2 ms·

What I'm seeing from TLS 1.3 discussion so far is that it's got improvements, big ones, but it can't really be TLS 2.0. Backwards compatibility with existing us

What I'm seeing from TLS 1.3 discussion so far is that it's got improvements, big ones, but it can't really be TLS 2.0. Backwards compatibility with existing uses and deployments is a big - but necessary - ball-and-chain around its foot. We could do a lot cleaner by doing a new protocol from scratch. But that is out of scope for the TLS WG. We'd need a new WG. And if we did, would anyone use it? Could we avoid repeating the mistakes of the past?

JoachimSchipper 12y ago
> Could we avoid the mistakes of the past? We're a lot better nowadays at building cryptographic protocols than when SSL was first developed, and SSL was never a particularly great protocol. One could also hope - but probably not expect - that we have learned some painful lessons from the security impacts of SSL's complexity.