are much more than commercial speeds were when the major carriers switch to the 95%ile billing model.
--
"Believe only half of what you see and nothing that you hear." - Dinah Craik

It's interesting... on the one hand we all acknowledge that our Internet access (from our chosen ISP) is a shared resource--we each pay our (over-)subscription for our portion and use whatever is available as needed/desired (first come, first served), so we're all effectively subsidizing each other's access. We, the customers, pay for the network. Companies and investors may pay to get a network started, but the customers pay many times over whatever was used to start-up the network. One would expect the company to actually take whatever amount of funds is needed from those payments to keep the network maintained and upgraded so as to provide the customer experience that they advertise: such and such a speed to enable such and such activities for everyone in their ever increasing customer base. (I'll ignore the fact that the advertised speed increases are little more than a ploy to make an ISP's network look "better" than the competition's as opposed to being based on any real need of the customer to go 20mbps instead of only 10mbps--though, yes, there are some who do "need" that.)

Then, someone--usually a representative, or several, of some ISP--comes along and says, "Gee, metering is the future."

These are 2 diametrically opposed ideas. We already pay for the network to be there and available such that it offer a certain level of performance on a shared basis, and then someone wants to charge more on the basis that it isn't really shared at all. Some will agree that those who use more should pay more, but the ISP seems to ignore that those who use less should pay less. Where's the "economy" tier? Where's the $9.95/mo for using less than 5 GB (or whatever reasonable amount for "low" usage)? The ISPs aren't interested in saving customers money; they just want more money for their coffers, and not so they can improve and upgrade the network, but to buy into bigger and fancier headquarters (a la Comcast) and give themselves bigger bonuses. We pay for it, so we should get some benefit from those payments. They're making a good profit as is... better than just "good" actually. Metered Internet access, as conceived by ISPs, is a rip-off.

Yes, their networks need to be upgraded periodically... and we've already paid for it.