> E.g. the claim that all ISPs simultaneously voluntarily decided to throttle Youtube to reduce foreign traffic
> The information that all ISPs voluntarily decided to throttle Youtube is implausible
> Also the information about mobile connection not being throttled
Why are you trying to build a strawman? That's not what I said. I've said "google kept abusing backbone networks" (e.g. IEXPs), which obviously means it's a matter of the Main Radiofrequency Centre, since it involves nation-wide infrastructure - not some "ISP volunteering".
And I’ve never said that “mobile connection is not being throttled”. In fact, I am stating exactly the opposite, pointing out that traffic shaping is an inherent feature for a mobile ISP. In contrast to broadband, where no one bothered with deep traffic manipulation before, so an ad-hoc throttling solution (yes, typically simply reusing existing law enforcement integrations) works like shit.
> This is not true. The connections to googlevideo are throttled by government-operated DPI, not by ISPs. You can verify this by sending following request from a Russian residential or mobile IP address to a Russian hosting provider Selectel:
One does not need a synthetic test such as yours. One can simply try playing a video from the same browser, switching connection between broadband connected Wi-Fi and a mobile hotspot and notice that broadband doesn’t seem to be working properly, but mobile actually works, even if it’s not Full HD. How come? Does your hypothesis regarding “not by ISPs, but by government-issued DPIs” explain the variance in ISPs behavior? No, it doesn’t. Just as it doesn’t explain why “blocked” YT seems to be “blocked” completely different from your typical weed growers forum. It works differently from how you imagine it.
> Why would they throttle the speed to allow their competitors to lure away their clients?
Speaking of which, apparently some broadband ISPs are now trying to implement throttling properly to give them an edge over the competition: https://www.kommersant.ru/doc/6919868
> The information that all ISPs voluntarily decided to throttle Youtube is implausible
> Also the information about mobile connection not being throttled
Why are you trying to build a strawman? That's not what I said. I've said "google kept abusing backbone networks" (e.g. IEXPs), which obviously means it's a matter of the Main Radiofrequency Centre, since it involves nation-wide infrastructure - not some "ISP volunteering".
And I’ve never said that “mobile connection is not being throttled”. In fact, I am stating exactly the opposite, pointing out that traffic shaping is an inherent feature for a mobile ISP. In contrast to broadband, where no one bothered with deep traffic manipulation before, so an ad-hoc throttling solution (yes, typically simply reusing existing law enforcement integrations) works like shit.
> This is not true. The connections to googlevideo are throttled by government-operated DPI, not by ISPs. You can verify this by sending following request from a Russian residential or mobile IP address to a Russian hosting provider Selectel:
One does not need a synthetic test such as yours. One can simply try playing a video from the same browser, switching connection between broadband connected Wi-Fi and a mobile hotspot and notice that broadband doesn’t seem to be working properly, but mobile actually works, even if it’s not Full HD. How come? Does your hypothesis regarding “not by ISPs, but by government-issued DPIs” explain the variance in ISPs behavior? No, it doesn’t. Just as it doesn’t explain why “blocked” YT seems to be “blocked” completely different from your typical weed growers forum. It works differently from how you imagine it.
> Why would they throttle the speed to allow their competitors to lure away their clients?
Speaking of which, apparently some broadband ISPs are now trying to implement throttling properly to give them an edge over the competition: https://www.kommersant.ru/doc/6919868