The Public Interest Advocacy Centre has issued a report that concludes that Canadian ISPs need to provide better disclosure about Internet speed and performance claims.
PIAC Calls For Better Disclosure on Internet Speeds
5 Comments
Recent Posts
The Law Bytes Podcast, Episode 187: Jeff Elgie on What the Bill C-18 Deal With Google Means for the Future of the Canadian News Sector 
Skillful Negotiation or Legislative Fail? Taking Stock of the Bill C-18 Deal With Google 
Salvaging Bill C-18: Government Upends Legislation To Bring Google Onside the Online News Act 
Accountability and Antisemitism: The Canadian Heritage Committee Needs To Step Up 
The Law Bytes Podcast, Episode 186: Andy Kaplan-Myrth on the CRTC’s Last Ditch Attempt to Fix Canada’s Internet Competition Problem
Good.
I don’t think they should be advertising seperate download/upload rates either. If the Internet is participatory, the upload speed is just as important. ISPs just want to eliminate uploading.
I test Rogers Internet speed regularly in Moncton and normally get somewhere between zero and 4 Mbps while Rogers claims that it is always over 20 an normally over 30 Mbps
Fishhead, one of the wireless providers that I could contract with has different upload vs download speeds… if they advertise one only, which would it be?
On the PIAC submission, there is a couple of competing concepts here that not all users of the ISPs fully understand. The rates that the user measures can be impacted by how the testing is done. For instance, if I were to make available a speed tester on a machine on a dialup connection, someone using that service from a 4 Mbps network would see a maximum rate of dialup rates (in my case, 26,400 bps), simply because my machine at home is the bottleneck. What the ISP should do, in my opinion, is advertise the speeds available for what is under their control; this means the physical connection plus any slowdowns that they put in their network. That doesn’t mean that a customer will see that kind of throughput…
Next in stores…
… milk cartons labeled “Up to 1 liter”.
Traffic shaping is the big problem. It is extremely difficult to detect so all ISPs find it advantageous. Note that there are many appliances available off the shelf for an ISP to buy and they don’t need to do this work directly. In fact the ISP won’t know the proprietary algorithms that are being used – it’s a black box to everybody except the seller thereof.
I mention this also because it has an impact on speeds that is both good and bad. Sometimes it is a good thing to even out the bandwidth between users, but mostly it is just a way for the ISP to oversell their own supply of bandwidth. In any case, I recommend that people who are checking their speeds, don’t do it with a “speedtest” website. Do it with your own know sources. You will be surprised just how radically different the speeds are when you use different protocols. That’s one way you will notice that it’s not just a straight pipe – it is a minute-to-minute moderated one.