Europeans Needed To Create Broadband Performance Measure 47
An anonymous reader writes "The European Commission has launched a project to recruit 10,000 volunteers across Europe to measure the performance of their broadband connection over two years. The trial, believed to be the world's biggest follows similar projects in the US and the UK, run by the EC's partner SamKnows. The data collected will be used to plan the next generation of services. Those interested in signing up to take part can do so here."
Re:So... (Score:4, Interesting)
You're supposed to put all your devices behind the "Whitebox". Using your solution would give them skewed results, according to their terms:
The Whitebox should be placed in between your existing router and your networked computers. Any devices that connected via ethernet cable to your existing router should instead connect to the Whitebox. This ensures that the SamKnows device is always aware of the network being used and will never run tests at a time when you require your full bandwidth to be available. (Whitebox faq [samknows.eu])
Re: (Score:2)
But what would the point of participating in the study be any more, if you reported improper data?
That's like choosing to volunteer in a study where they would measure time people spend in front of TV, and then you would do everything to watch it secretly at your neighbor's so that you don't look like a sucker.
Re: (Score:2)
Re: (Score:1)
But there's a flag on the front page, must be legit.
you made my day better :)
Sam Knows You (Score:3)
samknows.eu is the URL. Which is sort of a creepy sign right there I think.
Will these folks be blessed with any rights of immunity, against anything that is dug up? Might be a good method for a TOR onion router perhaps.
Re: (Score:2)
These were my thoughts as well, but FTFA:
Actual tests: (Score:2)
From the samknows FAQ:
The SamKnows Whitebox currently performs the following tests:
Multi-threaded HTTP download speed test
Multi-threaded HTTP based upload speed test
Availability of the connection
Jitter
Latency (both ICMP and UDP)
Packet loss (both ICMP and UDP)
DNS query resolution time
Re: (Score:1)
Letting it test throttling would be easy by just using up your limits before giving it the chance to measure.
Re: (Score:1)
Upload speed in UK tends to be a small fraction of download speed
It's similar in Poland. Usually something like 3:1, so 6 Mbps down and 2 Mbps up
I'll bet it doesn't test for throttling because it'd have to push through too much data and some people with data-caps would complain.
from their website: "NOTE: Our Whiteboxes download approximately 3GB per month and upload around 1GB. If you’re on a product with a low usage cap then we’d advise against signing up, or at least informing us beforehand so that we can apply a different testing profile." So at least you've been warned
Re:Actual tests: (Score:4, Insightful)
So in other words those of us with abysmally low bandwidth connections still being touted as 'broadband' would ruin our net experiences completely (as in, worse than not being able to stream even the lowest resolution videos) if we signed up to prove that our connections are really this horrid.
Re: (Score:2)
No, because the tests are only run when no network activity is detected.
Re: (Score:2)
Our ratios are far worse, mine is 10mbit/1mbit, when I was with Virgin Media it was 10 / 0.45mbit... when it wasn't being throttled, once throttled my upload with VM was as little as 20-30KBytes/sec.
The cynical part of me thinks artificial tests are pointless, the ISPs will investigate the boxes and then skew the figures to suit.
Re: (Score:2)
"I'll bet it doesn't test for throttling because it'd have to push through too much data and some people with data-caps would complain."
They warn in their requirements that the tests take about 3 GB monthly so ppl with a data cap shouldn't register or advise them in advance so they'll run a different test program.
Samknows have got suspiciously relevant... (Score:2)
You know those web sites which have existed as small, specific operations doing one or two things well for as long as you can remember? Always looking a little amateurish but getting the job done.
And you know how one day suddenly their web site goes all corporate and the old services are just sidelines? Looking slick but now seemingly just providing expensive consulting services.
And the following year they're doing something absolutely huge, like the fat kid known for baking tasty cakes who has just been gi
Re: (Score:2)
Its called EU grand money. They applied for a grand, got it and now spend like there is no tomorrow.
Re: (Score:2)
By that measure, anyone who has a care that goes from A to B should be happy, no matter how much they paid - whether or not the thing jerks like a kangaroo down the road, puffs out black smoke, or consumes twice as much petrol as others.
I believe this experiment is more about getting some numbers that can be used to combat the "Unlimited" downloads, "Up to 10Mbps" etc. advertising. When you have to provide a MINIMUM 10Mbps service, you start having to do things like: fixing your system to provide timely ac
Re: (Score:2)
How the hell we are still "measuring" throughput, packet loss and RTT in 2011????
My wife want to know if she can "WATCH YOUTUBE" , "SEND A PICTURE VIA EMAIL", "PLAY HER FAVORITE SHOW ON TV".
It's like the PC. People don't care about how many GHz, GB, Mcolors the box has. They care about what they can do...
It's like for cars: who's interested in checking the engine, knowing which brakes and knowing the maximum speed...
It's really time to get rid of this sticky measures...
throughput: Can it play a 1080p streaming video, or will it give up at 480p?
RTT: How well will my real-time online game work?
Re: (Score:2)
You want a car analogy? Here, Samknows and mlab have already made one for you: http://www.youtube.com/watch?v=RnIVMfBP4So [youtube.com]
measurement lab (Score:2)
http://www.measurementlab.net/ [measurementlab.net]
Re: (Score:2)
Samknows is one of the partners in this project.
Confidence in the EU (Score:2)
I have some confidence that this test will actually be used to improve services to customers. For example, the EU over the last couple of years has come down hard on the telecom industry, forcing them simply to become cheaper and to improve services.
It seems that although the EU takes 1984 as a guidebook rather than an example, they at least realize that its citizens must have an affordable and good quality information infrastructure if they want anything to eavesdrop on :-)
Re: (Score:2)
Speed test sites and consumer experience comparison sites have been around everywhere for years. This is an actual hardware device that should provide objective data to the EU
Dear European Commision: suck on a turd (Score:2)
This is just yet another swill trough project to keep worthless unelected bean counters in paid lunches for two years, plus the half decade that it'll take to draw up the Directive that will inevitably just mandate whatever Germany plans to do anyway.
So take your little spy boxes and hand them out to Frankfurters. They'll happily plug them in, they love obeying orders.
Re: (Score:2)
Oh do shut up. You haven't been paying attention if you haven't noticed how EU regulation has improved quality of service and pricing in the telecom sector in the past years.
And your derogative remarks on germans are really sad. Go get a life.
Re: (Score:2)
OK, Mr. Coward, how would YOU implement a way for the box to detect when the line is idle so that speed measurements can be made without interference from other network traffic?
Re: (Score:2)
Use SNMP to find out the current upload/download from the actual router. The same way Cacti does. Once the speed drops below pre-set level, try to run the test.
Re: (Score:2)
You would expect average people to configure SNMP on their routers?
Plugging in two cables seems more user-friendly.
I bet it is hard enough to find participants to this study already.
Re: (Score:2)
Isn't it on by default? Last time I configured a hardware (consumer level) router it had SNMP enabled by default IIRC.
Plugging in two cables works but not for my nonstandard configuration - I use a PC as a router and it also runs uTorrent (one other PC inside the network does it too). I also doubt that their device will be able to pass full gigabit (to/from my router PC).
I would participate but a Cacti graph of my actual usage would be more useful - my connection is almost never idle.
These statistics already exist (Score:1)
Why waste a fortune on measuring internet speed by HARDWARE (seriously, what's the point of that?) instead of using something like these numbers [speedtests.net] ?
Re: (Score:1)