and "balanced"
Everything here is my opinion. I do not speak for your employer.
January 2016
February 2016

2016-01-03 »

A pretty well-done report on broadband services, via the FCC:

GFiber isn't included, probably because we're too small.  If we had been, it would be a severe outlier and probably mess up their methodology:

  • Our Internet is ~10x faster than the fastest results in the survey

  • The survey doesn't even mention wifi, which for the ISPs shown, probably isn't the bottleneck in most cases for most users.  For GFiber, it's almost always the bottleneck.

  • They provide some kind of client device with test software that talks to m-lab for measurements.  Based on my own experiences, if this device is anything but an x86 (unlikely due to costs), it wouldn't be able to even saturate a gigabit service like GFiber.

We really are solving problems different from the ones everyone else is solving.

Separately, one major flaw in the report: it tests speeds to the nearest m-lab server.  That tests the "last few miles" bandwidth to your home, but an increasingly essential piece is your connection to the "real Internet."  Verizon FiOS does really well in this report, but I bet it wouldn't if we could evaluate the quality of their peering connections to eg. Netflix.

I'm CEO at Tailscale, where we make network problems disappear.

Why would you follow me on twitter? Use RSS.

apenwarr on gmail.com