Move out of incubator

[normanmaurer]

The QUIC codec has proven to be stable enough to move out of the incubator. We should make this happen.

[DanielAlmeidaJoao]

Hello!

I developed a simple Netty-quic client-server application in which the server sends a 3GB file to all the clients, and I noticed that as the number of clients increase, the clients really take too long to receive all the file. In the beginning, the event ConnectionIdleTimeout was being triggered, which left me baffled because the receiver were still to receive bytes of the file; I increased the value of the timeout, and then all the clients received all the file, but taking a really long time compared to the TCP version. I think there is something going on with the congestion control algorithm that QUIC is using; I tried RENO, CUBIC, and BBR. I tested this in the servers of my faculty, a high speed private network.

[normanmaurer]

It's hard to say what's going on here without any more details ... That said this has nothing to do with moving it out of the incubator.

[DanielAlmeidaJoao]

Without being rude, what more details do you want? Implement a simple QUIC and TCP client-server application in which the server sends a 3GB file to all the clients at the same time and compare the results of QUIC vs TCP.