Question about timing the server processing

[haggen]

I can see in the line below that what you call ServerProcessing is the time between when the connection's been established and we get the response's first byte.

https://github.com/rogerwelin/cassowary/blob/master/pkg/client/load.go#L117

But according to Go's httptrace docs there's a hook called WroteRequest that seems to me more accurate to the start of the remote server process.

I remember digging a little deeper into the request's lifecycle when I wrote a similar feature but I'm no expert so it's just me being curious.

Why did you go with GotConn hook instead?

Thanks in advance.