Simone Basso

> Packet loss is desired by lots of people. For ndt7, I was leaning towards telling people about %-age of bytes retransmitted from TCPInfo. It is not the packet loss...

As mentioned in another issue, OONI will switch to ndt7 soon, therefore it is becoming less relevant to us to focus on ndt5. I am again not sure about closing,...

BTW, issue #201 seems to be a case where the RTT is actually 20-ish but BBR thinks it is much smaller, therefore, it would be beneficial to see what applevel-RTT...

@darkk considering how other speed tests work, I think may be preferable to include an application level `/ping` phase to the experiment. I'm significantly less happy about modifying the expected...

Oh, noes, conflicts! I am going to solve them now!

@pboothe I have fixed conflicts, can you perhaps maybe take another look?

The download and the upload code have been refactored to look basically the same. The next step is to make sure we're sold on the design. If so, it's time...

Unassigning myself as this is not worrying us too much for now. (We are currently using xenial for CI, so this is still an issue, but also a feature 'cause...

> When a maximum comes from a model, then it makes sense that the model updates over time. It also makes sense that an expected value would change over time....

@elioqoshi indeed: I just meant to say that figuring out that is also part of that job!