> The QUIC protocol implements its own crypto-layer so does not make use of the existing TLS 1.2. So in addition to proving lower layer solution to things like

> The QUIC protocol implements its own crypto-layer so does not make use of the existing TLS 1.2. So in addition to proving lower layer solution to things like network congestion they also have to prove their crypto layer? Sounds like an equally large task, if not larger. And even if all TCP implementors decide to adopt the next task is to make servers and clients adopt the crypto layer.