Uncle Sam wants YOU to test your broadband connection

[Mustaine]

OMG, serious......? ? ?

Skuzzy, I'd REALLY like to hear your thoughts on the original topic.


Would YOU let one of those devices in your home?

Would YOU recommend anyone allows that device in their home?


A simple Yes or No, I don't want or need explanation on your thoughts.

Thanks

[Skuzzy]

I strongly agree with the rest you wrote, but not this:

Think of TWC hard cap trials (on top of existing soft caps), Comcast throttling torrent under pretense of QoS, lots of other examples.

Data throttling has nothing to do with the physical speed of your local connection.  That speed is constant.

OMG, serious......? ? ?

Skuzzy, I'd REALLY like to hear your thoughts on the original topic.

Would YOU let one of those devices in your home?

Would YOU recommend anyone allows that device in their home?

A simple Yes or No, I don't want or need explanation on your thoughts.

Thanks

No, I would not allow it.  No I would not recommend it.

[Eagler]

I think Orwell was off by about 30 years ... never thought it was possible until about 7 years ago - when the media showed their true colors.

[2bighorn]

Data throttling has nothing to do with the physical speed of your local connection.  That speed is constant.

But that's not the issue. It's about what ISP is selling to you (ie broadband/wideband) and what you actually get.

[Skuzzy]

But that's not the issue. It's about what ISP is selling to you (ie broadband/wideband) and what you actually get.

That is just it.  The ISP is not being dishonest.  They are not juggling anyone's actual connection speed.  Again, no one can guarantee actual data throughputs.  There are far too many variables involved to do that.

[Die Hard]

I think Orwell was off by about 30 years ... never thought it was possible until about 7 years ago - when the media showed their true colors.

[2bighorn]

The ISP is not being dishonest.

Since when?
I can give you hundreds of examples.

They are not juggling anyone's actual connection speed.

Yes they do. Bad signal, too many splitters, overloaded nodes, all examples when ISP know about the issue but doing nothing to fix it.
Example: http://www.dslreports.com/forum/r23199119-What-the-heck-is-going-on-in-SoCal-Part-3

Again, no one can guarantee actual data throughputs.  There are far too many variables involved to do that.

Almost every business account gets SLA with guarantees including data throughput. You're not telling me they are physically that different?

NOTE
When I'm talking about data throughput I go by ISPs definition, using their own speedtest servers.

[Skuzzy]

Those guarantees, for businesses, are for their local connection and not for Internet wide throughput.  No one can guarantee you will get the connection speed data throughput from a server on the other side of the world.  They would be insane to do so.

Physical issues with a connection is not an intentional ploy to reduce speed.  As to whether or not they actually fix it, is a support issue.

Data throughput and the physical speed of the connection are physically separate items.  On a properly maintained connection, you will always get 100% of the connection speed in data throughput to the first router.  You can bank on that, unless there is maintenance being done or any other normal support operation.

[eagl]

Skuzzy,

I know some guys who live in areas where the ISP oversold their local connection.  I also know some guys who got burned with comcast injecting false data or simply discarding packets to cause resends, to throttle back a data transfer to well below their advertised sustained download speed, even when no bottleneck (local or otherwise) existed.

THAT is dishonest, and you're saying it doesn't and didn't happen.  Nonsense, and you know it.  Some ISPs throttle locally when there is no real bottleneck, and some ISPs oversell local connections to the point where they cannot provide advertised performance on their own local subnets.  That is a problem, and I've had to deal with oversold local subnets a couple of times.  Very irritating.

[Skuzzy]

Again, there is a physical difference between the connection speed and the data throughput.  No matter how oversold a connection is, the actual clock/speed on the connection does not vary.  I guarantee the download speed, over your immediate connection, is never throttled and will always deliver 1:1 speed to data throughput on the immediate connection.

Data throttling techniques impact the data throughput, not the speed of the connection.  If you measured the actual bit rate of a single packet of data coming to your computer, it will be running at the connection speed a user purchased.

I am not making a statement as to how ethical or unethical anything is.  I am stating a simple fact.

Now, you want to get into marketing speak?  99% of the users have no idea how it all works.  99% of the users do not understand the difference between connection speed and data throughput.  ISP's are not being dishonest when they state your connection speed will be 50Mb/sec.  Are they being unethical when they cannot deliver the data throughput on a 1:1 ratio with the connection speed?  That is a can of worms.  From what point on the Internet, to your computer, should they be held accountable?

By the way, every ISP oversells thier connections.  They have to in order to be able to be competitive.  They have always oversold thier connections.  Some, more than others.  End user usage habits have changed though.  Now it is more and more difficult to do that without having problems.  ISP's are left with two solutions.  1) Get rid of the abusers or 2) Increase the prices to everyone so they can reduce how much they oversell connections.

Moving on; No single ISP has absolute control, on an end-to-end basis, for every site on the Internet.  8 times out of 10, most problems I see are not in the local ISP's network, but in the Tier 1 provider they are connected to.

As Tier 1 providers dwindle each year it is getting so there is little competition.  The competition that drove the Tier 1 providers to actually do what they are supposed to do.  Getting rid of competition has been an ongoing effort between the cable and phone companies.  Users happily signup for those services without realizing they are killing the Internet in the U.S.  Once there is nothing left but the phone and cable companies, and all the independents are gone, everyone will reap those just rewards.

So excuse me if I appear to have little sympathy for people who complain about a service, but are not willing to drop the service and move on.  Oh, you have no more service providers in your area now?  Gee, wonder how that happened.

[2bighorn]

Again, there is a physical difference between the connection speed and the data throughput.  No matter how oversold a connection is, the actual clock/speed on the connection does not vary.  I guarantee the download speed, over your immediate connection, is never throttled and will always deliver 1:1 speed to data throughput on the immediate connection.

That's really splitting the hair, besides, whenever you'd have any channel bonding (docsis 3, bonded T1s, etc) and a channel fallout, you'd cut your speed (just one example).

Also, consumer is not buying network timing/clock, but data throughput, and blaming the consumer for not knowing technical details is really a bit too much. Kinda like expecting every driver to know their car into smallest technical detail.

[Skuzzy]

It is not splitting the hair, and you keep bringing up technical problems as if the ISP does it on purpose.  Sure any number of technical problems can cause signal degradation to complete loss of signal.  That is going to happen.

I am not splitting hairs here.  It is like buying a car with a 500HP engine and expecting the car to 0-60 in 4 seconds flat.  The automobile company advertises the 500HP, but they do not say how fast it really is.  Is that deceptive?  

There is a physical different between data throughput and actual connection speed.  Guess what?  No matter what the actual speed is, you will never get 100% of that speed in data throughput due to protocol overhead.  Is that deceptive?

Just because you do not like the facts, does not make them any less true. And again, how far up the line do you hold the ISP responsible for the data throughput?

Your immediate connection (yes, I am assuming the connection is actually working correctly) is running as close to 1:1 data throughput to connection speed as it can be.  It always does.  Prove me wrong.  Show me a single data packet that does not come into your computer at the speed of the connection you pay for.  Anyone.

Also, consumer is not buying network timing/clock, but data throughput, and blaming the consumer for not knowing technical details is really a bit too much. Kinda like expecting every driver to know their car into smallest technical detail.

The consumer is ignorant.  I looked all over Verizon's WEB site (whom I have no love for) and I'll be damned if I can find any claims of data throughput.  They sell bandwidth, as most ISP's (I say most because I do not want to waste time on looking at all of them) do.  You assume that bandwidth actually equates to some physical data throughput.  That is an assumption which is not correct.  Whose fault is that?


Look, I do not really care about this stuff anymore, but I do care when people jump off the deep end without having all the facts.  All I am doing is presenting those facts.  No matter how you twist it, or try to get around it, the speed of the connection and the data throughputs are two different measures.  Until that is accepted, people will continue to make the wrong choices as to how to handle this.

The first question to ask is, where does the ISP's responsibility begin and end as it pertains to the quality of the Internet connection.  If you are going to try and govern the Internet, you better hope like heck someone is smart enough to ask the right questions are you are going to end up with a massive mess on your hands which we will all pay for.

Unfortunately, I am afraid there are people just like you and eagl (nothing personal guys) sitting on that panel which will decide things on behalf of the consumer, which will invariably screw the consumer.  None of you want to know what the real problems are.  You are just looking for a scape goat.  The ISP is the easy target.

You have made that clear by continually, and incorrectly, generalizing an aspect of service, in light of the facts, and avoiding the real questions that need to be asked and answered.

[2bighorn]

There is a physical different between data throughput and actual connection speed.

I tried earlier to understand that, but still, I'm not sure what you mean with that. In networking, speed was always throughput (bandwidth), and it still is.

1Gbps connection -> 1Gbps bandwidth -> 1Gbps throughput

Just because you do not like the facts, does not make them any less true. And again, how far up the line do you hold the ISP responsible for the data throughput?

How far? Until I leave their network. Even then, some things are still in ISP control through peering agreements.

Your immediate connection (yes, I am assuming the connection is actually working correctly) is running as close to 1:1 data throughput to connection speed as it can be.  It always does.  Prove me wrong.  Show me a single data packet that does not come into your computer at the speed of the connection you pay for.  Anyone.

Depends on definition of "speed".
In networking speed is measured with how much data you pass per unit of time (bitrate), and I go by that definition (as most other people go).

Look, I do not really care about this stuff anymore, but I do care when people jump off the deep end without having all the facts.  All I am doing is presenting those facts.  No matter how you twist it, or try to get around it, the speed of the connection and the data throughputs are two different measures.  Until that is accepted, people will continue to make the wrong choices as to how to handle this.

OK, maybe I'm wrong. If so, correct me.
Let me ask you
a) How we measure bandwidth?
b) How we measure data throughput?

The first question to ask is, where does the ISP's responsibility begin and end as it pertains to the quality of the Internet connection.  If you are going to try and govern the Internet, you better hope like heck someone is smart enough to ask the right questions are you are going to end up with a massive mess on your hands which we will all pay for.

Unfortunately, I am afraid there are people just like you and eagl (nothing personal guys) sitting on that panel which will decide things on behalf of the consumer, which will invariably screw the consumer.  None of you want to know what the real problems are.  You are just looking for a scape goat.  The ISP is the easy target.

You have made that clear by continually, and incorrectly, generalizing an aspect of service, in light of the facts, and avoiding the real questions that need to be asked and answered.

Sorry, but that's what you assume. Just because I'm not all cuddly when we talk about ISPs doesn't mean I'm for some government control of internet.

So to make it clear:
I do believe ISPs are screwing customers big time,
and I believe the solution is more competition, not more control.

Hence, in my opinion, FCC shouldn't fiddle much with anything but assure competition in all (communication) markets.

[Skuzzy]

I tried earlier to understand that, but still, I'm not sure what you mean with that. In networking, speed was always throughput (bandwidth), and it still is.

1Gbps connection -> 1Gbps bandwidth -> 1Gbps throughput[snip]</quote]

Sorry, I have other committments to deal with today, so I will try to explain this.

The physical speed of the connection represents the best case data throughput, but it does not represent the actual throughput.

Like a CPU that runs at 3Ghz.  Guess what, when an internal cache miss occurs, the aggregate throughput drops like a rock.  Same with a network connection.  Protocols slow the actual data throughput to something that is always less the the actual speed of the connection.  Get a retry on a bit error and the data throughput drops like a rock, while the speed of the physical connection never changes.

I take something back, or I should augment something.  I have been talking strictly about wired connections.  If we throw wireless into the mix, then all bets are off as wireless always varies its connection speed to reduce bit errors in the stream.

Have to go.  I'll try to get back to this tomorrow.

Does that help?

[Masherbrum]

Control most avenues of Information Sharing (largest being the Internet), you have a grasp on the Population.   

I agree with everyone who is intelligent enough to NOT have a box installed to test.