On 8 January 2015 at 10:56, Greg Wilkins <gregw@xxxxxxxxxxx> wrote
[heaps, which I will snip].
I agree pretty well with what Greg's written, and just want to add one incidental outcome that may be considered a positive: by defining a mechanism to allow HTTP/2 traffic to flow over the existing web, using existing HTTP/1 ports (even if most browsers will only use HTTPS ports), we've at least paved a way to allow other/future protocols to do the same.
I agree with James Snell's earlier sentiment as well, that HTTP/2 should aim for Experimental, rather than Standards track. We've bought the upgrade path (sorry, the ALPN-based TLS-as-an-opaque-tunnel path), so now we're able to experiment with new/improved web protocols, and this is the first such experiment.
Cheers
Matthew Kerwin
http://matthew.kerwin.net.au/
http://matthew.kerwin.net.au/