For me, here is the key quote that proves this is not about offsetting bandwidth congestion and infrastructure costs:
Congestion can happen even on networks with tremendous bandwidth, but
consumption pricing doesnât generally care about congestion (if it did,
ISPs could exempt all traffic in the middle of the night, for instance,
when congestion is generally absent).
If this is about reducing congestion, then we should be thinking about dynamic pricing for usage. Otherwise, they are lying.
Cost is the determinative factor in price only for public utilities and other socialist horrors. Here in the land of freedom, you will pay what the market will bear, and thatâs proportionate to how much value you receive.
If you only send the occasional email, any price at all might be too much - youâll just send them at work, or the library. If you regularly upload full-length movies, youâd better have a fiber optic connection, and youâll happily pay extra for it.
So itâs not that âbandwidth hogsâ are harmful; itâs that they would be willing to pay more than most of us. The problem is that TWC (or whoever) then tries to market the Gb service to everybody, at the Gb price. Then they want to not actually deliver that speed, because youâll never notice. And then the guy who really uses what he pays for is a âhog.â
This has been happening in Canada for at about a year, after being discussed for about another year. Canadians, being Canadian, whined a bit, but didnât want to rock the boat too much. We are now paying $5-$10/month more for internet usage than we used to in order to cover overages.
There is much to love about Canada, but the love of questioning and pushing back against authority is not one of them.
Speak for yourself, eh. Besides, in many other countries, people grouse and complain, sometimes very loudly, but they never seem to throw the bums out.
Anyways, here at the other end of the country, in the Atlantic Provinces, Aliant doesnât have caps. They never have had 'em. Bandwidth hogs? They donât care.
As I recall from microeconomics 101, this is âdifferential pricingâ which (in theory) only a monopoly can engage inâŚ
Well, âIn theoryâ is usually pretty unimaginative in economics. You donât need a monopoly if you have a reasonable-sounding explanation. âPeople who use more pay more,â triggers the âYeah, that sounds rightâ response.
The article is old, but as true as ever!
The big ISPs are all entrenched in the entertainment business. As soon as they realized the Internet would compete with cable, they clamped down on supply.
Thanks to our generosity in setting them up with monopoly access to our homes, they are now the fox guarding the Internet hen-house.
Fair enough. Maybe I should move out East. For the bandwidth (and itâs nice out there). If I could only find a jobâŚ
There may be a bunch of things wrong with Virgin in the UK, but if you pay for their top tier internet then you get all 220MB and thereâs no caps.
Which now means Iâm looking for a home router that can cope with 200MB/s and still do traffic management without being overloaded.
A few months ago, I in the Vangcouver airport, going from one flight to another. I stopped at a Starbucks to pick up an espresso.
âWhere are you from,â the lady at the cash asked.
- The other end of the country, I live on the East Coast.
- Oo! Itâs cold there!
- Maybe so but the houses cost one tenth of what they cost here.
Thereâs always pogey, eh.
What on Earth are you doing that needs 200MB? My connection is one eightieth of that, 100Mb/s, and is perfectly adequate. Once when our business ISP made a mistake, our web server got 200MB/s for a day - more than the gigabit Ethernet could cope with - and I amused myself downloading every resource we could possibly want into our server. I had an entire RAID volume filled by lunchtime, thanks to the bandwidth generosity of IBM and SUN, before they cottoned on.
Ack I wrote megabytes instead of megabits didnât I? I always get those units the wrong way round. Mea culpa.
Even so, we donât really need two hundred megabits of bandwidth either, even with five geeks in the house. Itâs nice all the same though 
If you are going north of 100MB/s and doing any actual packet munging, you probably donât want a âhomeâ router. The cheap ones will fall over and die. The expensive ones will bristle with more antennas than a stereotypical alien spiderbot and promise fancy wireless featues; but still fall over and die when it comes to actual packet crunching.
Your best bet is either a low power/multi NIC real computer(Soekris was the go-to back when I was looking, not sure what the state of the market is now) or a used low-end-of-enterprise device or a new item of the same family aimed at the âbranch officeâ use case. Iâve heard good things about Ubiquitiâs âEdgerouterâ line, though I havenât had time to really play with mine yet, so that is secondhand.
Basically, if the ârouterâ and the âaccess pointâ are in the same plastic box, itâs probably not going to be able to take it. Get a real router and add APs separately(you can get POE units with SNMP and some vague gestures toward manageability surprisingly cheaply from time to time; or you can run home âroutersâ as dumb APs, either way.
Okay, maybe Iâm missing something here, but my salary (as is most peopleâs) is based on what the market will bear. Iâve carefully selected the employer who I think values my skills the most. Thereâs lots to complain about with respect to ISPs, but complaining that they are engaged in differential pricing is pretty close to a complaint about them being involved in business.
What we can hope for someday is enough competition that the price is driven down to something equivalent to cost of goods. But even there Iâm not certain that weâre that much above it right now. Most smaller ISPs end up charging something a comparable price to the larger players, and yet they still manage to go bankrupt fairly regularly. And Iâm not certain thereâs a big player in the ISP market who isnât up to their eyeballs in debt.
I donât think being an ISP is the license to print money that this article suggests. It can certainly be profitable, but not more so than many other lines of business.
Still, what we really need is a few thousand more local ISPs like Chattanoogaâs efforts to bring decent competition.
I donât really buy this argument. Simply because the providerâs costs are largely fixed doesnât mean that all the customerâs costs should be fixed as well.
The fact that those that value bandwidth more pay more for more bandwidth means that those who cannot afford more, or do not need more donât have to pay as much - which seems equitable to me.
Big B is the big one. I thought you had actually signed up for a business fibre line, because there are people who do that. Because they can.
Keep in mind that in this context âwhat the market will bearâ equates to the least amount possible until someone they can pay less is found, or better, can be outsourced to for much less.
Unless, of course, youâre in the exec club which keeps paying each other exorbitant salaries - above what the market can actually bear as they routinely lay off workers to access that needed capital. But itâs an unspoken agreement between the execs that they represent ultimate value (when in some cases a coin toss provides the same results) and are actually those that could most easily be replaced for those making a fraction and doing a better job. Itâs a racket.
As much as I hate to defend data caps, most ISPs arenât big enough to get free peering, and do pay for upstream bandwidth. Not to mention that there are ongoing upkeep costs on fixed infrastructure that go up if you have more fixed infrastructure (which you need if youâre carrying more data), or the fact that in certain circumstances free data for certain services actually makes sense, such as if the free services are hosted on the same ISP or otherwise cheaply linked, bypassing upstream bandwidth costs (this is common in Australia - most ISPs host free download mirrors that host exactly the same content at their own expense). Admittedly this would be far more of a problem in Australia than in the much larger market of the US, but they can be justified in at least some circumstances even with fixed line based infrastructure.
