Slashdot Mirror
European ISPs Exaggerate Performance; US ISPs Slower But More Honest (itworld.com)
itwbennett writes: New studies of broadband Internet access across Europe and the U.S. published by the European Commission have found that European broadband Internet access providers advertised download speeds of 47.9 Mbps, but only delivered 38.19 Mbps, while U.S. providers delivered more or less what they advertised. But if you want fast fixed-line Internet access, you're still better off in Europe than in the U.S. Average DSL, fiber, and cable Internet speeds in Europe were all ahead of U.S. average speeds, and at lower prices.
"up to"
ISPs can get away with pretty much any speed that way
Ironically, if you want legal content for your blazing fast internet connection you are much better off in the US. The european versions of HBO and Netflix is not even close to make an high speed internet connection interesting, most other services are not even available..
I've been saying this for YEARS. Asia is even worse about lying. South Korea ISP's says they deliver 1Gbps. BALONEY. I used to live there. Their data rates were comparable for what I get here in the US.
I'm getting billed around 18€ monthly for 15/15Mbps internet and VOIP. I've never had any big issues with the advertised speed (Upload is actually 19Mbps sustained for some reason, means I can max out upload and download simultaneously and using fq_codel to track connections it's still manageable on the latency side).
More importantly, I use 1 - 1.5~TB of bandwidth a month, and have done so for the last 3 years. That's what I'd rather know. How many of these ISPs have monthly bandwidth limitations and how many of them traffic shape you into oblivion after you hit that cap.
What's the point in a nice cheap 100Mbit connection if the ISP throttles you after a few days?
The U.S. has tariffed rates.
If they don't meet specs on their connection compared to what they contracted the connection at, then they will be crucified.
This is why, when I lived in an apartment 10 feet too far away from the LATE, the wouldn't give me DSL, and would only offer ADSL. This was in Silicon Valley, where presumably, we'd have good Internet connectivity. They simply weren't willing to risk the legal ramifications, should the sell it to me, and it be 1% too slow, and me taking them to court over it, and them losing their regulatory approval everywhere because of it.
It looks like Europe is either under-regulated, or under-litigious, compared to the U.S..
The slow U.S. rollout of higher speeds has more to do with 20 year amortization on equipment, which is standard practice in the telecom industry, and the fact that you only have to be better than the competition to lock up all the consumers in a given market, and there is little competition.
That, and the U.S. is *big* and sparsely populated for the most part, and Europeans have absolutely no clue at the distances involved, which is why they totally fail on the "public transportation in the U.S." and "Internet access in the U.S." and "Taxi service in the U.S." arguments (you can get a Lyft in Alta, UT -- population 389 -- but if you expect a taxi, don't hold your breath, or expect to pay for it to come out from Salt Lake).
Particularly since many of the places I see that get talked about for having super-fast Internet seem to post speedtest results for a short distance, on network. That's not really useful because that can just mean that you basically have a big WAN with fast access to your own stuff, but no backhaul to support it. To really have a connection that you can claim gets the speeds advertised, you need to be seeing that speed to a server that is off of the ISP's network, and a few hundred miles/km away in another state/country. If you can get your speeds with tests like that, then you are actually getting what is advertised. If you see great results on the ISP's speedtest server that is 10 miles away but crap to everywhere else, they've sold you a fast link with no backhaul.
I'm real happy with my connection for that reason. It's 300mbit for $100/month but it really gets that. I see those speeds not just to my ISP's server, but to servers all over the US. Steam downloads go at like 40MB+/sec. So it is expensive to an extent, but I really get the speed I pay for, and I get it to anywhere that can handle it (when you start to talk fast lines the other end is the problem sometimes).
A fast last-mile means nothing if there isn't sufficient backhaul and peering at all level to support it.
Mine was honest.
I'm in a crappy DSL area and they said as much. They said give the best they could but it'd probably be around 4mbit/s. It's more like 4.5 sometimes 5.5, probably depending on the level of crosstalk from neighbours.
So mine was honest, but then they lost all my customer data to thieves because it's Talk Talk.
You win some, you lose some.
SJW n. One who posts facts.
I have lived in the US and in Sweden both in fairly newly built areas. I have also worked in various places related to IT and from experience the internet here in Sweden is much more reliable and faster than that of the US (Comcast). In the US I don't remember ever getting the speed I paid for and it was not unusual for the internet to be terribly slow at times or be completely down. Here in Sweden I currently pay for 100/100 internet and regularly get 95/95 or so which could be because of my own router/wireless etc. On good days it goes over 100/100 which is surprising :D recently I saw one area of Sweden where you could go from 100/10 speeds to 1000/1000 for an extra $20 which is super cheap.
http://www.digitaltrends.com/c...
I am interested to know the nesty of the South Korean ISPs
Anyone from South Korea? Care to elaborate?
They compare European speeds collected in October 2014 using their own methodology with a report published by the US FCC in April 2014. The US data was probably collected a year earlier than the European data and likely measured the speed differently.
The report itself notes that speeds in Europe increase by about 25% between 2013 and 2014:
The average download speed across all countries was 38.19Mbps during peak hours, a 25.7% increase from the previous year, slightly lower than the 39.69Mbps observed during the 24-hour measurement period. Average download speeds have therefore increased by nearly 10Mbps since October 2013, when the figures were 30.37Mbps and 31.72Mbps during the peak and 24-hour periods respectively.
>Internet speeds in Europe were all ahead of U.S. average speeds, and at lower prices.
But is that based on a simple exchange rate or as compared to average disposable income? Comparing prices from different economies/countries is not simple.
Answer (maybe): First they only considered prices in CA, NY, and CO. With the former two being likely some of the (if not the) highest pricing in the whole HUGE USA.
Then, they supposedly used a percentage of income instead of just "price", but it was TOTAL income, not "disposable" and presumably not post-taxed income. And there is no mention of if they included taxes on these services or not.
Throwing numbers at the above population density claim, the European Union has an average population density of 112/km^2. The United States has an average population density of 35/km^2.
If regulators like the FCC or FTC required that ISPs pro-rated their pricing based on the bandwidth they actually delivered, instead of what they claim to have provisioned, the 'market' would be held accountable. (On the other hand, rural populations would be out of luck without subsidies, which are in place in the U.S.)
I live in rural Am'rica, and CenturyLink is charging me for 10 Mbps ADSL. My Centurylink branded Xyxel modem reports 640 Kbps service... If my cost was pro-rated, I wouldn't mind so much, but paying Centurylink (formerly Qwest, formerly US West) for 'broadband' which isn't fast enough to support high fidelity audio is ridiculous.
In the interest of full disclosure: After years of complaining about the speed, a CS rep finally ran a remote diagnostic test which reported wiring issues, and a tech is scheduled to come out and investigate...
Captcha= 'Astatine' (Element 85, byproduct of decaying radioactive Bismuth)
While that's how it was advertised, it is unfortunately more than that.
Change is certain; progress is not obligatory.
When will we see the establishment of mandatory MINIMUM speed guarantees (based on 5 9's uptime - 99.999%)?
Speed monitoring is trivial. Let's all advocate for this basic change.
*** Don't be dull.***
I blame the Italians!
What do the speed-comparisons look like if you compare urban areas (say, an urban/suburban area with over 50,000 people in it) in the US to urban areas in Europe?
What about rural areas to rural areas, counting only those areas within, say, 30 minutes of non-rush-hour drive-time of an urban area.
What about more distant rural areas?
--
Why "30 minutes of drive time"?
If there are roads that can get you to a city in 30 minutes or less, I would expect there is a not-horrendously-expensive way* to run fiber to your general vicinity and from there a path for decent wired, fixed-wireless, or mobile internet service. I won't assume the same if you are on the other side of difficult-to-cross terrain or if you are a long, long way away from a populated area.
*Assuming of course that regulatory burdens or private landowners who ether refuse access outright or who see the fiber-operator as their personal gold mine don't make running fiber too expensive to put in. I don't know how it is in Europe but in most of America if the local or state governments sign off on running fiber from "point A to point B through path C" the affected landowners will be paid for an easement (if one does not already exist) but they won't be able to say "no" nor will they be able to demand exorbitant payments for new easements.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
The local small-town-midwest cable company, Mediacom, has always been honest about speeds. In fact, you always get slightly faster speed than what your supposed limit is.
For example if you had a 10/1 connection, you got something closer to 15/1.5. Now with a 50/5 connection it's more like a 53/7.
What you don't get is fairly priced TV services. Mediacom's internet and phone services are a good bang for the buck, the TV service....not so much. The situation locally is such that one of the Satellite services has a bundle where the internet is supplied by Mediacom. The other satellite service bundles with the local crap-DSL provider, Frontier. With them you're lucky if they offer you a 6Mbps connection.
Can't empathise with people who aren't getting their full 48Mb ... I'd happily take an over advertised 38Mb over my crappy 2Mb
One massive problem with the SamKnows study linked is the fact that the sample sizes are ridiculously small(less than 200 testers, broken down as 136 Fibre, 23 cable and 34 xDSL in Sweden, for example, which is nowhere near representative of actual distribution etc.). There's also no differentiation between various fibre methods. For example, in Sweden, the most common variant is FTTP+ethernet, while in the UK, FTTC+VDSL is very common, yet in this test they are lumped together, which helps skew the numbers for fibre overall.
You can get some insight into your own performance by joining Samknows- a worldwide survey of internet performance from various ISPs ( https://www.samknows.com/ ). You'll receive a monthly report with data & charts that shows your up/down speeds and averages for every day that month as well as latency, packet loss, and disentropy for your setup. Actually I don't think disentropy has been discovered until this moment.
Additionally as a subscriber, you can see the big picture data at their web site as discussed in TFA. Joining is free and you get a 'whitebox' to connect to your system which reports to Samknows in UK. You know it's safe because it's white and they promise not to spy on you.
...omphaloskepsis often...
Why as that moderated downward when it is true? We keep electing anti-tech local candidates.
For a while I ran our network out of 2 business standard ADSL lines from a reputable ISP. I could click on any speed test site and it would show me the speed almost meeting the "up to" level. But if you want to download something large (from far away or close by) the rate would start good, then drop and drop, and then just pulse along at an average of one fifth of the "up to" rate. That's 2 separate systems on 2 phone lines.
I think the basic problem is you're an idiot. You couldn't even be bothered to read the rest of my post, instead making snarky comments because my argument doesn't address al the points in the precise order you want them addressed in. Come back when you've grown up a little and are interested in having a proper debate and discussion among adults.
And you are ignorant.
You made a bunch of incorrect statements about population density, and you made a bunch of incorrect statements about population density by state with an exception for Alaska, even though 6 *other* states have the same population density as Alaska. I pointed out that the population density was *vastly* biased by averaging the numbers over about 14 states, several of which have 1000+ people per square mile (that's about 20X the average population density of Sweden, which is 54/square mile).
And yeah, were the population is incredibly dense (as in: >800 apartments in a single high rise building in New York), there's good internet connectivity.
You ignored the major premise that the infrastructure is not government owned, as it generally is in Sweden (many municipalities own their own fiber optic infrastructures, something which generally does not happen in the U.S., because the cable and telephone companies take the cities to court and charge unfair competition.
Basically, you pretty much ignored anything which disagreed with your premise that U.S. Internet connectivity sucks in all but two states, 48 U.S. states suck, and that economies of scale should apply equally everywhere, due to average population density over the entire U.S. (except Alaska, of course).
Educate yourself.
A part of Sweden called Norrland is about the same size as Kansas and has half the amount of people living there, and I still have 1 Gbps in my summer cabin there. But avarage population density has little to do with it, I think Svalbard would win that category though.. :-)
Backbone investment to remote places has just been very high priority in Sweden.
A network connection has at least a dozen different parameters that affect performance, yet again and again we see articles which seize upon one of them (or a bogus measurement of them) and term alarmist: "X is behind Y in Internet speed! This is horrible!"
In truth, network connections have minimum different capacities, CIRs (maybe; some ISPs offer them and others do not), latency, and jitter -- each of which is usually different depending upon where the user is going. There is no single number that can possibly capture all of this. But journalists, lobbyists, and politicians attempt to mislead by picking a dismal sounding number and claiming that it is in unacceptable.
The next time you hear anyone attempting to describe the quality or "speed" of an Internet connection with a single number, turn the page. Wittingly or unwittingly, he or she is misleading you.
I've stayed at my brother's place when his connection to Comcast High-Speed Internet (originally) had 50 mbps download speeds, now 100 mbps download speeds. Using Speedtest.net, I was getting around 44-47 mbps under the old setup and 90-92 mbps under the new setup.
But now, Comcast is preparing to roll out DOCSIS 3.1 service by 2017; they're converting all of their HD channels to MPEG-4 compression to free up bandwidth space to allow DOCSIS 3.1 service. In theory, DOCSIS 3.1 is capable of around 1 gigabit download speeds; just how fast Comcast will the new service be is still a major unknown, though I think at least 350-500 mbps download speeds is possible.
Mine doesn't exaggerate but you're not getting much, down 10mb up 1mb. No one is honest and everyone over sells. It's called a profitable business model. I'm just unfortunate to be in one of the countries with the slowest broadband running a lot of it over old copper and new fiber installations is moving at snail speed.