AT&T Won't Explain Meter Inaccuracy, Claims Data Proprietary

If money is involved, then they may be running afoul of various state "weights and measures" laws.

Here in Massachusetts (although we are Verizon turf), the weights and measures laws are somewhat strict such as a store can get a $100 fine if the price on the shelf and the price on the scanner does not match. They also inspect gas pumps to make sure that you are getting a true gallon and they inspect scales that are used for trade.

Maybe some class action lawyer could find a way to convince a judge to apply the various weights and measures laws since they do charge for overages.

When you pay for a gallon of gas, it should be a true gallon of gas. And in Mass, the price on the pump and the price on the signs have to match or you are entitled to the lower price.

....carries a lot of juice. Slashdot readers are given a big FU.

For their standard DSL offering, they count the overhead in the bandwidth calculation. When you factor in ATM, PPPoE, and TCP/UDP overhead, that racks up to about 16-25% overhead depending on the size of the average packet sizes you get using the connection.

To add insult to injury, this also affects the speed you get as they don't account for the overhead in the "up to" speed. So, if you have a plan of 3Mbps/512Kbps, the MAXIMUM you could ever get is about 2.5Mbps/415Kbps. So, you can never reach your "up to" speed, but they also knock you on the meter because of the overhead.

They recently added a $3 increase on all of the classic plans, because you know it costs more to deliver those same bits they were doing before.

This might be a great idea for our wonderful electric company LIPA. Come up with a secret proprietary way for calculating electric usage. The meters could be set to run on battery power so they are always spinning, even in a blackout! Plus they could multiply usage by a secret number, "e" perhaps, to up the usage.

Sounds like AT&T is making a good case for the government to step in and regulate their meters under the same weights and measures rules that apply to scales, gasoline and gas and electric meters.

AT&T has the "balls" to advertise a 2 year contract with no increases yet every month the TOTAL bill for my services keeps going up due to added "Just because we can" charges. They sent me the letter saying that CAPS are in place but have yet to say how I am supposed to know if I am approaching the CAP over it or how far over the CAP etc. Somehow this seems to be an extortion scheme of sorts. Thankfully I've fulfilled my contract terms so have no early termination fines to pay should I finally say enough is enough. America where Corporate dollars buy the best protection from the law.

It is always a gallon of gas, but due to temperature fluctuations, it may not be the same mass due to higher densities in cold temperatures, and lower densities in warmer temperatures.

ATT has the best protection from the law that money can buy so there is no danger from Federal (or state)-maybe locally there's a chance to get them on weights and measures violations but I am not holding my breath.

they need to be:

...documented exactly how they work
...audited by a third party
...and available to the consumer in real time

That's how my electric meter operates. That's how my water meter operates. ISPs who try opaque, inaccurate, undocumented "metering" need to be taken to task.

I am an ATT DSL customer in Michigan. After seeing numerous posts about the inaccuracy of ATT's metering method, I decided to run my own experiment. I turned off my modem at 11:30 PM on October 19, 2012 and turned it back on at 12:30 AM on October 21, so the modem was completely off for the entire day of October 20. Despite the impossibility of usage during this period, ATT says I used 340 meg down and 30 meg up for October 20. Maybe the metering is in another time zone, but draw your own conclusions.

1GB =1000mb

Bandwidth charges should also obviously be relative to bandwidth costs. My electric company is only allowed to pass on supplier cost with zero profit

... Local Time Zone as well.

I've said it before, and I'll say it again. Why is it so hard for companies like AT&T to get this right?

The example I use is Smoothwall Express. For those that don't know, its a Linux version of a firewall/router. Built into this, is a real time bandwidth measurement tool, complete with charts, that update in real time. Anytime I open the main web page for this, I can see exactly how much bandwidth I have used currently at the time the page loads, how much I have used for the day, and how much I have used for the month (both inbound and outbound numbers).

In fact, here is what mine shows right now

Current: 44.5 Kbit/s / 56.4 Kbit/s (Out / In)
Today: 79.9 MB / 982.1 MB (Out / In)
Month: 2.8 GB / 32.0 GB (Out / In)

What I'm getting at is that the people that developed Smoothwall figured it out, and they offer it to the end user FOR FREE!!!

Bottom line the technology exists, and apparently is not that hard to implement.

Just think of the business that Smoothwall (or companies like it) could get if they worked with AT&T and the other ISP's to implement this on their network. Its a win win!

--Brian

Which is why Costco pumps have a label on them now stating that they sell gas by volume not energy content. Some ambulance chaser thought they could win a lawsuit because the pumps weren't temperature compensating.

Why isn't it 1024Mb? I thought HD manufacturers were the only one's who favored marketing over mathematics?

It is not hard to get it right, they just do not want to get it right so they can overbill their customers. Need to pay for those Executive bonuses and shareholder dividends any way possible. It will only get worse as the State regulatory bodies keep bending over backward to ease regulations on Telcos.

Small fluctuations are expected. If someone's data is off 20% that's an issue. Even 5% is an issue.

ads and spyware add up to the cap too and if you leave your modem on 24/7 and if you use vonage it goes to the cap

WRONG. OK first B is BYTES and b is BITS So no 1000 megaBITS does not equal 1 gigaBYTE. 1000 megaBITS =125 megaBYTES

Also it's 1024 MB = 1 GB

In Canada, the gas is sold volumetric corrected to a temperature of 15C, which is the typical temperature of the gasoline in the underground tanks at the depth they are buried. Since the path of the gasoline in the above-ground portion of the pump is short there is no appreciable need to correct for the ambient atmospheric temperature.

You *might* gain some energy content advantage is you filled up in the winter just after the underground tanks were replenished, if the tanker had traveled 200 miles in -20F weather and chilled the gasoline.

I'm not defending HSI metering but the electric company is an allowed monopoly. Because of this, there are regulations in place that define how they can charge for their service.

Although I believe cable and telcos are really close to being monopolies, at this moment they aren't classified as such and are not regulated. I agree that they should charge fair prices for their products and services but right now, they can charge whatever they want and do whatever they want until we can legally provide sufficient evidence that they act as a monopolies and should be regulated.

Unfortunately they have a lot of money and firewalls in place to protect themselves from such consumer action. If possible, I'd favor figuring out a way to foster more competition. Unfortunately they employ a lot of money and firewalls to protect against competition. Kind of sounds like the record is skipping, doesn't it?

I get that part.

However, you cannot tell me that every IT person who works for said companies in the development of these meters has not raised the question during testing like "I know I downloaded 500 MB, but the meter is showing 800 MB. This is not working right, and we need to fix this before we move it to production".

Are you saying that if a member of the development team raised that issue, it either fell on deaf ears, or the person was let go? I know if I was working on the development of said application, and during the testing it was not working the way it was suppose to, I would bring it to the attention of whoever was in charge, and demand that it gets fixed.

The bottom line is you do not release software into production if it is not working the way it is suppose to.

Looking at it from another standpoint. Someone (probably a team of people) wrote the software that is in place in thousands of cash registers that scan the item, and add the price to your bill. If during the testing of that software, say an item was scanned and it came up a different price. Just think if that "mistake" was released without being fixed? The team that was in place on that project had to make sure that if the item scanned was $5.00, the program read the bar code correctly and charged you $5.00. Not $5.50, not $4.79, but $5.00. Its simple logic!

To me, the same rules of development and testing come in place here. Either code the meters to work, or don't put them into use. Its that's simple.

--Brian

quote:

---

Are you saying that if a member of the development team raised that issue, it either fell on deaf ears, or the person was let go? I know if I was working on the development of said application, and during the testing it was not working the way it was suppose to, I would bring it to the attention of whoever was in charge, and demand that it gets fixed.

The bottom line is you do not release software into production if it is not working the way it is suppose to.

---

That is exactly what I am saying/implying. Your "demand" to get it fixed will end up in you losing your job and/or just be laughed at. It is solely the bottom line that matters to the Managers and Executives. Would suggest getting a job at a Telco if you want to see what it is like, but as they outsource a majority of the work it is harder to get an actual "in house" programming job with the Telcos now than it was 10 years ago when I worked at SBC/Ameritech.

I'm sending this letter to the GA Secretary of State. I urge you all to do the same. In Georgia, weights and measures is handled by the Dept. of Agriculture. Each state is different.
----------------
You may be aware that internet service providers and cell phone companies have started imposing data caps on their customers usage, with high overage charges when they exceed the cap.

I am concerned that there is absolutely no oversight of the accuracy of the meters they use to bill customers. There are numerous reports of inaccurate meters, and in fact, some ISP's have temporarily discontinued the caps due to inaccurate metering.

There many reports of meters reporting data usage while the customer's equipment was unplugged or there was a power outage. Other technical users have measured their own data usage at their router and the usage results are different, and usually lower than what the providers are claiming.

When we consume gasoline, we are insured that we are getting a gallon because the pumps are inspected by the State and sealed, and consumers can verify the accuracy on their own. When we buy food, scales are also inspected for accuracy and consumers have easy access to other scales to verify their purchase.

When we buy data, there is no such oversight, inspection or guarantee that these usage meters are correct, and no way for the average consumer to verify that they are being billed correctly. This provides an open opportunity for providers, who are notorious for nickel and dime'ing their customers with so called "regulatory" non-fees, to overcharge customers with inaccurate data usage meters. More information on this can be found at www.dslreports.com

The laws of Georgia must be brought up to date to embrace the digital age and insure that meters that are used to measure and bill for data usage are accurate and verifiable. I urge the State to act quickly, because the use of voice over internet, video and other data consumption is expanding at lightning speed, and even small overcharges amount to millions of dollars of illicit income for the providers.

1024 MiB = 1 GiB. XiB/Xib is measured in 1024 increments not the 1000 (decimal) increments that are usually used in XB/Xb claims to inflate the speed (or deflate the volume) measure. Unless the measure is stated in Xi units (or there is a disclaimer that X is 1000 based) you can assume that you are getting a 1000 based measure not the accurate 1024/Xi one.

I agree. While I'm typically not for 'regulation', I would have to state that if companies that are using meters for profit, those meters should be certified for use by an independent company.

Eg. Household electric meters, gas meters, water meters, grocery store scales, gas station pumps, etc.

If AT&T is going to bill by the byte, then their billing must conform to a standard, and their devices must be certified for accuracy.

I should not be paying for external broadcast packets that do not make it to the LAN side of the modem, or corporate sniffers.

If you bothered to read the AT&T ToS, you would know that they sell "sync" speed. No matter what you measure down, if your modem holds sync at 3008 (for "Pro"), you are getting what you pay for at 2.5Mbps down.

too lazy to press the shift key duh.

The difference is that weights, volumes, lengths, etc. are physical measurements that can vary with temperature, humidity, physical wear, etc. so these need a common reference preferably tied to some fundamental physical measurements.

Bits on the other hand are a discrete invariant quantity so there is no possible dispute about the measurements of a bit.

What is more dispute-worthy here is the accounting behind the billing: what gets counted, why, where, when and how.

A lot of the billing errors may either be due to AT&T's systems having the wrong port associated with a given account and billing the wrong person. Another possibility is your IP address getting spammed by portscans and other activities which AT&T cannot really tell apart from regular traffic in which case the only way to avoid it is to turn off the modem when not in use. AT&T also cannot really tell packets dropped at your end and even if they could, I doubt there is any reason for AT&T to offer refunds for data mishandled at your end just like the gas station doesn't refund you if you spill gas.

Usage billing is all in the accounting. The likelihood of "measurement" errors from reading byte counters is practically zero.

It's the New Business, "We'll charge you what we want, when we want." model.