Sébastien Ros

The request is much smaller than that on plaintext when we use work.

Locally I think I got 9.8 Gbs with `iperf` and measured a max of 1.5M packets per second. Any benchmark that is over 1.5M is obviously and correctly using pipelining.

Yes I know ;) hence the comment

I will ask our colleagues who manage the lab, but as of last week I know they had not tried it yet.

Idea: Add a new dimension with the number of cores (1, 4, .. MAX). That would show how some frameworks behave in constrained environments.

And `Connection: keep-alive` is also implicit in HTTP 1.1

> pressing issues that may have a negative impact on the next round Can you share the details if we want to repro the environment?

Main benefit is that users who have to use a strong named assembly can use yours. https://docs.microsoft.com/en-us/dotnet/standard/library-guidance/strong-naming Until nobody is using it anymore I'd recommend to ship the package strongly...

> when I replace that with an object with a field Can you show what you mean?

Why a template? Why not a delegate, which could provide the context to render something else directly, or log.